Up-regulation of MicroRNA-21 Mediates Isoflurane-induced Protection of Cardiomyocytes

CARDIOVASCULAR disease is one of the leading causes of morbidity and mortality worldwide. Myocardial ischemia–reperfusion (I/R) injury alone accounts for approximately 450,000 deaths per year in the United States.¹  Therapies to reduce infarct size are predicted to increase patient survival rates, as final infarct size after an I/R event has been directly correlated to patient prognosis for future cardiac events.² 

One way to reduce the size of infarct before an I/R injury is through the use of preconditioning. Exposure to volatile anesthetics, such as isoflurane, has been shown to decrease myocardial infarct size in vivo and increase cell viability after oxidative stress in vitro.^3–7  Previously, we have shown that anesthetic preconditioning with isoflurane affected expression levels of nitric oxide synthase and heat shock proteins, as well as decreased levels of reactive oxygen species, resulting in increased cell viability.^3,8  However, knowledge of the underlying mechanisms remains incomplete. In this study, we investigate whether microRNAs are playing a role in isoflurane-mediated protection of cardiomyocytes.

MicroRNAs are 18- to 22-nucleotide sequences that suppress protein expression through degradation or suppression of messenger RNA (mRNA). MicroRNAs have previously been linked to the cardioprotective effects of ischemic, hypoxic, and heat shock preconditioning.^9–11  In 2009, the Zhang laboratory showed that overexpression of miR-21, similar to the up-regulation observed during ischemic preconditioning, significantly decreased infarct size after acute myocardial infarction.⁹  The Abdellatif laboratory has shown that miR-199a is a key regulator of hypoxia-triggered pathways, through actions on hypoxia-inducible factor-1alpha and Sirtuin 1.¹⁰  Micro RNAs -378 and -711 are suppressed during both ischemic and heat shock preconditioning, resulting in increased the expression of heat shock protein 70 and increased cell viability.¹²  So far, no studies have been done on microRNAs regarding anesthetic protection of cardiomyocytes.

Recent studies have also shown that isoflurane, sevoflurane, and other anesthetics can influence microRNA expression profiles in the liver and in neurons,^13–15  suggesting that isoflurane exposure may influence microRNA in cardiomyocytes as well. Thus, in this study, we used in vivo and in vitro rat models to investigate, for the first time, the role of miR-21 in isoflurane-mediated preconditioning. We hypothesized that isoflurane protects cardiomyocytes against oxidative stress through up-regulation of miR-21.

All experimental procedures and protocols were reviewed and approved by the Institutional Animal Care and Use Committee of the Medical College of Wisconsin (Milwaukee, Wisconsin; approval number 2602; 12-19-2011), and the investigation conformed to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health (National Institutes of Health Publication No. 85-23, revised 1996). All animals were housed with free access to food and water. Male Wistar rats (Charles River Laboratories International, Inc., Wilmington, MA) weighing between 300 and 400 g were used for in vivo microRNA screening. For in vitro microRNA screening, time-pregnant Sprague-Dawley female rats (Jackson Laboratory, Bar Harbor, ME) were housed in separate cages until gestation was complete. Male B6129SF1/J wild-type (WT) mice (Jackson Laboratory) and miR-21 knockout mice (Jackson Laboratory) weighing between 25 and 29 g were used for in vivo functional studies.

Neonatal cardiomyocytes were isolated as described our in previous studies.^8,16  Left ventricles were harvested from litters of 1- to 2-day-old Sprague-Dawley rat pups. Two litters were pooled per experiment to decrease variation within species, as well as to ensure that enough tissue was isolated for Western blots and time course studies. Cells were isolated using serial digestion with collagenase and pancreatin (Sigma-Aldrich, St. Louis, MO) and plated on gelatin-coated dishes at 1.0 × 10⁶ cells per 60 mm dish or 2.7 × 10⁵ cells per well in a 12-well plate.^* Cells were cultured under these conditions for 1 week before isoflurane exposure.

