Spaceflight has been shown to affect a number of mammalian body systems. This report provides experimental data on liver proteins of mice that were part of the Commercial Biomedical Test Module-2 flown on the space shuttle Endeavour (STS-118), a 13-day mission to the International Space Station in August, 2007. Several studies on these mice have now been published [1-7]. In earlier studies on humans, loss of Lean Body Mass (LBM), fat and water, as well as loss of strength and changes in plasma protein and amino acid levels were reported. An increase in urinary interleukin-6 (IL-6) on the first day of flight in the Columbia space shuttle indicated an acute-phase response from the liver [8]. Using a ¹⁵N-glycine tracer technique to study protein turnover in four Russian cosmonauts and two U.S. astronauts residing long term aboard the MIR orbital station, investigators found a nearly 50% decrease in protein synthesis [9]. In a review of the literature, Stein and Wade [10] noted that the shift towards increased activity of glycolytic enzymes associated with muscle atrophy, a major concern during spaceflight, is accompanied by increased gluconeogenesis in the liver. Leach et al. [11] showed that there are increased levels of 3-methyl histidine, creatinine and sarcosine due to muscle breakdown, and decreased levels of plasma albumin and transferrin inferring inadequate dietary protein intake on the early Soviet and Skylab missions. The protein depletion seen in astronauts upon landing after missions is followed by a post-flight anabolic phase so that muscles regain normal protein levels, a process that apparently can affect protein status in other body compartments [12]. These above studies, done with hippuric acid (¹⁵N glycine) administration and plasma sampling showed that the fractional synthetic rates of fibrinogen, complement C-3, ceruloplasmin, and haptoglobin were low on day 6 after landing compared to pre-flight measurements. These findings were consistent with limited amino acid availability due to substrate competition between muscles and other tissues.

The hypothesis of this present study was that there will be distinct changes in the profile of major liver proteins of the mouse that are introduced by conditions associated with spaceflight. This profile may occur in humans, as well after a mission in space.

Animals and liver sample collection

C57BL/6NTac female mice (n=10; Taconic Farms, Inc., Germantown, NY) were shipped at about 7 weeks of age to the National Aeronautics and Space Administration (NASA) Space Life Sciences Laboratory (SLSL) at the Kennedy Space Center. For both flight (FLT) and ground control (GRD) mice, similarity of housing conditions in animal enclosure modules (AEM) and acclimatization procedures prior to take off have been described [1]. The FLT mice flew onboard the Space Shuttle Endeavour (STS-118) for 13 days. By using telemetry from the shuttle, the AEM-housed GRD were exposed to environmental conditions comparable to FLT animals (i.e. temperature, humidity, CO₂) on a 48-h delay. The FLT and GRD mice housed in the AEMs were equipped with solid food bars and a water dispenser so that nourishment was available continually even after landing. Within 3-6 h after landing of the space shuttle, FLT and GRD mice were evaluated for muscle strength and scanned with nuclear magnetic resonance imaging to assess lean and fat mass composition (performed by Amgen investigators). Mice were then euthanized with 100% CO₂. The NASA, Amgen, Inc., Loma Linda University and University of Colorado Institutional Animal Care and Use Committees approved this study. A Material Transfer Agreement was also obtained for the transfer of mouse tissues. Liver tissues were alloted to the Loma Linda NASA Laboratory by Amgen.

Protein analysis

FFPE tissue blocks were prepared from the livers harvested from FLT and GRD mice (5 in each group, 10 animals total). The latter served as the basis for comparisons in this study. Sections were cut at 10 μm thickness and de-paraffinized, but not cover-slipped. Serial sections for protein sampling remained unstained and were stored briefly in water. Areas of the FLT and GRD liver sections selected for uniformity were removed from the slides with a 3 mm punch. These punched tissue samples, each containing approximately 30,000 cells, were placed in 20 μL of Expression Pathology digestion buffer (www.expressionpathology.com). (Expression Pathology Inc. Rockville, MD). The Expression Pathology protocol for digest preparation and analysis was followed precisely for all the samples. This consisted of first heating the punched tissue samples in the proprietary digestion buffer at 95°C for 1 h. This was followed by digestion with sequencing grade trypsin (www.promega.com) (Promega Corp. Madison WI), overnight at 37°C. Small aliquots were analyzed for protein content using a Micro BCA Protein Assay Kit (www.piercenet.com) (Thermo Fisher Scientific, Rockford, IL), prior to subjecting the remaining digest to dithiothreitol reduction. Peptide samples 1.5 μg by protein assay from each digest were then evaluated by LC/MS/MS on a Thermo-Electron LCQ Deca XP mass spectrometer (www.thermo.com) (Thermo Fisher Scientific), using nano-electrospray equipment produced by New Objective, Woburn, MA (www.newobjective.com). This consisted of reverse-phase collection of peptides on a 2 cm x 75 μm capture column followed by separation of peptides on a 10 cm x 75 μm vented analytical column [13], using Micro Magic RP-18AQ resin (www.michrom.com) (Michrom Bioresources, Inc., Auburn, CA). MS analyses were accomplished with a 5-part protocol cycle that consisted of one full MS survey scan (from 400 to 1700 m/z), followed by acquisition of collisional-induced dissociation tandem mass spectra of the four most intense ions in the survey scan. The nanoflow solvent gradient (linear 2-60% acetonitrile) extended over 3 h, using solvent B (95% acetonitrile with 0.1% formic acid) developed against solvent A (2% acetonitrile with 0.1% formic acid), at a flow rate of 250-300 nL/min. Samples were run as groups of five followed by water blank, preceeded and followed by a Michrom BSA trypsin-digested control followed by a water blank. Minimal carryover was detected in the water blanks.

