Plasma proteomic study of acute mountain sickness susceptible and resistant individuals

Although extensive studies have focused on the development of acute mountain sickness (AMS), the exact mechanisms of AMS are still obscure. In this study, we used isobaric tags for relative and absolute quantitation (iTRAQ) proteomic analysis to identify novel AMS−associated biomarkers in human plasma. After 9 hours of hypobaric hypoxia the abundance of proteins related to tricarboxylic acid (TCA) cycle, glycolysis, ribosome, and proteasome were significantly reduced in AMS resistant (AMS−) group, but not in AMS susceptible (AMS+) group. This suggested that AMS− individuals could reduce oxygen consumption via repressing TCA cycle and glycolysis, and reduce energy consumption through decreasing protein degradation and synthesis compared to AMS+ individuals after acute hypoxic exposure. The inflammatory response might be decreased resulting from the repressed TCA cycle. We propose that the ability for oxygen consumption reduction may play an important role in the development of AMS. Our present plasma proteomic study in plateau of the Han Chinese volunteers gives new data to address the development of AMS and potential AMS correlative biomarkers.


Discussion
Identifying the proteins involved in physiological and pathological processes associated with hypoxia would shed light on the mechanisms of diseases caused by lack of oxygen, such as AMS. Our present plasma proteomic study in plateau of the Han Chinese volunteers gives new data to address the development of AMS and potential AMS correlative biomarkers. In the present study, we test the plasma proteome of AMS− and AMS+ individuals in BL and short-term (9 hours) high-altitude exposure status. We found that proteins response to acute hypobaric hypoxic exposure in AMS+ individuals and AMS− individuals were different using iTRAQ proteomic analysis. These different proteins may help to address the pathology of AMS. Despite the fact that many theories explaining the development of AMS have been proposed during the past decades, the basic pathogenic mechanism of AMS is still fairly unclear. The blood brain barrier (BBB) theory, one of those hypotheses, suggests that hypoxia-induced hypoxemia will initiate an inflammatory response with the release of inflammatory mediators that contribute to an increase of the capillary pressure by over perfusion and vasodilatation, and elevate the permeability of the BBB by disrupting the BBB. This increases the potential for capillary leak and cerebral edema, which in turn causes the traction of the meninges and blood vessels, and high-altitude headache [9][10][11] . Recently, some studies suggested that reactive oxygen species (ROS) formed during short-term hypoxia exposure and might be involved in cerebral vascular leak and astrocyte swelling in the trigeminal areas which play a crucial role in causing AMS 12 . According to present evidences, inflammation response and antioxidant response may play important roles in developing AMS. The mechanism of AMS is not very clear yet, but increased consumption of oxygen must aggravate the symptoms of AMS at the high altitude.
After exposure to hypoxia, metabolic adaptation is essential for normal tissue functions. In cell models, maintaining oxygen concentration lays in a fast and reversible metabolic switch from oxidative metabolic process to anaerobic glycolytic process [13][14][15] . In human body, the molecular mechanism of the adaptive response to hypoxia is still not well known. TCA cycle is a crucial oxidative metabolic process in mitochondria to produce energy. Our proteomic analyses showed that levels of 10 pivotal metabolic enzymes of TCA cycle, including PDHA1, DLD, ACLY, ACON, IDH1/IDH2/IDH3A/IDH3B, OGDH, DLST, SUCLG1/SUCLG2/SUCLA2, SDHA/SDHB, and MDH2, were markedly decreased in AMS− group after acute exposure to high altitude (Fig. 2). While only SUCLG2 was down-regulated in AMS+ group after acute exposure to hypobaric hypoxia. These results are consistent with the recent proteomic study with muscle, which showed that both TCA cycle and glycolysis were repressed after exposure to hypobaric hypoxia 22 .
Low-level-oxygen is undoubtedly a trigger of AMS, it is reported that increased consumption of oxygen at the high altitude aggravated the symptoms of AMS 16 . Thus, the repressed TCA cycle in AMS− individuals after acute exposure to a high altitude of 3800 m, suggested a decreased consumption of oxygen, may be one of the factors of resistance to AMS. Moreover, as a pivotal metabolic process, TCA cycle provides precursors for certain amino acids and fatty acid synthesis, and intermediates of TCA cycle can be metabolized into amino acids and fatty acid and vice versa (Fig. 2). In the present study the alteration of several amino acids (valine, leucine, isoleucine, lysine, arginine, proline, alanine, aspartate and glutamate) and fatty acid could be explained (at least partly) by the suppressed TCA cycle. These results indicated that AMS− individuals could regulate the efficacy of TCA cycle to reduce the consumption of oxygen when exposure to hypobaric hypoxia, but AMS+ individuals could not. It seems that the minor ability of AMS+ individuals to reduce its oxygen consumption make them more dependent of stable and relatively high level of environmental oxygen, which contribute to their higher AMS susceptibility. But further research will be needed to confirm this point.