Each dish of neonatal rat cardiomyocytes was randomly assigned to the isoflurane-exposed or control group. Isoflurane (Baxter Healthcare Corporation, Deerfield, IL) exposure was performed in a sealed chamber placed in a 37°C incubator for 30 min at a concentration of 0.5 mM. The same protocol has been established in our previous studies and mirrors a clinically relevant dosage.^17,18  Control cells remained in the same incubator throughout the experiment. Isoflurane-exposed cells were removed from the sealed chamber and placed with control cells for 15 min to allow for washout of the volatile anesthetic before RNA isolation.

Each rat was randomly assigned to the isoflurane-exposed or control group. Rats were exposed to 30 min of 0.5 mM of isoflurane in a sealed plexiglass unit. Isoflurane was delivered using a VetEquip funnel-fill vaporizer (VetEquip Inc., Pleasanton, CA). A 32% O₂ environment was monitored using the Criticare Systems Poet IQ anesthesia gas monitor (SOMA Technology Inc., Bloomfield, CT). Control animals were exposed to 32% O₂ for 30 min. Left ventricles were harvested after a 15-min washout period in the original housing units. Left ventricles were pooled in groups of three for each experiment to decrease variation between groups of animals on the 12 arrays used for microRNA screening.

Total RNA was isolated using the miRNeasy Mini Kit (Qiagen Inc., Valencia, CA). Cells were homogenized in QIAzol lysis reagent. Afterward, chloroform was added and samples were centrifuged to separate aqueous and organic phases. The aqueous phase was extracted and combined with ethanol to facilitate RNA binding to miRNeasy Mini spin columns. Samples were washed twice in buffer, and total RNA was eluted using RNase-free water. cDNA was synthesized by adding RNA to a mixture of miScript Reverse Transcriptase Mix, 10× miScript Nucleics Mix, and 5× miScript HiFlex Buffer (Qiagen Inc.). The mixture was incubated at 37°C for 60 min and then at 95°C for 5 min to convert RNA to cDNA.

MicroRNA profiling was conducted using the miFinder miScript microRNA Polymerase Chain Reaction (PCR) Array (Qiagen Inc.) according to the manufacturer’s instructions. These arrays contain 84 lyophilized microRNA primers distributed in single wells along with four housekeeping genes (RNAs Rnu-6, U87, 4.5 S-V1, and Y1). Samples without reverse transcriptase were run to confirm that no genomic DNA was present in the sample. The microRNAs present in this assay represent 84 of the most abundantly expressed microRNAs across all mammalian tissues. A total of 12 microRNA arrays were run for in vivo rat hearts and 12 for in vitro cardiomyocytes: six control plates and six exposed to isoflurane, as designed in previous microRNA screening studies.¹⁹  Differentially expressed microRNAs were validated in technical triplicate using miScript Primer Assays for miR-21 (Qiagen Inc., Lot Nos. 117241738 and 117259379). qRT-PCR was conducted using the Bio-Rad iCycler Real-Time PCR Detection System (Bio-Rad Laboratories Inc., Hercules, CA). cDNA was diluted and combined with QuantiTect SYBR Green PCR Master Mix and miScript Universal Primer using low-adhesion pipette tips (VWR, Radnor, PA). Data collected from these experiments defined a threshold cycle number (Ct) of detection for the microRNAs present in each sample. MicroRNAs were excluded from the study if Ct values were greater than or equal to 35. Primer specificity was tested by running qRT-PCR analysis with miScript Primer Assays specific for miR-21 in a solution containing only synthetic miR-21 mimic (Qiagen Inc.), as well as by size fractionation of stem-loop RT-PCR products on 2% agarose gel. Melt curve denaturation levels also occurred in the predicted 70° to 80°C range (see fig., Supplemental Digital Content 1, https://links.lww.com/ALN/B128).

The neonatal rat cardiomyocytes were transfected with doses ranging from 1 to 25 nM of locked nucleic acid (LNA)-modified anti-miR-21 (Exiqon, Skelstedet, Denmark) using the Lipofectamine 2000 transfection system (Life Technologies, Carlsbad, CA). The LNAs comprise a single ribose ring connected by a methylene bridge at the 2’-O and 4’-C atoms, allowing for the nucleotides to bind in an optimal conformation. To transfect the cardiomyocytes, LNAs were incubated briefly with the Lipofectamine reagent to allow for anti-miR incorporation into lipid bubbles. The mixture was then exposed to cells for 20 h. Scrambled LNA anti-miRs (Exiqon) were used as a control. After an exposure to anti-miR-21, cells were lysed and tested for miR-21 expression using qRT-PCR analysis as described under “MicroRNA analysis by quantitative reverse transcriptase PCR” in this Materials and Methods section.

