SEARCH

SEARCH BY CITATION

Keywords:

  • arachidonic acid;
  • bipolar disorder;
  • brain;
  • cyclo-oxygenase;
  • prostaglandin;
  • valproic acid

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

Sodium valproate, a mood stabilizer, when chronically administered to rats (200 mg/kg i.p. daily for 30 days) significantly reduced the brain protein levels of cyclooxygenase (COX)-1 and COX-2, without altering the mRNA levels of these enzymes. COX activity was decreased, as were the brain concentrations of 11-dehydrothromboxane B2 and prostaglandin E2 (PGE2), metabolites of arachidonic acid (AA) produced via COX. In contrast, the brain protein level of 5-lipoxygenase and the concentration of its AA metabolite leukotriene B4 were unchanged. In view of published evidence that lithium chloride administered chronically to rats, like chronic valproate, reduces AA turnover within brain phospholipids, and that lithium post-transcriptionally down-regulates COX-2 but not COX-1 protein level and enzyme activity, these observations suggest that mood stabilizers generally modulate the release and recycling of AA within brain phospholipids, and the conversion of AA via COX-2 to PGE2 and related eicosanoids. If targeting this part of the ‘AA cascade’ accounts for their therapeutic action, non-steroidal anti-inflammatory drugs or selective COX-2 inhibitors might prove effective in bipolar disorder.

Abbreviations used
AA

arachidonic acid

COX

cyclooxygenase

cPLA2

cytosolic phospholipase A2

G3PDH

glyceraldehyde 3-phosphate dehydrogenase

iPLA2

Ca2+-independent phospholipase A2

LO

lipoxygenase

LTB4

leukotriene B4

PGE2

prostaglandin E2

PLA2

phospholipase A2

PPAR

peroxisome proliferator-activated receptor

PUFA

polyunsaturated fatty acid

sPLA2

secretory phospholipase A2

TxB2

thromboxane B2

Bipolar disorder is a psychiatric illness with a prevalence of 1.5%, high risk of suicide, and great social and economic costs (Zarate and Toehn 1996). Valproic acid has proven effective in the treatment of acute mania and bipolar disorder in controlled clinical trials (Guay 1995; Post et al. 1996), but its mechanism of action is not fully understood. One possibility is that it targets part of the brain arachidonic acid (AA; 20 : 4 n-6) cascade (Rapoport and Bosetti 2002). This cascade involves the phospholipase A2 (PLA2)-initiated hydrolysis of esterified AA from the stereospecifically numbered (sn)-2 position of phospholipids, followed by conversion of AA to bioactive eicosanoids via cyclo-oxygenase (COX)-1 or COX-2, lipoxygenase (LO) or cytochrome P450 epoxygenase enzymes. A fraction of the released AA is re-incorporated into brain phospholipids via specific acyl-CoA transferases and acyl-CoA synthetases (Shimizu and Wolfe 1990; Rapoport et al. 1997; Fitzpatrick and Soberman 2001). AA and its metabolites are critical to signal transduction, transcriptional regulation, neuronal activity, apoptosis and a number of other physiological processes in the central nervous system (Leslie and Watkins 1985; O'Banion 1999; Kam and See 2000).

We previously showed that chronic administration of lithium chloride to rats, producing a therapeutically relevant mean brain lithium concentration of about 0.8 mm (Bosetti et al. 2002b), reduced AA turnover within brain phospholipids in awake rats (Chang et al. 1996), as well as AA conversion to prostaglandin E2 (PGE2) via COX-2 (Bosetti et al. 2002a). In contrast, COX-1 protein level was unaffected by lithium chloride (Bosetti et al. 2002a).