In the other hand, some intermediates of the TCA cycle were identified as inflammatory signals. For instance, succinate, a metabolic intermediate of the TCA cycle, playing an important role in adenosine triphosphate (ATP) generation 17 , has been recently proposed for a novel function apart from the TCA cycle. Succinate activates hypoxia-inducible factor-1α (HIF-1α) through inhibiting prolyl hydroxylase domain (PHD) enzyme function and enhancing ROS production 18 . Subsequently, succinate enhances glycolytic process via elevated HIF-1α which is a transcriptional regulator in the switch to glycolysis and increased ROS which inhibits mitochondrial function, in turn boosting glycolysis as a result 17,19 . In our present study, levels of 10 crucial metabolic intermediates of glycolysis, aldose 1-epimerase (GALM), fructose-1,6-bisphosphatase 1 (FBP1), fructose-bisphosphate aldolase (ALDOA), phosphoglycerate kinase (PGK1), phosphoglycerate mutase 1 (PGAM1), beta-enolase (ENO3), pyruvate dehydrogenase alpha 1 (PDHA1), dihydrolipoyl dehydrogenase (DLD), alcohol dehydrogenase (AKR1A1), and aldehyde dehydrogenase (ALDH1B, ALDH6A1, ALDH7A1, ALDH8A1), were significantly decreased in AMS− group after short-time of hypoxia exposure (Fig. 3). The repressed glycolysis may be caused by inhibiting TCA cycle, especially by the decreases of succinate (attenuate SUCLG1, SUCLG2, SUCLA2, SDHA, and SDHB). And then, the reduced ROS production and HIF-1α accumulation may be causes of resistance to AMS in AMS− individuals after rapid ascent to high altitude. However, further in-depth studies are needed to confirm this point.
Cytosolic citrate is required for fatty acid biosynthesis and metabolized to produce reactive oxygen intermediates and prostaglandins, important inflammatory mediators 20 . Our present results showed that ATP-citrate synthase (ACLY) was decreased, suggesting that the decreases of citrate may in turn inhibit the inflammatory process. Furthermore, the fatty acid biosynthesis (fatty acid synthase, FASN) may be decreased in plasma samples of AMS− individuals, which may be partly resulted from the citrate reduction. The decreased level of citrate would lead to NO and ROS reduction, in turn decreasing hypoxia induced oxidative damage.
Protein synthesis is a fundamental cell process of ribosome. Ribosome is the site for biosynthesis which cost lots of energy and oxygen, assigned to the task of decoding the genetic code. There are considerable evidence showing that hypoxia can significantly affect the ribosome assembly and proteins synthesis 26,27 . After acute exposure to high altitude of 3800 m, the ribosome biosynthesis pathway was markedly down-regulated in AMS− group, but not in AMS+ group. Reduced ribosome means attenuated ribosome assembly and protein synthesis, lead to less ATP and oxygen consumption. Consisted with this, in the AMS− group, the proteasome pathway (involving PSMC3, PSMA3, PSMD12, PSMD1, PSMD2, PSMD11, PSMA7, PSMC4, PSMB1, PSMA2, PSMA1, PSMC5, PSMA6, PSMD7) was also greatly suppressed, suggesting that the life of proteins is prolonged and the proteins biosynthesis is less needed. In contrary, after acute exposure to hypoxia, only four ribosomal proteins (Rps2p, Rps9p, Rps24p, Rpl15p) are significantly decreased in AMS+ group. These results demonstrated that reduced ribosome assembly and prolonged protein life potentially decreased oxygen consumption, which might be involved in the resistance to AMS of AMS− individuals. The exact mechanism of ROS functioning in AMS is not very clear yet. Many studies showed that the hypoxia could increase the production of ROS in vivo and suggested that ROS and other oxidative proteins might play important roles in causing AMS. Line of evidence suggested that ROS was formed during acute exposure to hypoxia and contributed to cerebral vascular leak and astrocyte swelling in the trigeminal areas 12 . This has led some researchers to hypothesize that hypoxia increase the production of ROS in the brain, which were subsequently responsible for developing AMS [28][29][30] . Unfortunately, however, exogenous antioxidant (L-ascorbic acid, α-tocopherol acetate, and α-lipoic acid) therapy as a strategy to prevent AMS in human has been proved ineffective 31 . One recent plasma proteomic study of AMS in hypoxic chamber showed that antioxidant properties (i.e. peroxiredoxin-6 (Prdx-6), glutathione peroxidase (GPx), and sulfhydryl oxidase 1(QSQX1)) rose in AMS+ individuals but not in AMS− individuals 5 . Their exploratory analyses suggested that enzymatic antioxidant systems were enhanced after exposure to hypobaric