Cardiomyocytes were exposed to 0.5 mM of isoflurane in an anesthetic chamber, as described under “Exposure of neonatal rat cardiomyocyte culture to isoflurane” in this Materials and Methods section, while the control plates were kept in a 37°C incubator. After a 15-min washout period, 50 μM of hydrogen peroxide was used to cause oxidative stress to the cardiomyocytes. After 6 h of exposure to hydrogen peroxide, lactate dehydrogenase (LDH) release was measured by absorbance of indicator dyes as calculated by the Epoch Spectrophotometer (Biotek, Winooski, VT). After media samples were taken for LDH analysis, cells were stained with PI (Life Technologies). Nuclei were stained with Hoechst 33342 (Life Technologies) at a concentration of 1:800 diluted in cell culture media, and 1.0 mg/ml of PI solution was diluted in media at a concentration of 1:500. Hoechst-positive nuclei were counted as a control using fluorescence microscopy, and PI-positive cells were taken as a percentage of the control.

Cardiomyocytes were transfected with 10 nM of programmed cell death protein 4 (PDCD4) Stealth RNAi short interfering RNA (siRNA) Select RNAi (Life Technologies, Oligo ID: RSS329752) using the Lipofectamine 2000 (Life Technologies) transfection system for 24 h. PDCD4 knockdown of mRNA expression was confirmed by qRT-PCR analysis, as described under “MicroRNA analysis by quantitative reverse transcriptase PCR” in this Materials and Methods section. Knockdown of PDCD4 protein expression was confirmed by Western blot. Negative control cells were transfected with Stealth RNAi siRNA Negative Control Med GC (Life Technologies).

miR-21 mRNA targets PDCD4 and phosphatase and tensin homolog (PTEN) were investigated using rat RT² PCR Assays (SABiosciences, Valencia, CA) according to the manufacturer’s instructions. Total RNA was extracted using the Qiagen RNeasy kit as described under “Total RNA isolation and cDNA synthesis” in this Methods, and cDNA was prepared using the RT² First Strand Kit (Qiagen Inc.). RNA was combined with genomic DNA elimination buffer and incubated at 42°C for 5 min to ensure cDNA purity. Samples were chilled on ice for 1 min and then mixed with 5× RT buffer 3, primer and external control mix, RT enzyme mix 3, and water. cDNA was diluted and combined with RT² SYBR Green/Fluorescein Master Mix and distributed evenly across the wells of a 96-well plate in technical triplicates. qRT-PCR was conducted using the Bio-Rad iCycler Real-Time PCR Detection System. Samples were exposed to an initial 95°C hot-start activation step, followed by 40 cycles of 94°C denaturation and 60°C annealing/extension phases. Data collected from these experiments defined Ct values of the mRNAs present in each sample.

Expression levels of the housekeeping genes β-actin and β-2 microglobin were used as controls to normalize samples. Samples without reverse transcriptase were run to confirm that no genomic DNA was present in the sample. mRNAs with Ct values greater than or equal to 35 were excluded from the study.

Cells were lysed and sonicated in RIPA lysis buffer (Cell Signaling, Danvers, MA) containing phosphatase inhibitor cocktail (Roche Diagnostics, Indianapolis, IN). Lysates were centrifuged at 10,000g for 10 min at 4°C. Pellets were discarded, and the total protein concentration of the supernatants was determined by DC Protein Assay Reagents Package kit (Bio-Rad, Hercules, CA). The samples were boiled for 5 min at 97°C. Twenty-five micrograms of protein was loaded per lane for sodium dodecyl sulfate–polyacrylamide gel electrophoresis gel separation and then transferred to nitrocellulose membrane. Membranes were blocked with blocking buffer (Thermo Fisher Scientific, Waltham, MA) and then incubated overnight at 4°C with primary antibodies against PTEN, actin (Cell Signaling), or PDCD4 (Rockland Immunochemicals, Inc., Limerick, PA). After the incubation of membranes with secondary antibodies conjugated to horseradish peroxidase (Cell Signaling) for 1 h at room temperature. Labeled proteins were detected with chemiluminescence detection reagent (Cell Signaling). The membrane was transferred to x-ray film and exposed. The intensity of the protein bands in films was quantified using ImageJ software,^† and the data were reported as percentage of control.