Valproate, like lithium, reduces AA turnover within brain phospholipids (Chang et al. 2001) and decreases concentrations of bioactive COX and LO products in rat platelets (Szupera et al. 2000). We therefore thought it of interest to see if chronic valproate, like chronic lithium, would have effects on brain enzymes regulating AA turnover and metabolism, cytosolic PLA2 (cPLA2), COX and 5-LO enzymes, and on the concentrations of their AA-derived products PGE2, thromboxane B2 (TxB2) and leukotriene B4 (LTB4). We hypothesized that if these structurally different drugs, at relevant concentrations, were targeting common parts in the AA cascade, these effects might be related to their therapeutic action in bipolar disorder.

Animals and valproate administration

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

The study conformed to the Guidelines for the Care and Use of Laboratory Animals (NIH Publication no. 80–23). Adult male Fischer-344 rats, weighing 200–250 g (Charles River Laboratory, Wilmington, MA, USA) were acclimatized for 1 week in our animal facility, in which temperature, humidity and light cycle are controlled, and had free access to food and water. Twenty rats received sodium valproate (100 mg/mL; Sigma, St Louis, MO, USA) in 0.9% saline by intraperitoneal injection once a day at 200 mg/kg for 30 days (Chang et al. 2001), to produce a therapeutically relevant plasma concentration (Iyer et al. 1995; Loscher and Honack 1995; Bowden et al. 1996). Twenty rats (controls) were injected with an equivalent volume of 0.9% saline. After 30 days of daily injection, the rats were killed by carbon dioxide inhalation and immediately decapitated. Whole brains, including cerebellum and brainstem, were rapidly excised, frozen in 2-methylbutane at −50°C, and stored at −80°C until use. Rats used to measure eicosanoid concentrations (four for each group) were killed with an overdose of sodium pentobarbital (100 mg/kg i.p.), then subjected to high-energy head-focused microwave irradiation (4.8 kW, 3.5 s; Cober Electronics, Stanford, CT, USA) to stop metabolism.

Western blot analysis

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

The brains were homogenized using a Teflon–glass homogenizer in 4 mL ice-cold homogenization buffer (pH 7.5) containing 25 mm Tris, 150 mm NaCl, 5 mm EGTA, 2 mm DTT, 10 µm phenylmethylsulfonyl fluoride, and the phosphatase inhibitors aprotinin (9.2 µm), leupeptin (0.11 mm) and pepstatin A (10 µm). The cytosolic fraction was obtained from the homogenates by centrifugation at 18 000 g for 10 min at 4°C. Protein concentrations of the homogenate and of the cytosolic fraction were determined by the Bradford method (Bradford 1976), using bovine serum albumin as a standard. Either cytosolic protein or whole brain homogenate (50 µg) was loaded on NuPage gels (Invitrogen, Carlsbad, CA, USA) and transferred to a nitrocellulose membrane as directed by the manufacturer. Western blotting was carried out as reported (Bosetti et al. 2002a), using primary antibodies for COX-1 (monoclonal, 1 : 500; Cayman Chemicals, Ann Arbor, MI, USA), COX-2 (polyclonal, 1 : 500; Cayman Chemicals), 5-LO (polyclonal, 1 : 200; Santa Cruz Biotechnology, Santa Cruz, CA, USA), cytochrome P450 epoxygenase CYP2C11 (1 : 1000; BD Gentest, Woburn, MA, USA) or actin (1 : 15 000; Sigma), followed by a secondary antibody conjugated with horseradish peroxidase (1 : 1000; Bio-Rad, Hercules, CA, USA). Immunoblots were visualized by means of a chemiluminescence reaction (Pierce, Rockford, IL, USA). The protein level was quantified by measuring the integrated optical density of the bands, after subtraction of background by AlphaEase Stand Alone software (Alpha Innotech, San Leandro, CA, USA). Image analysis of the blots was performed on optical density-calibrated images captured with a video camera (Alpha Innotech). Only bands below the saturation limit were analyzed.