Data analysis

Runs were first evaluated by the Thermo Bioworks software generating .dta files, then Perl .mgf files (www.perl.com) for evaluation by the Mascot search engine (www.matrixscience.com), using an International Protein Index (IPI) mouse FASTA library (www.ebi.ac.uk/IPI), modified to a concatenated format so as to detect false discoveries. These results were then moved into Scaffold (version 3.3.3) (www.proteomesoftware.com). Merged resultant files from Mascot and X!-Tandem (www.thegpm.com) within Scaffold gave a final readout of the analysis, using the spectral quantitative value display option with filter settings of: Min Protein 95%, Min # Peptides 2, min Peptide 95%. This gave 67 proteins identified for this dataset. The aim to obtain highly reliable but not necessarily comprehensive quantitative data was attained in positive and negative format by expression as log₂ fold changes (FLT/GRD). Thus IPA, which utilizes fold change data as well as P-values for reliability, gave good pathway analyses with this format. By combining FLT and GRD data within Scaffold and entering each analysis as a separate biosample, a T-test analysis was generated between the FLT and GRD groups, and also a log₂ fold change for each from FLT AVG / GRD AVG calculations was created. The Quantitative Value display option with Total Ion Current (TIC) was chosen as the quantitative method.

Statistical and pathway analyses

For this type of data, the T-test method of analysis has been shown to be the best [14]. These two groups (FLT and GRD) of data were then analyzed through the use of Ingenuity Pathways Analysis (IPA) (Ingenuity Systems, Redwood City, CA) (www.ingenuity.com) to generate significant pathway networks, and for comparison with canonical pathway networks within the IPA databases. Data was thus entered into IPA using the IPI accession numbers, the P-values and log₂ fold change values, the latter allowed demonstration of positive (up-regulated) and negative (down-regulated) values for the flight group within IPA (Table 1). The entire contents of the Scaffold analysis dataset, Table 1 (67 proteins) was fed into IPA because even proteins of lower significance (P-value>0.050) have an advantageous influence on the outcome of the pathway analysis (Figure 1).

Table 1: 67-protein dataset generated in Scaffold with windows set at high stringencies (95% for both minimum protein and peptide identification probabilities, with 2 unique peptides). Proteins are sorted with respect to P-Values. 8 proteins have values <0.05. Averages of TIC counts for each of the 5 FLT analyses and 5 GRD analyses are shown as well as standard deviations (SD FLT & SD GRD), fold changes (FC) and LOG₂ FC. Fold changes were calculated from FLT/GRD ratios. For computational purposes, all zero values within Scaffold were reset to 1000000.

Figure 1: Through the use of the Ingenuity Pathway Analysis software and interactive website, we linked 6 of the 8 liver proteins, and their pathways, which we found to be significantly altered by spaceflight (Table 1). These are all enzymes (CSP1, ALDOB, HSPD1, GNMT, MAT1A and RGN); these proteins in red or green are up or down regulated respectively, at least 1.5 fold over GRD values (except for CPS1 which is 1.2). All of the end-product proteins except PC (pyruvate carboxylase) are represented in Table 1.

In our present MS study comparing FLT versus GRD mice, measurement of statistically significant (P<0.05) changes of eight proteins was achieved in a specific mouse organ, the liver, via an innovative methodology that allows analysis of FFPE tissue for total tissue protein composition. The Expression Pathology reagents and procedure provided a simple method for obtaining these data from FFPE tissues (without this statistical data, no one could have accomplished this study). Scaffold yielded a spectral quantitative value and total ion current for each peptide identified within each analysis of the MS/MS data. Of the 8 proteins, only carbamoyl-phosphate synthetase and 60 kDa heat shock protein, a chaperonin, gave significant positive log₂ fold change values. Carbamoyl-phosphate synthetase was the protein present in highest concentration. This enzyme provides the entry point of ammonia into the urea cycle, and is found primarily in the liver [15]. The urea cycle is the major mechanism for ridding the organism of catabolic ammonia. The other enzymes of the urea cycle (ornithine carbamoyltransferase, argininosuccinate synthetase, argininosuccinate lyase and arginase-1) were also apparent in the MS spectra, but their levels were not significantly altered between FLT mice and GRD controls. These are also present in Table 1. Their inter-relationship is shown in Figure 2. Because it is up-regulated, carbamoyl-phosphate synthetase appears to be the major urea cycle regulatory enzyme in these mice.

