Background

Rapid aeromedical evacuation (AE) is often used for the transport of critically injured patients to medical facilities following on-site triage. The physiologic and cognitive impacts of such flights are incompletely understood, and there is experimental evidence that the transport itself could contribute to secondary injury to the brain and other organs1,2. Efforts to understand the mechanisms that cause AE-related secondary injury will potentially improve patient safety and outcomes by facilitating more accurate risk assessment related to the timing of evacuation and the development of mitigation strategies to counteract any identified negative effects.

Like most fixed wing aircraft, the cabins of planes used for AE are routinely pressurized up to a cabin altitude of approximately 8000 ft (2438 meters), which is well-tolerated by healthy individuals. Acute hypobaric exposure has a transient effect on mechanisms like oxygen transport in order to compensate for the increased pressure, and extended exposure can lead to long-term physiological adaptation3. But most studies into the pathophysiological effects of hypobaria involve exposure to extreme altitudes (>12,000 ft), such as those experienced by mountain climbers, to study the molecular basis of high-altitude pulmonary edema (HAPE) and high-altitude cerebral edema (HACE)4,5. Conversely, the effects from less extreme AE-relevant hypobaria are more obscure.

The most commonly cited explanation for adverse effects of hypobaria is systemic hypoxia and associated impairment of O2 delivery to vital organs6,7. Thus, critical care practitioners often provide supplemental O2, sometimes up to 100%, to avoid systemic hypoxia. However, supplemental O2 in excess of what is necessary to avoid hypoxia can be toxic due to the production of reactive oxygen species and subsequent tissue damage caused by increased oxidative stress. Some studies have reported worse outcomes in brain injured rats when exposed to hyperoxic hypobaria in comparison to normoxic hypobaria8,9. Moreover, hyperoxia is associated with increased mortality in patients with traumatic brain injury (TBI), stroke, and resuscitation following cardiac arrest10,11. Retrospective clinical studies provide evidence for increased in-hospital mortality for patients treated with hyperoxic ventilation, although results from some subgroups of patients, including those with TBI, are inconclusive12,13.

The effect of hypobaria post-injury is worthy of exploration, however injury causes an inflammatory cascade that may conceal the effects induced by the hypobaric transit itself. For instance, exposure to repeated AE-relevant hypobaria has its own observable effects in the healthy (non-injured) human brain14,15. These changes may have little acute impact on healthy individuals, but they could have a deleterious consequences on already-injured patients as hypobaria itself can cause adverse effects on multiple organs16,17,18,19, particularly after TBI2,8,20. There is growing evidence obtained from lab animal studies that exposure to hypobaria within a few days after trauma can exacerbate physiological and cognitive deficits8,9,17,20,21,22 and increase inflammatory signaling2,17 compared to animals maintained under normobaric conditions2,8,16,17,18,19,20.

The goal of this study was to analyze gene expression in the lungs, heart, blood, and brain (hippocampus) of rats immediately following exposure to AE-relevant conditions, including level of hypobaria (mild: 4000 ft/1219 m equivalent or 654 mmHg; moderate: 8000 ft/2438 m equivalent or 562 mmHg), for durations of 5 and 10 hr, under normoxic (21% O2) or hyperoxic (100% O2) conditions20.

Methods

Exposure of rats to hypobaria and/or hyperoxia

The animal use protocol was reviewed and approved by the U.S. Air Force Surgeon General’s Office of Research Oversight and Compliance (FWR-2018-0001A) and by the University of Maryland, Baltimore Institutional Animal Care and Use Committee (1217007). These studies were conducted in a facility accredited by AAALAC, in accordance with the Guide for the Care and Use of Laboratory Animals (NRC, 2011) and were performed in compliance with DODI 3216.1.

Adult (3–6 months) male Sprague Dawley rats (Inotiv, Inc, West Lafayette, IN) weighing 300–350 g were placed in a flight chamber and exposed to normobaria (767 mmHg; 0 ft altitude equivalent), or two levels of hypobaria, 654 mmHg or 562 mmHg (equivalent to 4000 ft and 8000 ft altitude, respectively)8, for 5 or 10 hr. The chamber pressure was decreased incrementally over the initial 30 min to simulate ascent and increased over the final 30 min to simulate descent. The chamber was also perfused with either air (21% O2; normoxia) or 100% O2 (hyperoxia) throughout the simulated “flight.” Immediately after descent, rats received 4–5% isoflurane until completely anesthetized followed by euthanasia via decapitation. Lung, heart, and hippocampus tissues were collected and snap frozen on dry ice before storage at −80 °C. Blood was collected in PAXgene tubes (Qiagen, Santa Clara, CA) and stored at −80 °C.

