Chronic Psychosocial Stress in Mice Is Associated With Increased Acid Sphingomyelinase Activity in Liver and Serum and With Hepatic C16:0-Ceramide Accumulation

Chronic psychosocial stress adversely affects human morbidity and is a risk factor for inflammatory disorders, liver diseases, obesity, metabolic syndrome, and major depressive disorder (MDD). In recent studies, we found an association of MDD with an increase of acid sphingomyelinase (ASM) activity. Thus, we asked whether chronic psychosocial stress as a detrimental factor contributing to the emergence of MDD would also affect ASM activity and sphingolipid (SL) metabolism. To induce chronic psychosocial stress in male mice we employed the chronic subordinate colony housing (CSC) paradigm and compared them to non-stressed single housed control (SHC) mice. We determined Asm activity in liver and serum, hepatic SL concentrations as well as hepatic mRNA expression of genes involved in SL metabolism. We found that hepatic Asm activity was increased by 28% (P = 0.006) and secretory Asm activity by 47% (P = 0.002) in stressed mice. C16:0-Cer was increased by 40% (P = 0.008). Gene expression analysis further revealed an increased expression of tumor necrosis factor (TNF)-α (P = 0.009) and of several genes involved in SL metabolism (Cers5, P = 0.028; Cers6, P = 0.045; Gba, P = 0.049; Gba2, P = 0.030; Ormdl2, P = 0.034; Smpdl3B; P = 0.013). Our data thus provides first evidence that chronic psychosocial stress, at least in mice, induces alterations in SL metabolism, which in turn might be involved in mediating the adverse health effects of chronic psychosocial stress and peripheral changes occurring in mood disorders.


INTRODUCTION
Stress, defined as the physiological response of the body to any demand (1), serves the principal goal to mobilize energy for appropriate, fight or flight' response. The primary response includes the activation of the sympathetic nervous system and the hypothalamus-pituitary-adrenal axis, resulting in the secretion of catecholamines and glucocorticoids from the adrenal gland. Albeit this adaptation promotes survival of physical threats to homeostasis, chronic psychosocial threats are well-known to adversely affect human health (2)(3)(4).
Chronic psychosocial stress, particularly in conjunction with viral hepatitis, cirrhosis, and hepatocellular carcinoma is thought to contribute to the development and progression of liver disease (5). For instance, a comparison of the mortality in the general population between periods of economic crisis and periods prior or after the crisis revealed an increased all-cause mortality, due among others to an increased incidence of chronic liver disease (3). Moreover, chronic stress is a well-known risk factor for the development of obesity and metabolic syndrome (6,7). Animal studies further support the hypothesis that chronic stress induces (8) and aggravates (9, 10) liver injury, causes hepatic oxidative stress (11,12) and insulin resistance (13), alters hepatic metabolism and gene transcription (14), and disrupts the regulation of lipid synthesis (15). Sphingolipids (SL) comprise a class of lipids with important structural functions and relevance in cell signaling (16). Bioactive SL play a role in the regulation of cell growth, death, senescence, adhesion, migration, inflammation, angiogenesis, and intracellular trafficking (17). Ceramides (Cer) constitute a family of lipid species (18) that are central to SL metabolism as they serve as precursors for the biosynthesis of plasma membrane SL such as glycosphingolipids (GSL) or sphingomyelins (SM), and, alternatively, can be metabolized into other bioactive SL such as ceramide-1-phoshate, sphingosine, and sphingosine-1phosphate. Additionally, Cer modulate a number of biochemical and cellular processes induced by stressor exposure, including apoptosis, cell-cycle arrest and cell senescence. Moreover, several extracellular challenges, such as tumor necrosis factor (TNF)α, chemotherapeutic agents and heat, cause Cer accumulation (19). Increased Cer levels, in turn, are supposed to contribute to the development of several human diseases, including liver diseases (20). Cer can be synthesized de novo from serine and palmitoyl-CoA or by re-acylation of sphingosine in a salvage pathway. Cer can be also generated by the breakdown of SL from biological membranes. Acid sphingomyelinase (here referred to as ASM for human protein and Asm for mouse protein) is one of several mammalian sphingomyelinases that catalyzes the breakdown of SM to Cer and phosphorylcholine (21). Activity of ASM is sensitive to cellular stress and is activated, among others, by TNF-α (22), oxidative stress (23), and ionizing radiation (24). In line with these findings, recent studies provided evidence that both stress-associated disorders, as for instance major depression (25), chronic heart failure (26), acute and chronic alcohol consumption (27,28), and chronic hepatitis C infection (29), as well as various chronic unpredictable stressors (25,30), have been associated with increased ASM activity and Cer levels, respectively.
However, it is unknown to date whether this holds true for chronic stressors which are psychosocial in nature. We therefore analyzed mice subjected to the chronic subordinate colony housing (CSC) paradigm, a pre-clinically validated mouse model for chronic psychosocial stress (31). Importantly, besides typical stress symptoms such as adrenal hypertrophy, thymus atrophy, and increased plasma nor-epinephrine levels, CSC exposure reliably causes anxiety and spontaneous colitis, increases the risk for colon cancer (32), and induces hepatic inflammation and oxidative stress (11). Our analysis of serum samples and liver specimen revealed that chronic psychosocial stress is associated with increased Asm activity, increased levels of C16:0-Cer, a decline in C24:0-Cer, and increased expression levels of Cers5, Cers6, Gba, Gba2, Ormdl2, and Smpdl3b mRNA. Thus, a shift in the SL composition toward an accumulation of C16:0-Cer might be the origin of the adverse health effects associated with chronic psychosocial stress.