RNA isolation and RT–PCR

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

Total RNA was isolated with a RNeasy Maxy kit (Qiagen, Valencia, CA, USA). Total RNA (1 µg) was reverse transcribed using a RETROscriptTM kit (Ambion, Austin, TX, USA), by 100 units murine Moloney leukemia virus in RT buffer containing 50 mm Tris-HCl, pH 8.3, 75 mm KCl, 3 mm MgCl2, 5 mm dithiothreitol, 5 µm oligo (dT) primers and 10 units Rnase inhibitor in a volume of 20 µL. The RT mixture was incubated at 44°C for 1 h. A 1-µg sample of each RNA was incubated similarly in the absence of RT to ensure that PCR products resulted from amplification from the specific mRNA rather than from genomic DNA contamination. The RT reaction was stopped by incubating the mixture at 92°C for 10 min. The cDNA was amplified in an automated thermal cycler (GeneAmp 9700; Perkin Elmer, Norwalk, CT, USA) with 2.5 units Platinum Taq Polymerase (Invitrogen/Life Technologies, Carlsbad, CA, USA), 1.5 mm MgCl2, 150 µm dNTPs and 0.5 µm specific oligonucleotide primers for COX-1 (forward: 5′-CTCACAGTGCGGTCCAAC-3′; reverse: 5′-CCAGCACCTGGTACTTAAG-3′; 424 bp) (Bernard et al. 2000) and for COX-2 (forward: 5′-ACTTGCTCACTTTGTTGAGT3′; reverse: 5′-TTGATTAGTACTGTAGGGTT-3′; 581 bp) (Feng et al. 1995) in a total volume of 50 µL. Specific glyceraldehyde 3-phosphate dehydrogenase (G3PDH) primers (983 bp; Clontech, Palo Alto, CA, USA) were used as an internal control to normalize the sample amounts. cDNA was heated at 94°C for 4 min, then the PCR mastermix with 10% dimethylsulfoxide was added. The reactions were heated at 75°C for 4 min (‘hot start’ method). After an initial 5-min denaturation at 95°C, the DNA was amplified in 30 cycles of denaturation at 94°C for 20 s, primer annealing at 55°C for 30 s, and extension at 72°C for 40 s, with a final extension at 72°C for 5 min for COX-2; 30 cycles of denaturation at 94°C for 1 min, primer annealing at 56°C for 1 min, and extension at 72°C for 1.5 min, with a final extension at 72°C for 4 min were used for COX-1. Agarose gels (1.3%) were stained with ethidium bromide and the bands were quantified by AlphaEase Stand Alone software. Integrated densities were normalized to G3PDH values to yield a semiquantitative assessment of individual transcript levels. Preliminary experiments confirmed that the PCR conditions and the image analysis system were in the linear range of detection.

Measurement of PLA2 activity

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

One half of each brain was homogenized in 3 mL lysate buffer (10 mm Tris-HCl, pH 7.8, containing 1% Nonidet P-40, 0.15 m NaCl and 1 mm EDTA), then chilled on ice for 30 min and centrifuged at 4000 g for 25 min at 4°C. Arachidonoyl thio-PC (1-O-hexadecyl-2-deoxy-2-thio-R-(arachidonoyl)-sn-glyceral-3-phosphorylcholine) was used as a synthetic substrate to detect PLA2 activity. Hydrolysis of the arachidonoyl thioester bond at the sn-2 position by PLA2 releases free thiol, which is detected by Ellman's reagent (Reynolds et al. 1994). PLA2 activity was determined in the supernatant using a cPLA2 assay kit (Cayman Chemicals), in the presence and absence of the specific inhibitor of Ca2+-independent PLA2 (iPLA2), bromoenol lactone (Ackermann et al. 1995), which was incubated for 15 min at 25°C at a concentration of 10 µm before the assay. Activity was calculated by measuring the absorbance at 414 nm, using the 5,5′-dithiobis(2-dinitrobenzoic acid) (DTNB) extinction coefficient of 10.66 per mm, and reported as nmoles per minute per gram cytosolic protein.