RNA isolation and quantification

Samples of about 50–100 mg were collected from heart (lower apex), lung (lower apex of right inferior lobe), and hippocampus and placed into −80 °C RNAlater™ ICE (Thermo Fisher Scientific, Waltham, MA) and stored at −20 °C overnight. The following day, tissue was homogenized on ice and RNA was isolated with TRIzol® (Thermo Fisher) and PureLink™ mRNA mini kit (Thermo Fisher) according to manufacturer’s instructions. Blood mRNA was isolated according to manufacturer’s PAXgene Blood RNA protocol. DNase was added to columns to prevent DNA contamination. Isolated RNA was then quantified using a NanoDrop ND-1000 (Thermo Fisher), and quality and integrity of RNA samples was assessed by an Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA) at the UMB Center for Innovative Biomedical Resources (CIBR). RNA was required to achieve an RNA integrity number of 7 or higher before microarray analysis.

Microarray hybridization

Microarray profiling of lung, heart, and blood RNA was performed using Clariom S Rat Assays (Thermo Fisher), while profiling of hippocampus RNA was performed using Rat Gene 2.0 ST Array chips (Thermo Fisher), according to manufacturer’s instructions. Briefly, 150 ng RNA was transcribed to double-stranded cDNA, followed by synthesis and purification of cRNA. 2nd-cycle ss-cDNA was then synthesized and then purified after RNA hydrolysis using RNase H. ss-cDNA was fragmented and labeled before hybridization onto the microarray chips, which were washed and processed using an Affymetrix GeneChip system 3000 7 G (Thermo Fisher) to generate CEL files.

Analysis of microarray files

Affymetrix image data (CEL files) was analyzed by the UMB Institute for Genome Sciences using the R statistical analysis tool. The R ‘oligo’ package23 was used to access the intensity values which were normalized using the ‘Robust Multichip Average (RMA)’ normalization method provided by the R’limma’ package24. The samples were assessed and corrected for any batch effects before downstream analyses (see USAGE NOTES). A general linear model was utilized to fit the expression values for each gene. Principal component analysis (PCA) was used to identify outliers through visual inspection for downstream exclusion (Fig. 1). The absence of noticeable grouping indicates that there was no obvious driver of sample distribution beyond biological variation. The same analysis was performed for all tissues with final n-values listed in Table 1. Subsequent tables and figures reflect final groups excluding outliers in all tissues.

Fig. 1
figure 1

Principal Component Analysis of Rat Lung Microarray Data. PCA plots showing clustering of lung RNA samples from all treatment groups. Samples are characterized into simulated altitude (0 ft, 4000 ft, or 8000 ft), oxygen (21% or 100%), and duration of exposure (5 hr or 10 hr). (a) The first and second components of all 60 samples are shown. Samples within the red square were identified as outliers and were not considered for differential expression analysis or qPCR validation. (b) The first and second components of samples after removing outliers are shown.

Full size image
Table 1 Final n-values for each tissue type in each treatment group.
Full size table

Statistical analysis

The R ‘limma’ package uses a moderated t-statistic test based on a Bayesian model to assess differential expression. In order to focus on transcripts that showed the greatest change and therefore most likely reflect physiological response to the treatments, initial thresholds were set at >1.5x change (positive or negative) with p-values < 0.1 compared to the baseline group (Altitude: 0 ft; Duration: 5 hr; Oxygen: 21%). P-values were adjusted using the Benjamini-Hochberg method in R. The total number of up- or down-regulated transcripts in each tissue based on these thresholds are listed in Table 2. The identified transcripts were then utilized to assess pathway enrichment using QIAGEN Ingenuity Pathway Analysis (IPA) (QIAGEN Inc., Santa Clara, CA; https://digitalinsights.qiagen.com/IPA)25. The significant pathways enriched were selected using a 5% p-value cut-off. Heatmaps indicating pathway enrichment based on treatment group for lung (Fig. 2), heart (Fig. 3), blood (Fig. 4), and hippocampus (Fig. 5) are available for reference. All figures were generated using the R statistical analysis tool.

Table 2 Differentially expressed transcripts in all rat tissues.
Full size table
Fig. 2
figure 2

Heatmap based on microarray analysis of lung RNA after indicated treatments. Canonical pathway enrichment based on differentially expressed transcripts from lung RNA. Treatment groups were compared to baseline (0 ft, 21% oxygen, 5 hr; not shown). Color gradient represents -log10(p-value).

Full size image
Fig. 3
figure 3

Heatmap of canonical pathway enrichment based on differentially expressed transcripts in heart RNA. Treatment groups were compared to baseline (0 ft, 21% oxygen, 5 hr; not shown). Darker shading indicates greater significance.

Full size image
Fig. 4
figure 4

Heatmap of canonical pathway enrichment based on differentially expressed transcripts in blood RNA. Treatment groups were compared to baseline (0 ft, 21% oxygen, 5 hr; not shown). Darker shading indicates greater significance.