Myocardial I/R injury in vivo was produced by occluding the left coronary artery, as previously described.^20,21  WT and miR-21 knockout mice were randomly divided into isoflurane-exposed or control groups. After instrumentation was completed, all mice were stabilized for 30 min and subjected to 30 min of coronary artery occlusion followed by 2 h of reperfusion. Isoflurane was administered at 1.0 minimum alveolar concentration for 30 min via an isoflurane-specific vaporizer (Ohio Medical Instruments, Cincinnati, OH) followed by a 15-min washout period before coronary artery occlusion. Control mice were not exposed to isoflurane. The infarct area was delineated by perfusing the coronary arteries with 2,3,5-triphenyltetrazolium chloride via the aortic root, and the area at risk was delineated by perfusing phthalo blue dye (Heucotech Ltd., Fairless Hills, PA) into the aortic root after tying the coronary artery at the site of the previous occlusion.

All rats, mice, and neonatal rat cardiomyocyte cell culture dishes were randomly assigned to isoflurane or control groups, as well as scramble or transfected groups. Western blot analysis was conducted blindly, with samples divided into randomly numbered groups. Values are expressed as means ± SEM. t Test confirmed statistically significant differences between microRNAs in single groups. We used an analysis of variance test followed by a Tukey test to compare microRNAs in multiple groups. P values less than 0.05 (two-tailed) were considered to be significant. MicroRNA Arrays were analyzed using the miScript microRNA PCR Array Data Analysis Web Portal (SABiosciences), and the Benjamini and Hochberg procedure was applied to control the false discovery rate in our samples.²²  Three microRNAs (miRs-137, -142, and -196b) had a single data point greater than 2 SDs from the mean and were considered to be outliers. After these data points were discarded, the isoflurane-exposed groups were no longer significantly different from the control. All other statistical analysis was performed using SigmaPlot 12.0 (Systat Software, Inc., San Jose, CA).

Array-based analyses investigated the effect of isoflurane on expression levels of 84 microRNAs in 36 in vivo rat hearts pooled into groups of three and in vitro cultured neonatal rat cardiomyocytes from 24 litters of Sprague-Dawley pups pooled into two litters per group (for a total of n = 6 per group). Eleven microRNAs were found to be differentially expressed between isoflurane and control groups. All 11 microRNAs were up-regulated with a Benjamini–Hochberg-adjusted P value to control the false discovery rate at 0.05 level (fig. 1A).²²  Bioinformatics analysis (Ingenuity Pathway Analysis Software, Redwood City, CA) was used to nominate a candidate for further functional studies. MicroRNAs were investigated that directly targeted proteins in cell signaling pathways previously reported to mediate anesthetic preconditioning. These microRNAs were compared with the in vivo and in vitro array data, as well as against known literature.

MicroRNA-21 emerged as the strongest candidate for further functional studies, as it was linked to multiple cardioprotective cell signaling pathways, exerting effects on PDCD4, Akt, and PTEN.^9,23–25  In addition, miR-21 amplified at the lowest Ct values of all candidate microRNAs identified, indicating a higher relative abundance, and potentially, a more robust functional effect.

Samples were taken after 30 min of isoflurane exposure, as well as after 3-, 6-, and 24-h time points to determine miR-21 expression changes over time. At 30 min after isoflurane exposure, miR-21 was significantly up-regulated (219.2 ± 17.1%). Three hours after the initial up-regulation, miR-21 expression returned to baseline values (fig. 1B). qRT-PCR sample purity was confirmed by melt curve analysis and the presence of a single PCR product after size fractionation on a 4% agarose gel (see fig., Supplemental Digital Content 1, https://links.lww.com/ALN/B128).

Neonatal rat cardiomyocytes were exposed to 20 h of multiple doses of LNA-modified anti-miR-21 or scramble antagomirs. qRT-PCR analysis revealed a dose-dependent response of miR-21 knockdown at 1 nM (56.7 ± 8.1%), 5 nM (37.0 ± 4.9%), and 25 nM (2.1 ± 0.6%) of transfection reagent (fig. 1C).