Measurement of COX activity

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

COX activity was determined as reported previously (Bosetti et al. 2002a) in brain homogenate cytosolic fraction diluted 1 : 10 with lysate buffer [10 mm Tris-HCl, pH 7.8, containing 1% Igepal CA-630 (Sigma), 0.15 m NaCl and 1 mm EDTA], in the presence of 10 mm phenol, 18.2 mm l-epinephrine, 4.6 mm glutathione and 9.3 µm hematin. The reaction was started by adding 0.1 mm AA, and the mixture was incubated at 37°C for 10 min. The reaction was terminated by adding 250 µL 1 m HCl. PGE2 was extracted by ethyl acetate and determined using a PGE2 immunoassay kit (Cayman Chemicals), as directed by the manufacturer. A sample not allowed to react with AA was prepared and assayed in the same manner, and used as a blank determination.

Eicosanoid concentrations

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

Eicosanoid concentrations were determined in microwaved brain extracts. Brains were weighed, then extracted in 18 volumes of hexane : 2-propanol (3 : 2 by volume) using a glass Tenbroeck homogenizer. The prostaglandins were purified from the lipid extract by the method of Powell (Powell 1985). The concentration of PGE2 was determined by ELISA (Oxford Biomedical, Oxford, MI, USA). 11-DehydroTXB2 and LTB4 were determined using a specific immunoassay kit (Cayman Chemicals).

Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

Brain PLA2 activity, measured in the presence and absence of the iPLA2-specific inhibitor, bromoenol lactone, was not changed by valproate treatment (Table 1). On the other hand (Figs 1a and b), valproate significantly reduced the mean ratio of COX-2 protein to actin protein (by 33%) compared with control (0.68 ± 0.08 vs. 1.02 ± 0.11; p < 0.05, n = 6). The ratio of COX-1 protein to actin protein was also significantly decreased (by 38%) compared with control (0.52 ± 0.06 vs. 0.84 ± 0.07; p < 0.01, n = 6). Similar results were obtained using the whole brain homogenate, which confirmed that the observed decline in COX proteins in the cytosolic fraction did not result from their translocation from the cytosol to a membrane fraction (Figs 1c and d). The ratios of COX-1/actin were 0.65 ± 0.04 for valproate-treated and 1.0 ± 0.05 for control homogenate samples (n = 6, p < 0.01). Those for COX-2/actin were 0.81 ± 0.05 for valproate-treated and 1.08 ± 0.04 for control homogenate samples (n = 6, p < 0.05).

Table 1.  PLA2 activity in the cytosolic fraction of brain homogenate from valproate-treated and control rats
 Total PLA2 activity (nmol per min per g protein) cPLA2 + sPLA2 activity (nmol per min per g protein)
  1. Data are mean ±SEM of 10 independent samples for each group. Each sample was assayed in duplicate. Arachidonoyl thio-PC was used as a synthetic substrate to detect PLA2 activity. Hydrolysis of the arachidonoyl thioester bond at the sn-2 position by PLA2 releases free thiol, which is detected by Ellman's reagent (DTNB). The sum of cPLA2 + sPLA2 (calcium-dependent PLA2 enzymes) activities was assayed in the presence of the specific inhibitor of iPLA2, bromoenol lactone (10 µm).

Control (n = 10)677.3 ± 32.5627.8 ± 25.4
Valproate (n = 10)704.9 ± 53.9650.1 ± 47.1
image

Figure 1. Brain COX-1 and COX-2 protein levels in valproate-treated (VP) and control (Con) rats. (a) Representative immunoblots of COX-1, COX-2 and actin in the cytosolic fraction. (b) Optical density (OD) of COX-2 and COX-1 relative to actin in the cytosolic fraction. (c) Representative immunoblots of COX-1, COX-2 and actin in whole brain homogenate. (d) OD of COX-2 and COX-1 relative to actin in whole brain homogenate. Values are mean ± SEM (n = 6). Each sample was derived from an individual rat brain. *p < 0.05, **p < 0.01 versus control (Student's t-test).