hypoxia in AMS+ individuals but not AMS− individuals. They speculated that enhanced antioxidant systems might be overcompensation for hypoxia-induced oxidant production in AMS+ individuals and questioned that oxidant production played an important role in the development of AMS. While, our proteomic analysis results showed that the abundance of proteins related to TCA cycle, glycolysis, ribosome, and proteasome were significantly reduced in AMS− group, but not in AMS+ group after acute ascent to high altitude. We speculate that oxygen consumption, not oxidant production, may play an important role in the development of AMS. Oxidant production, such as ROS, may decrease resulting from less oxygen consumption.
It was noticed that the vascular endothelial growth factor (VEGF) signal pathway was suppressed in the AMS− group after acute hypoxic exposure, but not in the AMS+ group. Many studies have showed that VEGF seems likely to be a biomarker of AMS. VEGF is a critical angiogenic factor and hypoxia-induced protein 32,33 that can increase vascular permeability 34,35 . Blocking VEGF can prevent hypoxic brain edema 36 . Although in the present study we did not observed changes in VEGF, we found that CEPN3, which is a key regulator of VEGF signal pathway, is significantly decreased after exposure to hypoxia in AMS− individuals, but not in AMS+ individuals. Coupled with previous studies, we assume that VEGF signal pathway may participate in the development of AMS.
One recent study showed that Tibetans have a hyporesponsive HIF signal pathway and blunted physiological responses to hypoxia living at sea level 37 . It is hinted that HIF signal pathway may be involved in the development of AMS. Several studies demonstrated that ALDOA 38 , PGK1 39 and MMP-2 40 are targets of HIF-1 signal pathway. Our present proteomic results showed that ALDOA, PGK1, and MMP-2 were down-regulated in AMS− individuals after acute exposure to hypoxia, but not in AMS+ individuals (Fig. 4). Our present study demonstrated that HIF-1 signal pathway may be repressed in AMS− individuals after rapid ascent to high altitude, but further researches are needed to confirm this point.

Study Limitations.
Plasma from peripheral blood samples may not accurately reflect the cerebral environment. However, peripherally circulating proteins can affect cerebral endothelial permeability 41 , so making possible to evaluate AMS susceptibility through plasma proteomic analysis. Considering ethical limitations and participants safety, peripheral blood sample collection was unique choice for AMS biomarker investigation. Secondly, a technical challenge of global plasma proteomics is that the plasma proteome encompasses a larger number of proteins and a more dynamic protein concentration range than any other single human sample 42 . Even though after several depletion steps, it is not possible to detect proteins of very low abundance that may be of importance for AMS. Thirdly, we analyzed pooled samples by mass spectrometry rather than individual samples. Sample pooling reduces biologic variance between specimens in proteomic and microarray analyses [43][44][45][46] and, consequently, facilitates the identification of the most robust differences between groups 43,47 . A comprehensive evaluation of the effect of sample pooling for proteomic analysis demonstrated that, for the majority of proteins, data obtained from pooled samples accurately represented the mean protein levels of individual samples that composed the pool rather than being skewed by one or two samples 43 . Although reducing biological variation masks unique features of individual samples that compose the sample pool, this strategy is advantageous when the research focus is to elucidate common features of a given population or disease group 45 , as was the case in our study. Further supporting the validity of the proteomic methods we used, a recent paper confirmed that differentially expressed proteins identified by mass spectrometry in high-altitude natives versus sea-level controls were consistent with the results obtained by Western blot and enzyme-linked immunosorbent assay 48 . Finally, although our work has provided some important clues, we do not make a further verification of our proteomic results on whole animal or cell culture models. The primary reason is subjective symptoms of AMS. Coupled with the elusive mechanisms of AMS, especially high altitude headache, we cannot model the AMS in animals. To further test the proteomic results in human we need numerous samples but this implies insurmountable ethical difficulties.

