Introduction

Decompression sickness (DCS) is caused by tissue bubble formation from dissolved inert gases during a rapid reduction in the ambient pressure surrounding the body, as may occur with ascent from depth (for example, diving or mining) or ascent to altitude (for example, extravehicular activity from spacecraft or flying in unpressurized aircraft).1 DCS is classified as either type I or type II according to the clinical symptoms.2 Type I DCS generally presents with joint pain, rash, skin marbling or localized edema. Type II DCS includes more serious symptoms such as cardiorespiratory failure, paresis and anesthesia. Thus, DCS involving the spinal cord is defined as type II.

Spinal cord DCS (SC-DCS) is probably the most dangerous form of the commonly encountered syndromes of DCS. It often affects the lower cervical or lower thoracic regions of the spinal cord; however, its predilection for these sites is not understood. Currently, there are two major theories to account for the symptoms associated with SC-DCS. The first theory, the one for which the best experimental evidence exists, is obstruction of venous outflow of the spinal cord in the epidural plexus. This plexus is a series of blood vessels with a large cross-sectional area and, therefore, a relatively slow flow. Direct experimental observations in dogs with SC-DCS confirm that bubbles coalesce in this area and eventually lead to complete cessation of blood flow.3 Once this process results in an ongoing and progressive diminution of blood flow to the cord, the second process of in situ formation of a gas phase within the tissue of the spinal cord either begins or is compounded. Pathological studies in animals suggest that in situ bubble formation takes place, and some of the aspects of the clinical picture of DCS are consistent with the development of in situ bubble formation in the spinal cord.4 However, the molecular pathological mechanism of this in situ bubble formation is far from clear.

Gene chip technology has been widely applied in various fields, including disease diagnosis, disease therapy, as well as drug screening. Data analysis enables us to understand the genetic alterations of diseases and also provides a novel approach for predicting disease incidence and progression. Using microarray chips, we examined the molecular alterations in the spinal cord tissues of a rabbit model of type II SC-DCS, which has not yet been reported. Our findings may provide a new theoretical basis for the incidence of SC-DCS.

Materials and methods

Statement of ethics

We certify that all applicable institutional and governmental regulations concerning the ethical use of animals were followed during the course of this research.

Animals

Male New Zealand white rabbits (n=20, weight 2–3 kg) from a single vendor (Shandong Lukang Inc, Jinan, Shandong, China) were housed in cages at our animal care facility, where they acclimated for 5 days before any procedures. They were fed standard rabbit chow twice daily with free access to water.

Subjects and experimental series

On the basis of a previous report and our preliminary study, animals were placed into a hyperbaric chamber (YC3800/0.3–36 VII, Yantai Moon Hyperbaric Oxygen Chamber Co, Ltd). The chamber was pressurized with room air at an absolute pressure of 0.7 MPa within 5 min, and the pressure was kept constant for 60 min. Then, the chamber was decompressed gradually over a 5-min period at a constant rate. Next, the animals exhibited acroparalysis in one or more than one limb, suggesting the successfully established type II SC-DCS model.5 The control animals were also placed in a chamber, which was decompressed to normal conditions according to the naval air decompression table.6 Within 3 days after treatment, no DCS was found in these control animals.

Within 30 min after removal from the chambers, DCS (n=10) and control (n=10) animals were anesthetized with 0.1 ml kg−1 ketamine and lumianning. Fresh thoracic spinal cord tissues were dissected, and the animals were killed. For molecular biology analysis, some of the tissues were placed into liquid nitrogen followed by preservation in a freezer at −80 °C. Two samples were randomly selected from each experimental group and were used for microarray analysis. Other samples were fixed with 4% formaldehyde, embedded with paraffin, sectioned and used for morphological analysis under a light microscope or for immunohistochemistry.

RNA isolation

Total RNA from the spinal cord of the DCS group or the control group was isolated using the Trizol reagent, according to the manufacturer’s instructions (Invitrogen Life Technologies, Carlsbad, CA, USA). Total RNA from each sample was quantified using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific, Waltham, MA, USA), and the RNA integrity was assessed using standard denaturing agarose gel electrophoresis. The RNA was further purified on an affinity resin column (RNeasy; Qiagen, Chatsworth, CA, USA).