Download figure to PowerPoint

These decreases were not due to down-regulation of COX-1 or COX-2 gene transcription, as RT–PCR showed that the mRNA levels of COX-1 and COX-2, normalized to the G3PDH mRNA level, did not differ significantly between valproate-treated and control rats (n = 8) (Figs 2a and b). However, the use of a more sensitive technique such as real-time PCR or examining specific brain regions might increase the chance of finding decreases in mRNA levels that may not be detected with semiquantitative RT–PCR in whole brain.

image

Figure 2. Brain COX-1 and COX-2 mRNA levels in valproate-treated (VP) and control (Con) rats. (a) Representative ethidium bromide gel illustrating COX-1, COX-2 and G3PDH mRNA levels in rat brain, assessed by RT–PCR, after i.p. valproate injection for 30 days compared with that in controls. (b) COX/G3PDH ratio in brain of controls and valproate-treated rats (n = 8). Each sample was derived from an individual rat brain.

Download figure to PowerPoint

COX enzyme activity, measured as the rate of formation of PGE2 in brain extract, was significantly lower in brains from valproate-treated rats than in control rats (74.4 ± 7.8 vs. 138.0 ± 18.9 pg PGE2 per min per g cytosolic protein; n = 8, p < 0.01) (Fig. 3). Furthermore, the brain PGE2 concentration was decreased significantly (by 50%) in valproate-treated compared with control rats (26.2 ± 6.8 vs. 52.0 ± 6.4 ng per g wet-weight brain; n = 4, p < 0.05) (Fig. 4). The brain concentration of 11-dehydroTxB2 was also decreased significantly (by 74%) (n = 4, p < 0.05) in the valproate-treated group (8.0 ± 2.4 vs. 30.7 ± 9.0 pg per g wet-weight brain; n = 4, p < 0.05).

image

Figure 3. COX activity in valproate-treated and control rats. Brain cytosolic fraction from control (Con) and valproate-treated (VP) rats was assayed for COX activity. Data are mean ± SEM of eight independent samples for each group, each assayed in duplicate. Each sample was derived from an individual rat brain. **p < 0.01 versus control (Student's t-test).

Download figure to PowerPoint

image

Figure 4. Eicosanoid concentrations in microwaved brains from control (Con) and valproate-treated (VP) rats. Data are expressed as mean ± SEM of four independent samples, each assayed in duplicate. Each sample was derived from an individual rat brain. Brain eicosanoid concentrations were determined using specific immunoassay kits. *p < 0.05 versus control (Student's t-test).

Download figure to PowerPoint

No effect of valproate on 5-LO or LTB4

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

The ratio of 5-LO protein level to actin level was not significantly changed by valproate treatment (1.24 ± 0.19 vs. 1.33 ± 0.25; n = 6). No significant difference was found in the brain LTB4 concentration between valproate-treated and control rats (93.3 ± 13.2 vs. 80.8 ± 17.8 pg per g wet-weight brain; n = 4), which is consistent with evidence that the 5-LO protein level was not affected by valproate. Furthermore, the protein level of cytochrome P450 epoxygenase (CYP2C11) was not changed significantly (1.25 ± 0.13 vs. 1.39 ± 0.14; n = 6), indicating a selective effect of valproate on the COX pathway.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References

The biosynthesis of prostanoids within the AA cascade involves the PLA2-mediated release of AA from phospholipids, oxygenation of AA to form prostaglandin endoperoxide H2 by COX-1 and COX-2, and conversion of prostaglandin endoperoxide H2 to bioactive eicosanoids, prostaglandins, prostacyclin and thromboxanes, via specific synthases (Wolfe and Horrocks 1994; Herschman 1996; Smith et al. 2000; Fitzpatrick and Soberman 2001). Of the two COX isoforms, COX-1 is distributed throughout the brain, but is more abundant in the forebrain (Yamagata et al. 1993; Kam and See 2000). COX-2, which is not usually expressed in most tissues, is expressed constitutively and represents the predominant COX isoform in brain and spinal cord, where it may be involved in synaptic signaling (Kaufmann et al. 1996), cerebral blood flow (Niwa et al. 2000) and behavior (Yamamoto and Nozaki-Taguchi 1996).