Conclusions
Our present study found that proteins related to TCA cycle, glycolysis, ribosome, and proteasome decreased in AMS− individuals after acute exposure to high altitude, but not in AMS+ individuals. These results suggested that AMS− individuals could reduce oxygen consumption via repressing TCA cycle and glycolysis, and reduce energy consumption through decreasing protein degradation and synthesis compared to AMS+ individuals. The inflammatory response might be decreased resulting from the repressed TCA cycle. We speculate that oxygen consumption may play an important role in the development of AMS. However, further researches are needed to confirm this point. Subjects. We recruited 23 local Chinese volunteers aged 25 to 35 years. They primarily resided at an elevation of 400 m or lower in the area of Xi'an, China. Recruitment, screening and exclusion criteria used in this study were as described previously 5,49 . Briefly, screening procedures to assess general health status and to determine eligibility prior to participation included medical history check, physical and neurological exam, standard blood and urine analysis, and a maximal exercise test. We excluded those who 1) had any health problems; 2) had an abnormal complete blood count, chemistry panel, or liver function results; 3) were pregnant or intended to become pregnant within the near future; 4) history of migraine, headache, seizure, or head injury with loss of consciousness; 5) with mechanical limitations or metal implants that prohibited exercise; 6) were smoker; 7) were regular on prescription medications; inability to reach a workload of least 200 W during the incremental exercise test; 8) or altitude exposure greater than 2500 m within three months of study. All subjects in each group were the Han Chinese and not related.

Ethics statement. This study has been approved by the Institutional
Study design and ascent profile. Peripheral venous blood samples were collected and AMS symptoms were evaluated at low altitude (400 m; Xi'an, China) 24 hours (baseline, BL) in advance. To acutely expose volunteers to high altitude, all subjects were transported to the altitude of 3800 m (Yushu, China) from Xi'an (400 m) in 3 hours by flight. AMS symptoms of subjects were evaluated immediately when landed. This is time point 0 h. The following studies were conducted in Yushu Bayi Hospital, which has a phase I clinical trial qualification and can treat high-altitude illness. Two hours after exposure to the altitude of 3800 m, all subjects completed three 5-minute sub-maximal exercise bouts on a cycle with 15 minutes rest in between in the ward. This is to increase the likelihood of developing AMS 16 . The whole process was monitored by one doctor and two nurses.

Evaluation of AMS status.
To evaluate AMS status, subjects completed self-reported sections of the Lake Louise Questionnaire (LLQ) prior to exposure to the 3800 m high altitude (BL and 0 hour) as well as after 3, 6, 9 and 12 hours of hypobaric hypoxia. Subjects were asked to quantify their degree of headache, gastrointestinal upset, fatigue, and dizziness on a four-point scale (no symptoms = 0, mild = 1, moderate = 2, severe = 3). Subjects with a cumulative LLQ score equal or greater than 3 with headache after ascent to high altitude within 6 to 12 hours were defined as AMS+ individuals. Those with the sum of LLQ score equal or less than 2 or without headache after exposure to hypobaric hypoxia were considered as AMS− individuals 50 .
Plasma collection and sample preparation. Plasma collection and sample preparation was performed as described previously 5 . Blood samples were collected in a semi-recumbent position from an antecubital vein by an indwelling intravenous cannula (BL and 9 hours), and then placed in EDTA-coated blood collection tubes. The plasma was separated from blood cells by centrifugation and stored in 0.2 mL aliquots at −80 °C until analysis.