Neonatal rat cardiomyocytes were exposed to 20 h of 25 nM anti-miR-21 or scramble antagomirs. qRT-PCR analysis revealed similar baseline levels of miR-21 between nontreated and scramble-transfected groups. Nontreated and scramble-transfected groups also exhibited up-regulation of miR-21 after isoflurane exposure similar to previous experiments. In anti-miR-21–transfected cells, both the isoflurane-exposed (0.01 ± 0.002) and control cells (0.10 ± 0.06%) exhibited significantly decreased the expression of miR-21 compared with nontransfected controls (fig. 1D), confirming a complete knockdown.

Measurements of cell viability after 6 h of oxidative stress with 50 μM hydrogen peroxide revealed an increased LDH release in nontransfected (442.3 ± 77.0%), scramble-transfected (478 ± 66.9%), and miR-21–transfected (481.9 ± 34.9%) cells compared with nonstressed control cells. Exposure to isoflurane decreased LDH release in nontransfected (269.8 ± 49.2%) and scramble-transfected (251 ± 20.5%) cells but not anti-miR-21–transfected (425 ± 18.1%) cells (fig. 1E).

The mRNA expression of miR-21 targets PDCD4, and PTEN were evaluated via qRT-PCR. PDCD4 mRNA expression was significantly decreased after isoflurane exposure. Western blot analysis also revealed that cells exposed to isoflurane exhibited significantly decreased the expression of PDCD4 protein when compared with controls. Transfection with anti-miR-21 increased the endogenous levels of PDCD4 and abolished the isoflurane-induced down-regulation of PDCD4 (fig. 2A).

Both mRNA and Western blot analysis revealed no change in PTEN expression after isoflurane exposure. Anti-miR-21 transfection also had no effect on endogenous PTEN expression (fig. 2B).

Neonatal rat cardiomyocytes were exposed to 10 nM of PDCD4 siRNA or medium guanine/cytosine content negative control siRNA. qRT-PCR analysis revealed that PDCD4 mRNA expression after siRNA knockdown was reduced to 25.7 ± 3.0% of control values (fig. 3A). Western blot analysis showed that PDCD4 protein expression was reduced to 8.8 ± 1.3% of control values (fig. 3B).

Measurements of cell viability after 6 h of oxidative stress with 50 μM hydrogen peroxide revealed that anti-PDCD4 transfection caused a protective effect in cardiomyocytes that mimicked the protective effect of isoflurane. Oligo-transfected cardiomyocytes were preconditioned with isoflurane (11.8 ± 1.9% PI-positive cells compared with 21.3 ± 3.6% of cells not exposed to isoflurane), while there was no significant difference in the protective effect of isoflurane in PDCD4-transfected cells (7.5 ± 2.3% control cells compared with 8.6 ± 2.1% isoflurane-exposed cells). As a comparison, anti-miR-21–transfected cells were also stained with PI. Scramble-transfected cells were successfully preconditioned with isoflurane (10.6 ± 2.2% compared with 20.0 ± 1.3% non–isoflurane-exposed cells). Transfection with 25 nM of anti-miR-21 abolished the protective effect of isoflurane (21 ± 2.7% compared with 17.5 ± 2.2%) (fig. 3C). Representative PI staining images show decreased cell viability after miR-21 knockdown and increased cell viability after PDCD4 knockdown (fig. 3D).

There were no significant differences in area at risk between WT and miR-21 knockout mice subjected to I/R injury with or without isoflurane exposure (fig. 4A). Coronary occlusion followed by reperfusion resulted in an infarct size of 59.1 ± 4.2% and 62.3 ± 4.6% in WT and miR-21 knockout mice, respectively. Pretreatment of WT mice with isoflurane significantly decreased infarct size (44.1 ± 2.9%). However, disruption of miR-21 abolished isoflurane-induced decreases in infarct size (56.2 ± 3.2%) (fig. 4B).