Animal Specimens
Liver and serum specimens of male C57BL/6 mice (Charles River, Sulzfeld, Germany) that were either exposed to CSC (n = 8) for 19 days or kept as SHC (n = 8) were analyzed. Animals and procedures are described in detail in Czech et al. (11). Briefly, four CSC mice were housed together with a larger dominant male in a polycarbonate observation cage (38 × 22 × 35 cm) for 19 d consecutively. Prior to CSC exposure, all potential male dominant mice were tested for their aggressive behavior. Males that started to injure their opponents by harmful bites were not used for the CSC procedure. To avoid habituation, each dominant male was replaced by a novel dominant male on days 8 and 15. Serum of independent CSC and SHC groups (each n = 8) was used to confirm results. All experimental protocols were approved by the Committee on Animal Health and Care of the Government of Oberpfalz (Permit Number: 54-2531.2-16/08) and conform to international guidelines on the ethical use of animals.

Preparation of Tissue Lysates and Determination of Asm Activity
Asm activity was determined from liver homogenates and from serum. Tissue, homogenates and serum were stored at −80 • C prior to the analysis. For the preparation of liver homogenates, pieces of 10-20 mg tissue were homogenized in 0.5 ml sucrose lysis buffer (250 mM sucrose, 1 mM EDTA, 0.2% Triton X-100) using a TissueLyser LT bead mill (Qiagen). Raw lysates were centrifugated with ≥10,000 g at 4 • C for 10 min, and supernatants were transferred to new tubes. The protein concentrations were determined using bicinchoninic acid kit (Sigma). For the determination of Asm activity from liver homogenates, 1 µg of protein were incubated with 0.58 µM N-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-dodecanoyl)-sphingosylphosphocholine (BODIPY R FL C 12 -sphingomyelin; D-7711; Life Technologies, Darmstadt, Germany) in 50 µl reaction buffer (50 mM sodium acetate pH 5.0, 0.3 M NaCl, 0.2% NP-40) for 2 h at 37 • C; after incubation, 3 µl of the reaction volume was spotted on a silica gel 60 plate (Macherey-Nagel; Düren, Germany), and Cer and SM were separated by thin layer chromatography using 99% ethyl acetate/1% acetic acid (v/v) as a solvent (33). Intensity of BODIPY-conjugated Cer and SM fractions were determined using a Typhoon Trio scanner (GE Healthcare, München, Germany) and quantified with QuantityOne software (Biorad, München, Germany). For each liver specimen, Asm activity was determined twice from two independent pieces of tissue with similar results. Activity of secretory (S-) Asm was measured with the same protocol using 2 µl of serum and with additional 500 µM ZnCl 2 in the reaction buffer.