Protein preparation. The set of 23 subjects was ranked according to the LLQ scores after exposure to 3800 m in 6 to 12 hours. The top 7 individuals with the highest LLQ scores were defined as the subset of AMS−. The bottom 7 individuals with the lowest LLQ scores were regarded as the AMS+ group. Although the criteria reduced sample size, it maximized the differences between AMS− and AMS+ individuals, and avoided the inclusion of subjects with ambiguous AMS status 4,5 . A series of preprocessing steps, which increased the detection sensitivity of expressions of hypoxia-induced proteins changes in AMS+ group and AMS− group, were implemented before analyzing the plasma samples. First of all, plasma samples collected from the 7 AMS− subjects and 7 AMS+ subjects were pooled at each time point respectively. This resulted in four pooled samples for analysis (two at BL for AMS− and AMS+ groups, and other two at 9 hours for the two groups). Secondly, to reduce the complexity of samples and facilitate the identification of lower-abundance proteins, the highly abundant proteins were depleted using ProteoMinerTM Kits (Bio-Rad Laboratories, Hercules, CA, USA) 51 after pooling the samples. Samples were eluted in Lysis buffer (7 M Urea, 2 M Thiourea, 4% CHAPS, 40 mM Tris-HCl, pH = 8.5), reduced with 10 mM DTT at 56 °C for 1 h, and followed by alkylation with 55 mM IAM in a darkroom for 1 h. The reduced and alkylated protein mixtures were precipitated by adding 4 fold volume of pre-chilled acetone at -20 °C overnight. After centrifugation at 30,000 g at 4 °C, the pellet was dissolved in 0.5 M TEAB (Applied Biosystems, Milan, Italy) and sonicated in ice. The solution was afterwards centrifuged at 30,000 g at 4 °C, and an aliquot of the supernatant was taken for determination of protein concentration by Bradford protein assay method 52,53 . The remaining supernatant was kept at −80 °C for further analysis.
Proteomic analysis. One hundred micrograms of total protein were taken out of each sample and the proteins were digested with Trypsin Gold (Promega, Madison, WI, USA) (protein: trypsin = 30:1) at 37 °C for 16 h. The samples were then processed according to the protocol attached to 8-plex iTRAQ reagent (Applied Biosystems).
The peptides were subjected to nanoelectrospray ionization followed by tandem mass spectrometry (MS/ MS) in a Q EXACTIVE (Thermo Fisher Scientific, San Jose, CA) coupled online to the HPLC. Intact peptides were detected in the Orbitrap at a resolution of 70 000. Peptides were selected for MS/MS using high-energy collision dissociation (HCD) operating mode with a normalized collision energy setting of 27.0; ion fragments were detected in the Orbitrap at a resolution of 17500. A data-dependent procedure that alternated between one MS scan followed by 15 MS/MS scans was applied for the 15 most abundant precursor ions above a threshold ion count of 20000 in the MS survey scan with following dynamic exclusion duration of 15 s with electrospray voltage of 1.6 kV. Automatic gain control (AGC) was used to optimize the spectra generated by the orbitrap. The AGC target for full MS was 3 × 10 6 and 1 × 10 5 for MS2. For MS scans, the m/z scan range was 350 to 2000 Da. For MS2 scans, the m/z scan range was 100-1800 Da. Data analysis. Raw data acquired from the Orbitrap were converted into MGF files using Proteome. Proteins identification was performed by using the Mascot search engine (Matrix Science, London, UK; version 2.3.02) against Uniprot-human database (http://www.uniprot.org) containing 143397 sequences. For protein identification, a mass tolerance of 20 ppm was permitted for intact peptide masses and 0.05 Da for fragmented ions, with allowance for one missed cleavages in the trypsin digests. Gln-> pyro-Glu (N-term Q), Oxidation (M) and deamidated (NQ) are the potential variable modifications. Carbamidomethyl (C), iTRAQ8plex (N-term), and iTRAQ8plex (K) are fixed modifications. To reduce the probability of false peptide identification, only peptides with significance scores (≥20) at the 99% confidence interval by a Mascot probability analysis greater than "identity" were counted as identified. Each confident protein identification involved at least one unique peptide. For protein quantitation, it was required that a protein contained at least two unique peptides. The quantitative protein ratios were weighted and normalized by the median ratio in Mascot. We only used ratios with p-values < 0.05, and only fold changes of >1.5 was considered as significant.
Functional analysis. Functional annotations of the proteins were conducted using Blast2GO program against the non-redundant protein database (NR; NCBI) . The Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/kegg/) 54 and the Cluster of Orthologous Groups (COG) database (http://www.ncbi.nlm.nih.gov/COG/) were used to classify and group these identified proteins. Biological processes associated with AMS− and AMS+ individuals were further analyzed by Gene Ontology (GO), which is an international standardization of gene function classification system. KEGG pathway is a collection of manually drawn pathway maps representing our knowledge on the molecular interaction and reaction networks. Statistics. Differences in age, height, weight, BMI, and LLQ scores between AMS− and AMS+ individuals were identified by independent sample t-tests as appropriate. Data are presented as 'mean ± standard deviations' . Criteria for significance were set a priori at two-tailed p < 0.05.