In this study, we showed for the first time that isoflurane induces up-regulation of miR-21 and that this up-regulation is linked to protection of cardiomyocytes. Lipofectamine-mediated knockdown of miR-21 completely abolished the protective effect of isoflurane and resulted in increased the expression of PDCD4, a well-established target of miR-21.²³  Inhibition of PDCD4 resulted in a protective effect on control cells that mimicked the protective effect on isoflurane-treated cardiomyocytes. However, not all miR-21 targets that we investigated were suppressed. PTEN, a protein suppressed by miR-21 during ischemic postconditioning,²⁴  showed no difference in expression levels after isoflurane exposure or miR-21 knockdown during our study. From this, we conclude that anesthetic preconditioning acts through a miR-21/PDCD4-mediated pathway. This protective effect was correlated in vivo, with miR-21 knockout mice failing to show significant decreases in infarct size after exposure to isoflurane when compared with control mice.

Our array results showed that preconditioning with isoflurane for 30 min altered the profiles of microRNAs in cardiomyocytes (fig. 1). We used a Benjamini–Hochberg-adjusted P value for false discovery rate and found that 16 microRNAs were differentially expressed. When a more conservative Bonferroni correction is applied, one third of the microRNAs remain significant. In particular, isoflurane up-regulated members of the miR-30 family, as well as miR-210 and miR-21. These microRNAs have been previously linked to cardioprotective pathways.^26–28  The miR-30 family (miR-30a, b, c, d, and e) has recently been shown to be down-regulated in response to hydrogen peroxide–induced stress in both neonatal rat ventricular cells and rat aorta vascular smooth muscle cells.²⁸  miR-210 has been shown to be up-regulated in cardiomyocytes under hypoxic conditions and confers cardioprotection.^27,28  Especially, acute up-regulation of miR-21 was involved in ischemic and hypoxic protective signaling mechanisms in multiple cell types including cardiomyocytes.^15,24,29  In the heart, acute up-regulation of miR-21 has been shown to decrease myocardial infarct size, increase cell viability, and indirectly up-regulate the expression of numerous cardioprotective proteins.^25,29  Up-regulation of miR-21 has been linked to the protective effects of ischemic preconditioning in both an in vivo rat model and the cultured cardiomyocytes.^9,23  In other organs, miR-21 has been linked to protection as well. In 2013, Jia et al.¹⁵  showed that up-regulation of miR-21 contributed to the protective effect of xenon preconditioning before I/R injury in the mouse kidney model. These studies, as well as the link to multiple cell signaling pathways involved in anesthetic preconditioning, led us to focus our study on the dissection of the important role of miR-21 in isoflurane-mediated cardiomyocyte protection.

After isoflurane exposure, miR-21 initially increased and then decreased back to baseline levels over a 24-h period (fig. 1B). Knocking down miR-21 attenuated the isoflurane-induced decrease of LDH in cardiomyocytes exposed to hydrogen peroxide (fig. 1E), suggesting that acute up-regulation of miR-21 after isoflurane exposure is sufficient to contribute to protection of cardiomyocytes. Several studies conflict on the long-term effects of miR-21 up-regulation, but an acute up-regulation is well established as protective.^30,31 

Our results revealed that exposure to isoflurane decreased the expression of both PDCD4 mRNA and protein levels compared with nonexposed controls. This decrease was abolished when miR-21 was knocked down. Endogenous levels of PDCD4 protein were raised, and the decrease after isoflurane exposure was not significant compared with nonexposed controls (fig. 2A). In addition, inhibition of PDCD4 expression led to protection in non–isoflurane-exposed cardiomyocytes that was similar to the preconditioning effect of isoflurane (fig. 3). It should be noted that the magnitude of PDCD4 suppression was greater than the down-regulation seen after isoflurane exposure, so it is not surprising that the amount of protection conferred is greater. This suggests that PDCD4 is the downstream target mediated by miR-21 up-regulation. PDCD4 is widely studied as a tumor suppressor³² ; however, it has recently emerged as a target of miR-21 with powerful downstream effects. In renal I/R injury, miR-21 knockdown increased renal I/R injury accompanied by up-regulation of PDCD4.³³  In a cardioprotective response to chronic oxidative stress, neonatal cardiomyocytes were shown to up-regulate the expression of miR-21. Inhibition of miR-21 was linked to an increase in the expression of PDCD4 and decreased cell viability.²³ 