Quantification of Ceramide and Sphingomyelin Species by Mass Spectrometry
Cer and SM were extracted and quantified as described previously (25). Briefly, lipid extraction was performed from 100 µl liver homogenates containing 50 µg total protein using C17:0-Cer and deuterated C16-d31 SM (N-palmitoyl-d31-D-erythro-sphingomyelin; Avanti Polar Lipids) as internal standards. Sample analysis was carried out by rapid-resolution liquid chromatography-MS/MS using a Q-TOF 6,530 mass spectrometer (

Extraction of RNA and Synthesis of cDNA
Total RNA was isolated from pieces of liver tissues (<30 mg) using a TissueLyser LT bead mill (Qiagen) and peqGOLD Trifast reagent (Peqlab, Erlangen, Germany) according to manufacturers' instructions. RNA qualities and concentrations were assessed using a Nanodrop ND-1000 UV-Vis spectrophotometer. SuperScript VILO cDNA synthesis kit (Invitrogen) was used to reverse transcribe 1 µg RNA into cDNA using 2 µl 5x VILO reaction mix and 1 µl 10x SuperScript enzyme mix in a final volume of 10 µl. After completion and termination of the RT reaction, cDNA was diluted with 190 µl LowTE and stored at −20 • C.

Quantitative PCR Analysis
Quantitative real-time PCR was performed using a LightCycler 480 real-time PCR system (Roche, Germany) and SYBR-green chemistry. In detail, qPCR reactions contained 5 µl FastStart Essential DNA Green Master, 1 µM of each primer and 2.5 µl diluted cDNA (corresponding to 12.5 ng RNA) in a total volume of 10 µl. Temperature profile used was: 95 • C for 5 min followed by 45 cycles of amplification (95 • C for 10 s, 60 • C for 20 s, 72 • C for 30 s) and by melting curve analysis. After run, PCR product specificity was assessed by the inspection of single peak melting curves, and threshold cycles (Ct) were determined with the second derivative maximum method using the LightCycler 480 software (release 1.5.0). Gene-specific primers were either selected from the literature or from PrimerBank (https://pga. mgh.harvard.edu/primerbank/) or designed via the Universal Probe Library Assay Design Center (http://qpcr.probefinder. com/organism.jsp). Reference genes were tested and selected according to their gene expression stability (34), and a normalization factor was calculated based on the geometric mean of the reference genes Rpl32, Rpl38, Hprt, and Gusb (34) using the SLqPCR package in R version 3.2.2. Relative mRNA expression levels were calculated in Microsoft Excel using the 2 − Ct method (35). The sequence of primers is given in Table 1.

Statistical Analyses
Statistical analyses were performed using the IBM SPSS Statistics version 21. Continuous variables were tested for deviation from normal distributions using the Kolmogorov-Smirnov test. Statistical significance of measurements between SHC and CSC mice was determined using ANOVA. A two-sided P < 0.05 was considered to indicate statistical significance. All results are presented as mean value ± standard deviation (SD). Graphs were created with GraphPad Prism for Windows, Version 4.01 (Graphpad Software, LaJolla, CA) or with MS Excel 2010.

RNA Expression of SL Metabolizing Enzymes Is Upregulated in CSC Mice
To check if other pathways in addition to the SM hydrolyzing pathway via Asm were involved in the change of C16:0-Cer levels upon chronic psychosocial stress, we conducted a gene expression analysis of 31 genes involved in the metabolism of Cer in liver tissue ( Table 1). Three genes, ceramide synthase 1 (Cers1), ceramide synthase 3 (Cers3), and serine palmitoyltransferase, long chain base subunit 3 (Sptlc3) were not expressed in hepatic tissue (data not shown). A comparison of the normalized mRNA  expression levels of the other 28 genes revealed that the majority was higher expressed in CSC compared to SHC mice ( Table 2). Cers5, Cers6, Gba, Gba2, Ormdl2, and Smpdl3b mRNA were significantly higher expressed in CSC mice compared to control mice (Figure 3).