Microarray expression analysis

To determine the relative gene expression profiles in SC-DCS, we analyzed the relative abundance of mRNA using microarray analysis with the Agilent Array platform (Agilent Technologies, Santa Clara, CA, USA). Sample preparation and microarray hybridization were performed based on the manufacturer’s standard protocols. In brief, 1 μg of total RNA from each sample was amplified and transcribed into fluorescent cRNA by using Agilent’s Quick Amp Labeling protocol (version 5.7, Agilent Technologies). The labeled cRNAs were hybridized onto the Whole Rabbit Genome Oligo Microarray (4 × 44K, Agilent Technologies). After washing the slides, the arrays were scanned by an Agilent scanner (G2505B). Agilent Feature Extraction software (version 10.7.3.1) was used to analyze the acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX v11.5.1 software package (Agilent Technologies). The differentially expressed genes between the DCS group spinal cords and the control group spinal cords were filtered based on fold-change cutoffs of ⩽0.5 and ⩾2 for downregulated and upregulated genes, respectively.

Quantitative real-time reverse transcription polymerase chain reaction (qRT–PCR)

qRT–PCR was performed as described previously.7 Reverse transcription was performed using 1 μg of total RNA from each sample and the MMLV Reverse Transcriptase 1st Strand cDNA Synthesis Kit (Illumina, San Diego, CA, USA), according to the manufacturer’s instructions. qRT–PCR was performed using SYBR Green in a Rotor-gene 3000 thermal cycler (Corbett Research, Sydney, New South Wales, Australia). The PCR primer sequences were designed for the selected genes and β-actin, which served as the constitutive control, based on the gene sequences reported in GenBank and were chemically synthesized (Table 1). The specificity of the PCR was confirmed by examining the dissociation reaction plot subsequent to qRT–PCR. PCR of each sample was conducted in triplicate. Data were analyzed using the comparative cycle threshold method.

Table 1 Primer sequences used for qRT-PCR
Full size table

Immunohistochemistry

The spinal cord tissues were fixed in formalin, embedded in paraffin, and cut into 4-μm-thick sections. The sections were pretreated, dewaxed in xylene and hydrated before antigen retrieval. After endogenous peroxidase inhibition, the sections were incubated with a polyclonal antibody against rabbit tumor necrosis factor alpha (TNF-α; 1:300; BG-A042081, Bluegene, Shanghai, China) overnight at 4 °C. After thorough washing with phosphate-buffered saline, the corresponding secondary antibodies were applied and incubated at room temperature for 30 min. Reaction products were visualized by incubation with 3,3′-diaminobenzidine and then counterstained with hematoxylin. Negative controls were prepared by substituting the primary antibody with isotype-matched irrelevant antibody.

Evaluation of immunostaining

TNF-α staining was judged by two pathologists (Mao RQ and Cao YC) who were blinded to any clinical details related to the cases. TNF-α-positive cells were counted from 10 randomly selected high-power fields. The percentage of positive cells among the total number of cells counted was calculated to obtain the TNF-α-positive rate (%).

Statistical analysis

All statistical analyses were performed with SPSS version 11.0 software (SPSS, Chicago, IL, USA). A P-value <0.05 was considered statistically significant.

Results

Pathology

Under a light microscope, vacuolar lesions were observed in the white matter lesions of the spinal cord in DCS animals (Figure 1b). In addition, axons undergoing Wallerian degeneration with loosened and swelling tissue structures, accompanied with glial cell proliferation, were observed in the spinal cord tissues of DCS animals. However, an intact tissue structure of the spinal cord as well as normal morphology of nerve cells were found in control animals (Figure 1a).

Figure 1
figure 1

Compared with the control (a), abundant vacuoles were found in the white matter of spinal cord tissues of DCS animals (b). Axonal swelling (arrow), interstitial hemorrhage and congestion of blood vessels were also detected in the DCS spinal cord. ( × 40 Magnification).

Full size image

Differential gene expression between SC-DCS and control tissues

Using the Whole Rabbit Genome Oligo Microarray, we analyzed the gene expression of spinal cord tissues obtained from DCS and control animals. Gene expression analysis showed that the levels of 609–984 genes were upregulated, whereas 488–948 genes were downregulated in the DCS animals as compared with the controls. Nine genes showed more than a twofold increase (Table 2) and 17 genes exhibited over a twofold decrease (Table 3), simultaneously in DCS animals. These differentially expressed genes were distributed among different functional categories, such as inflammation, ion channels, the cell cycle, material transport and transcription factors.

Table 2 Upregulated genes (more than twofold increase) in SC-DCS tissues compared with the controls
Full size table
Table 3 Downregulated genes (less than twofold increase) in SC-DCS tissues compared with the controls
Full size table

Validation of the microarray data using qRT–PCR

Using the qRT–PCR technique, we validated the microarray chip data. To this end, parathyroid hormone and TNF genes, both of which were upregulated in the microarray analysis, as well as acyl-CoA synthetase and voltage-gated channel genes, both of which were downregulated in the microarray analysis, were included. Compared with the controls, the expression levels of parathyroid hormone and TNF genes were elevated (parathyroid hormone, 6.18-fold increase; TNF, 6.46-fold increase), whereas the levels of acyl-CoA synthetase and voltage-gated channel genes were decreased (acyl-CoA synthetase, 2.28-fold decrease; voltage-gated channel, 3.16-fold decrease) in the DCS animals (all P<0.05). These results were consistent with the microarray hybridization data, confirming the microarray analysis results.