In this study, we found that chronically administered valproate post-transcriptionally reduced COX-1 and COX-2 protein levels as well as net COX enzyme activity in rat brain. In addition, brain concentrations of 11-dehydroTxB2 and PGE2, AA products formed via the COX enzymes, were significantly reduced compared with control levels. In contrast, the protein level of 5-LO and the brain concentration of LTB4, produced by 5-LO-mediated oxidation of AA, were unaffected by valproate.

These results, and previous observations that chronic valproate reduces AA turnover within rat brain phospholipids (Chang et al. 2001), support the view that valproate down-regulates the part of the brain AA cascade that involves AA release and turnover within brain phospholipids, and conversion of AA by COX-1 and COX-2 to enzyme-specific bioactive eicosanoids. Because COX enzymes are activated irreversibly following catalysis, COX activity is thought to be determined by the amount of enzyme protein, and thus regulated predominantly by transcription and translation (Parfenova et al. 1998). As we found that chronic valproate reduced the protein levels of COX-1 and COX-2, activity changes in both isoenzymes may contribute to the observed decrease in total COX activity. However, we measured COX activity at saturating concentrations of AA that are probably absent under basal in vivo conditions. Although COX-1 and COX-2 have similar values of Km for AA, COX-2 metabolizes the bulk of available AA under conditions of limiting substrate (below 0.5 µm). In contrast, both isoforms metabolize AA, and COX-1 is slightly more active, at substrate concentrations above 2.5 µm (Swinney et al. 1997). These data suggest that the contribution of each isoenzyme in vivo depends on the AA concentration and/or its intracellular localization.

A number of the effects of chronic valproate on the brain AA cascade overlap with those produced by chronic lithium chloride administration. These effects include a decreased turnover of AA within brain phospholipids, reduced brain protein and activity levels of COX-2, and a reduced brain concentration of the COX-2-derived AA metabolite PGE2. On the other hand, lithium chloride reduces the brain arachidonoyl-CoA concentration (Chang et al. 2001), as well as the gene, protein and activity expression of cPLA2 (Chang et al. 1996; Chang and Jones 1998; Rintala et al. 1999). These observations support the hypothesis that drugs that target AA release and turnover with brain phospholipids and the conversion of AA to eicosanoids via COX-2 may in general be effective in mania and bipolar disorder. Although non-steroidal anti-inflammatory drugs or specific COX-2 inhibitors have not been tested in controlled trials to treat manic symptoms, their potential therapeutic effect is suggested by evidence that aspirin, a non-selective COX inhibitor, had a beneficial mood-modulating effect when used for anti-thrombosis (Ketterer et al. 1996).

The hypothesis that targeting AA release, recycling and conversion to PGE2 by COX-2 underlies the therapeutic actions of valproate and lithium is consistent with reports that dietary n-3 polyunsaturated fatty acid (PUFA) supplementation is beneficial in mood disorders (Stoll et al. 1999; Noaghiul and Hibbeln in press). This may be because n-3 PUFAs, such as docosahexaenoic acid (22:6 n-3) or eicosapentaenoic acid (20:5 n-3), can inhibit AA conversion to prostaglandins by COX-2 (Corey et al. 1983; Rubin and Laposata 1992; Mirnikjoo et al. 2001).