DISCUSSION
In this study, we showed for the first time that chronic psychosocial stress in mice leads to the activation of Asm and, consequently, increased Cer levels, both assessed in hepatic tissue. In detail, increased Asm activity in mice exposed to chronic psychosocial stress was associated with increased concentration of C16:0-Cer and a decline of C24:0-Cer. Increased C16:0-Cer concentrations and especially a shift from very-longchain C24:0-Cer to long chain C16:0-Cer in the SL composition is frequently observed (36) and confers susceptibility to cellular apoptosis (37) and steatohepatitis and insulin resistance (38). Thus, activation of Asm seems likely to constitute an important link between chronic psychosocial stress and its adverse health effects in the liver, but also in other organs. Significance of differentially expressed genes under chronic psychosocial stress was estimated using t-test. Genes with P < 0.05 are given in bold.
Frontiers in Psychiatry | www.frontiersin.org Our analysis further revealed that chronic psychosocial stress affects other key aspects of sphingolipid regulation besides activation of ASM. By means of gene expression analysis we found that several enzymes involved in SL metabolism-Cers5 and Cers6, Gba, Gba2, Ormdl2, and Smpdl3b-were upregulated under conditions of chronic psychosocial stress. Ceramide synthase (CerS) 5 and 6 (MGI symbol Cers5 and Cers6) belong to a group of enzymes that catalyze the formation of ceramides by N-acylation of sphingoid bases. Each CerS has a high specificity toward the acyl chain length and, thus, the active CerS isozymes determine the fatty acid composition of Cer and the derived SL (39). Of the six mammalian CerSs we found four-CerS2, CerS4, CerS5, and CerS6-to be expressed in hepatic mouse tissue, while we could not detect transcripts of CerS1 or CerS3. These results are in line with a previous report (40). Increased CerS5 or CerS6 activity can mediate the accumulation of C16:0-Cer (41,42). Of note, several reports indicate that CerS6 activity is controlled via transcriptional regulation (42,43). Thus, increased expression of CerS5 and CerS6 following chronic psychosocial stress might be responsible for the shift in the cellular SL composition from very-long-chain Cer to long-chain C16-Cer and, thus, contribute to the adverse health effect of stress. However, since CerS are also regulated by posttranslational mechanisms (44,45), enzymatic activities of CerS5 and CerS6 under conditions of chronic psychosocial stress should to be determined before claiming a role in stress-induced health issues.
We also found two glucosylceramidases, Gba and Gba2, to be transcriptionally upregulated under chronic psychosocial stress. This could also contribute to increased Cer levels, resulting from the hydrolysis of glucosylceramide. Chronic psychosocial stress also induced increased expression Ormdl2. Ormdl2 belongs to the evolutionarily conserved family of ORM-like proteins, which are central regulators of SL metabolism (46). Ormdl proteins are negative regulators of de novo ceramide synthesis via inhibition of serine palmitoyltransferase (SPTLC), the first and rate-limiting enzyme in SL production. However, Ormdl proteins can also stimulate de novo synthesis of complex SL downstream of SPTLC in yeast (47) via a mechanism that potentially involves CerS activity (48). Sphingomyelinase-like phosphodiesterase 3b (MGI symbol: Smpdl3b) is a GPI-anchored SM phosphodiesterase with relevance for cellular lipid composition (49). Knockdown of Smpdl3b in RAW264.7 macrophages results in decreased Cer level, indicating that the protein acts as an enzyme and generates Cer.
In summary, our analysis provides first evidence that chronic psychosocial stress has an impact on the hepatic SL metabolism in mice. Whether this relates to the primary stress response, e.g., secretion of glucocorticoids and catecholamines,-not mutually exclusively-or to the strong connection of SL with inflammation and oxidative stress (50) needs to be determined in future studies. Similar changes of the SL metabolism might be involved in the development of stress-associated pathologies in humans, and key enzymes of SL metabolism such as ASM (51) and CerS6 (52) might constitute therapeutic targets to prevent or treat such conditions.

AUTHOR CONTRIBUTIONS
MR and SR conceived and designed the experiments. MR and CR wrote the manuscript. MR, LH, JM, LJ, BK, DL, and AF performed experiments. MR, CR, JK, and EG analyzed the data. CH contributed reagents, materials, and analysis tools. EG and JK provided funding. All authors reviewed the manuscript.