Full size image
Fig. 5
figure 5

Heatmap of canonical pathway enrichment based on differentially expressed transcripts in hippocampus RNA. Treatment groups were compared to baseline (0 ft, 21% oxygen, 5 hr; not shown). Darker shading indicates greater significance.

Full size image

Reverse transcription and microarray validation with qPCR

Reverse transcription of 1 µg lung RNA was carried out with High-Capacity cDNA Reverse Transcription kit (Thermo Fisher) according to manufacturer’s instructions. qPCR was performed using primers designed through the National Center for Biotechnology Information (NCBI) primer design tool (Table 3) and purchased from IDT (Integrated DNA Technologies, Coralville, IA). Primers and annealing temperatures (Ta) were optimized to ensure efficiency was between 90% and 110% (Table 1). qPCR was performed with PowerUp™ SYBR™ Green Master Mix (Thermo Fisher) on a QuantStudio 3 qPCR machine (Thermo Fisher) using the following protocol: initialized with 50 °C for 2 min then 95 °C for 2 min, followed by 40 cycles of 95 °C for 15 s, 15 s at the primers’ respective annealing temperatures, and 72 °C for 1 min. Relative quantification was performed using the QuantStudio 3 software (Thermo Fisher) to compare treatment groups against the control group. β-Actin used as a housekeeping transcript to normalize gene expression.

Table 3 List of primers used for qPCR validation.
Full size table

A diagram of the entire workflow can be found in Fig. 6.

Fig. 6
figure 6

Workflow diagram.

Full size image

Data Records

All individual microarray CEL files associated with this study are available at the NCBI Gene Expression Omnibus (GEO) repository (https://www.ncbi.nlm.nih.gov/geo/) with accession number GSE26313126. All microarray sample files, including those designated as outliers, are included. The metadata record for each sample is also available at that location and a brief description of the dataset can be found in Table 4.

Table 4 Summary of the microarray CEL files available at the GEO database.
Full size table

Technical Validation

After removal of outliers identified by PCA, qPCR validation was performed on specific transcripts using the same RNA samples that were submitted for microarray analysis in order to assess the quality of microarray results. Validation focused on the lung mRNA results. Candidates selected for validation were DE genes that either are known to be responsive to oxidative stress (ex. Hif3a, Osgin1, Sod3) or were identified as DE genes in 2 or more treatment groups. Gene expression was normalized to β-actin (Actb) expression and compared to baseline samples to determine fold-change. In total, 69% of the transcripts tested both agreed with the direction (increase or decrease) of the expression and exceeded the fold-change of the microarray analysis compared to baseline. 28% of transcripts matched with the direction of differential expression but the magnitude of the fold-change did not meet that of the microarray analysis. Only 1 transcript in 1 group completely contradicted the microarray results (Table 5). Considering the increased sensitivity of qPCR as compared to microarray, we concluded that these results are adequate to validate the microarray analysis27.

Table 5 qPCR validation of select transcripts in rat lung.
Full size table

Usage Notes

Microarray analysis is a useful tool for broad-scale gene expression analysis, but it can be an expensive and time-consuming process. This study includes an extensive sample list across multiple tissues and should be useful for research into exposure of healthy or injured animals and humans to these AE-relevant conditions. The canonical pathway analysis provided can be used as an initial reference point for identifying pertinent pathways or mechanisms within each tissue before exploring the raw data itself. Along with the CEL files, TXT files featuring the results of pair-wise comparisons between each treatment group after removal of outliers and the respective tissue’s control group (Altitude: 0 ft; Duration: 5 hr; Oxygen: 21%) are available at GEO26.

Although batch correction was not needed for the vast majority of the data, the hippocampus microarray analyses were performed at different times and therefore were completed using different manufacturing lots of the Rat Gene 2.0 ST Array chips. The hippocampus files should therefore be separated into 2 batches as noted in the sample list (Table 4) and corrected before further analysis.

It should be noted that this study has several limitations. The qPCR validation was performed to assess the quality of the microarray using the same RNA samples, so these data have not been validated with new animals. The experiments only included young male rats, therefore the data may not reflect gene expression in female or older animals. This study was also limited to 2 durations of relatively mild exposure. Although gene expression was altered acutely through exposure to hypobaria and/or hyperoxia, later time points could further elucidate the effects these environments induce in the longer term as well as create a longer timeline in which to identify trends. Exposure to mild atmospheric perturbations for 10 h may not be long enough to observe all downstream effects of some genetic cascades, such as the cellular hypoxia response.

Ethics approval and consent to participate

The study protocol was reviewed and approved by the University of Maryland-Baltimore Institutional Animal Care and Use Committee (1217007) and the U.S. Air Force Surgeon General’s Office of Research Oversight and Compliance (FWR-2018-0001A). Animals were handled, and studies were conducted, under a program of animal care accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International and in accordance with the National Research Council’s 2011 Guide for the Care and Use of Laboratory Animals (in compliance with Department of Defense Instruction 3216.1).