TNF-α protein expression in the spinal cord

TNF-α-positive staining, with the appearance of brown granules, was detected in the cytoplasm of neurons in the gray matter and glia in the white matter of spinal cord tissues. Strong immunostaining of TNF-α and an increased number of cells expressing TNF-α were found in the spinal cord in the gray matter and the white matter in DCS animals (DCS, 16.3±4.5 per high-power field, n=10; Control, 6.5±3.1 per high-power field, n=10; P=0.003; Figure 2).

Figure 2
figure 2

Immunohistochemical analysis of TNF-α expression in the spinal cord. Compared with the control (a), an increased number of cells expressing TNF-α was observed in the white matter of spinal cord tissues of DCS animals (b). ( × 40 Magnification).

Full size image

Discussion

Gene expression profile analysis showed that genes related to inflammation, the cell cycle, methylation, apoptosis and protein translation were upregulated in injured spinal cords of DCS animals. Since it has been reported that TNF is involved in the progression of DCS,8, 9 we were quite interested in the altered expression of TNF in DCS animals. We further confirmed the elevated expression of TNF in spinal cord tissues of DCS animals from the transcription and protein expression levels. TNF-α and TNF-β are two subunits of TNF, and TNF-α accounts for 75–95% of the total biological activities of TNF. In the nervous system, TNF-α is mainly expressed by microglia, astrocytes, blood-borne macrophages, neurons and vascular endothelial cells. Being a cytokine with various biological actions, TNF-α participates in immune and inflammatory responses.8, 9 TNF exerts its biological activities by binding to its receptor on the cell membrane. After binding, TNF triggers signal transduction pathways, including apoptosis and the nuclear factor kappa-light-chain-enhancer of activated B cells and JNK pathways.10, 11

Second, the upregulation of the DNA methyltransferase 3 (DNMT3) gene indicated that DNA methylation might contribute to spinal cord injury in DCS. DNMT is required for the establishment and maintenance of DNA methylation. In mammalian cells, three active methyltransferases, namely DNMT1 and DNMT3A/3B, and one inactive methyltransferase, DNMTL, were identified. DNMT1 contributes to the maintenance of DNA methylation, and DNMT3 is required for de novo methylation. DNA methylation is one of the most important mechanisms of mammalian epigenetics and regulates cell differentiation as well as individual development by affecting the stability of the chromosome structure, the selectivity of gene expression, and DNA replication, recombination and repair.12, 13 Finally, we also observed increased expressions of parathyroid hormone, cyclin A1-interacting protein 1 and meiosis-expressed gene-1-like genes, suggesting the disease complexity of DCS.

Seventeen genes related to inflammation, ion channels, lipid metabolism and signal transduction were found to be downregulated in DCS animals. Of these 17 genes, the 18-kDa cationic antimicrobial protein, macrophage migration inhibitory factor, defensin-4, interleukin-12A, interferon-αA and neutrophil sterilization protein are involved in inflammation, immune response and stress.14, 15, 16, 17 The leucine-rich repeat (LRR) family member LRRC48 was obviously downregulated. The LRR protein is widely distributed in eukaryotic and prokaryotic cells and tissues. Some of the LRR proteins are specifically expressed in the nervous systems. Serving as cell adhesion or ligand-binding proteins, these proteins participate in several physiological activities of nerves, including synapse formation, neurite outgrowth and development, and the transfer and release of neurotransmitters. Until now, the effect of LRRC48 in DCS has been unclear. We speculate that the downregulated LRRC48 in spinal cord tissues of DCS animals may contribute to the abnormal transfer of neurotransmitters.18, 19 Our results further demonstrated that the voltage-gated channel gene was altered in DCS animals, which may be because of altered neuronal functions.20 Furthermore, some of the differentially expressed genes could be identified by the sequence number on Genebank, whereas their functions remain unknown and need to be clarified.

In conclusion, we analyzed the gene expression in spinal cord tissues of DCS animals using gene expression profiling and identified multiple differentially expressed genes involved in the incidence and progression of DCS. Our findings reveal the disease complexity of DCS and confirmed the previous hypothesis regarding the involvement of the TNF gene in DCS. Thus, our study provides an experimental basis for understanding the mechanism of spinal cord injury in DCS.

Data Archiving

There were no data to deposit.