Interestingly, not all reported miR-21 targets previously implicated in preconditioning pathways were influenced by the presence of miR-21 or isoflurane. PTEN has been widely shown to be a direct target of miR-21 that is involved in cardioprotection and was initially hypothesized to be a main contributor to isoflurane-induced cardioprotection. Ischemic postconditioning showed a significant up-regulation miR-21 in mouse hearts, which correlated with a significant up-regulation in the PTEN/Akt signaling pathway.²⁴  Extensive investigation of PTEN was conducted in our model, at multiple time points up to 6 h after exposure to isoflurane. However, all time points failed to show a significant change in PTEN expression after isoflurane exposure or miR-21 knockdown. Specifically, 30 min of isoflurane exposure failed to decrease PTEN mRNA or protein levels compared with control values. miR-21 knockdown also failed to influence both endogenous and isoflurane-exposed levels of PTEN compared with control values (fig. 2B).

To ensure that the results seen in vitro correlated to in vivo conditions, we tested whether the miR-21 knockout mouse could be successfully preconditioned with isoflurane. miR-21 knockout mice are created using a targeting vector, which replaces the 93-bp precursor to miR-21 with a pGK-gb2 loxP/FRT-flanked neomycin resistance cassette.³⁴  These mice are healthy and viable and are capable of cardiac remodeling in response to stress.³⁵  However, when exposed to isoflurane, these mice displayed no decrease in infarct size after I/R injury when compared with non–isoflurane-exposed controls. Both the in vivo and in vitro data provide direct evidence that miR-21 is playing a critical role in the cardioprotection against I/R injury mediated by isoflurane.

One limitation of this study is the use of neonatal rat cardiomyocytes and hydrogen peroxide–induced stress to dissect the important roles of miR-21 in isoflurane-induced protection of cardiomyocytes. This in vitro model may not be able to completely recapitulate the in vivo environments. However, the results found in neonatal rat cardiomyocytes parallel our work in the miR-21 knockout mouse, as well as previous work in our laboratory, studying anesthetic preconditioning in whole animal models, isolated heart preparations, and human-induced pluripotent stem cell–derived cardiomyocytes.^36–38  In addition, neonatal rat cardiomyocytes allow for a broad range of studies and an in-depth view of microRNA signaling pathways in a cardiomyocyte-specific model. When compared with isolated adult rat cardiomyocytes, neonatal myocytes have been shown to maintain a phenotype and contractile profile comparable with in situ hearts during I/R injury.³⁹ 

The use of hydrogen peroxide to mediate reactive oxygen species–induced injury has also been used in previous studies to investigate the role of microRNAs in ischemic preconditioning and in anesthetic preconditioning studies done previously in our laboratory.^16,23,40  Reactive oxygen species–induced stress plays an important role not only in I/R injury but also in myocardial infarction, heart failure, and hypertrophy.^41–44  Understanding the changes in expression levels of genes after reactive oxygen species–induced injury can provide important insight into the pathogenesis of these diseases.

In conclusion, the results obtained in this study reveal for the first time that isoflurane directly affects microRNA expression profiles in cardiomyocytes and that miR-21 is acting to protect cardiomyocytes after isoflurane exposure, most likely through its actions on PDCD4. This shows that while isoflurane-induced cardioprotection shares many pathways with ischemic, hypoxic, and heat shock preconditioning, the mechanisms are not identical. Investigating unique, isoflurane-mediated cardioprotective pathways may lead to mechanisms by which infarct size after I/R injury may be decreased, leading to improved patient outcomes after cardiac events.

The authors gratefully acknowledge the statistical services of Qun Xiang, M.S., and Aniko Szabo, Ph.D., of the Biostatistics Consulting Service at the Medical College of Wisconsin (Milwaukee, Wisconsin), supported, in part, by the National Center for Advancing Translational Sciences, National Institutes of Health, Bethesda, Maryland, through grant no. 8UL1TR000055.

Supported by grants P01GM066730 and R01HL034708 from the National Institutes of Health, Bethesda, Maryland; by FP00003109 from Advancing Healthier Wisconsin Research and Education Initiative Fund, Milwaukee, Wisconsin (to Dr. Bosnjak); and by grant R01GM112696 from the National Institutes of Health (to Dr. Bai).

The authors declare no competing interests.