Figure 2: Enzymes of the urea cycle, all of which are identified in our dataset (Table 1). CPS1 = carbamoyl-phosphate synthetase, OTC = ornithine carbamoyltransferase, ASS1 = argininosuccinate synthetase, ASL = argininosuccinate lyase and ARG1 = arginase-1. Captions and colors are the same as in Figure 1. From Boyer (15).

Ammonia in excess of that transformed into urea is toxic to cells via the glutamate dehydrogenase (GDH) catalyzed reaction, even though nontoxic glutamate is produced. The toxicity is due to the concomitant depletion of alpha keto glutarate, by the GDH-catalyzed reaction and consequently, the other Krebs cycle intermediates [15]. Thus, a metabolic result of detoxifying excess ammonia by GDH activity lowers the amount of oxaloacetate, the substrate required for entry of acetyl coenzyme A into the Krebs cycle; with the result that energy metabolism is inhibited. Up-regulation of GDH is indicated in FLT liver samples, although not statistically significant (Table 1). Even though an elevation of GDH may seem advantageous for ammonia detoxification, the downside would be decreased in the capacity for energy production from glucose metabolism. The data for carbamoyl phosphate synthetase and GDH strongly suggest a compensatory mechanism in the FLT mice to deal with an ammonia overload caused by excessive protein catabolism. Skeletal muscle is the largest active protein pool in the body, and is the major site of protein catabolism. It is reported that blood urea nitrogen levels double during the first month of spaceflight. In astronauts, a 6-day mission resulted in a 3.5 kg loss of LBM [8]. Glycine N-methyltransferase, a methyl group transferring enzyme which participates in detoxification chemistry in liver cells [16] is down-regulated (Table 1). Studies have shown an insulin c-peptide elevation in astronauts of the American Space Shuttle missions, as high as 40% above normal after 9 days of flight [8]. This might infer a hyper insulinemia in an attempt to increase cellular uptake of glucose or perhaps indicate an insulin resistance, both of which relate to an intracellular glucose deficit contributing to increased protein catabolism. In contrast with the data which indicate protein catabolism, Ferrando et al. [17] state that the primary adaptation to space flight is not an increase in protein breakdown, but rather the body’s inability to maintain protein synthesis in skeletal muscle. They feel that the most apparent factor in the loss of LBM is reduced energy intake during spaceflight (25-30% on some spaceflights). The data suggest in agreement with Stein & Gaprindashvili [8], that insulin resistance may also be a factor. The down-regulated fructose-bisphosphate aldolase B levels were observed and are also consistent with this observation (Table 1). Insulin positively regulates fatty acid synthesis by increasing fatty acid synthase (FAS) mRNA transcription [18], but FAS as it appears in (Table 1) is markedly down-regulated. However, the P-value for FAS is below the significance cutoff of 0.050 in our study because the quality of the spectra as they appear in Scaffold is sub-optimal. Thus, our FAS data should be considered unreliable.

The smallness of this protein data set (67 proteins) may stem from the MS instrumentation that was available, but even though not complete, it represents the potential of utilizing FFPE tissue for conducting studies of this type and represents a beginning of understanding, of what takes place in the mammalian liver with weightlessness followed by stress of landing. Eight of the 67 proteins show statistical differences between FLT and GRD. Six were mapped by IPA. Some of these are linked to detoxification pathways within the liver (carbamoyl-phosphate synthetase, glycine N-methyltransferase, S-adenosylmethionine synthetase), and some to carbohydrate metabolism (fructose-bisphosphate aldolase B, alpha-enolase). 60 kDa heat shock protein was up-regulated, probably because of its relation to stress. Regucalcin was highly down-regulated possibly limiting osteoporosis, which is a major problem with space flight [19]. Overexpression of regucalcin has been shown to induce bone loss in rats [20]. Ribonuclease UK114, also known as heat-responsive protein 12 [21], was down-regulated possibly due to the stress of space flight as well.

The authors are grateful for Amgen, Inc., for sponsoring the flight investigation and generously providing the tissues required to conduct this study. In particular, we thank HQ Han and David Lacey, the principal investigators at Amgen. We also thank Ramona Bober and the rest of the staff at NASA Space Life Sciences Laboratory (SLSL) at the Kennedy Space Center for their support, and the students and technicians from the University of Colorado for assistance with tissue collection. Dr. Gregory A. Nelson and Tamaki Jones at Loma Linda University also assisted in various aspects of this study.