Whereas transcriptional down-regulation of cPLA2 can explain the ability of lithium to reduce AA turnover within brain phospholipids, the apparent absence of an effect of valproate on cPLA2 implies that valproate reduces AA turnover by another mechanism. One possibility, suggested by the observed radiolabeling of membrane phospholipids of cultured GT1-7 neurons exposed to [3H]valproate (Siafaka-Kapadai et al. 1998), is that valproate is incorporated directly into membrane phospholipids via valproyl-CoA, so as to reduce AA availability or inhibit cPLA2. However, this mechanism is unlikely, as valproyl-CoA, the activated form of valproate required for its esterification into a phospholipid, could not be detected in brains of rats treated with valproate (Becker and Harris 1983; Deutsch et al. in press). A second possibility is that chronic valproate reduces AA turnover by binding to the peroxisome proliferator-activator receptor (PPAR) (Lampen et al. 1999). PPARs comprise a family of three nuclear receptors (α, β and γ) which can heterodimerize with retinoid X receptors to inhibit gene expression of a number of genes linked to lipid metabolism (Kliewer et al. 1992) by antagonizing the effects of the positive transcription factors activator protein 1 (AP-1), signal transducers activators of transcription (STAT) and nuclear factor κB (Ricote et al. 1998). One target gene for PPAR encodes for an acyl-CoA synthetase that activates AA to arachidonoyl-CoA and regulates its recycling within phospholipids (Basu-Modak et al. 1999). Therefore, the decrease in AA turnover by valproate (Chang et al. 2001) may be related to an effect on the re-incorporation of AA into phospholipids via arachidonoyl-CoA synthetase or arachidonoyl-CoA transferase enzymes.

Studies in rats subjected to chronic administration of lithium chloride or valproate suggest that reduced AA substrate availability might lead to decreased protein expression of COX-2, but the mechanism for this is not evident. It may arise from a reduced rate of formation of PGE2 due to reduced AA availability. This would be expected to inhibit COX-2 translation mediated by the 3′-untranslated region of COX-2 mRNA (Faour et al. 2001). On the other hand, the reduction by valproate of COX-1 protein implies a lithium-independent mechanism.

The decrease in brain PGE2 concentration by valproate, as well as by lithium chloride, might have important therapeutic implications. AA and its bioactive products act as second messengers and can modulate brain ion channels, neurotransmitter uptake, blood flow, synaptic transmission, autonomic and behavioral responses, gene transcription, pain, fever and the sleep–wake cycle (O'Banion 1999). In addition to the known hypothalamic functions of PGE2, including fever induction, promotion of wakefulness, cardiovascular control and luteinizing hormone-releasing hormone (LH-RH) release, it has been suggested that PGE2 can modulate viscerosensory, somatosensory and visual inputs, as well as central integration of autonomic and limbic functions (Matsumura et al. 1992). Prostaglandin D2 and PGE2 play a key role in sleep–wake regulation and may be disturbed in patients with bipolar disorder, who usually show an abnormal circadian rhythm and sleep–wake cycle (Hayaishi 1999; Frank et al. 2000). It also has been suggested that affective disorders arise from an alteration in the immune system; an increase in the concentration of pro-inflammatory cytokines in the brain might alter the neurotransmitters as well as the hypothalamic–pituitary–adrenal axis and result in ‘sickness behavior’ (Leonard 2001).

The hypothesis of a dysfunctional AA cascade in mood disorders (Rapoport and Bosetti 2002) is supported by the finding that the level of PGE2 is increase in plasma, saliva and cerebrospinal fluid of subjects with major depression (Linnoila et al. 1983; Calabrese et al. 1986; Nishino et al. 1989), and by studies suggesting an association between bipolar disorder and genes involved in the cascade such as prostaglandin receptors, phospholipase A2 and COX-2 (Bennett and Horrobin 2000).

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and valproate administration
  5. Western blot analysis
  6. RNA isolation and RT–PCR
  7. Measurement of PLA2 activity
  8. Measurement of COX activity
  9. Eicosanoid concentrations
  10. Statistical analysis
  11. Results
  12. Valproate does not affect PLA2 activity, but down-regulates COX-1 and COX-2 protein levels, COX activity and concentrations of COX-derived eicosanoids
  13. No effect of valproate on 5-LO or LTB4
  14. Discussion
  15. Acknowledgement
  16. References
  • Ackermann E. J., Conde-Frieboes K. and Dennis E. A. (1995) Inhibition of macrophage Ca2+-independent phospholipase A2 by bromoenol lactone and trifluoromethyl ketones. J. Biol. Chem. 270, 445450.
  • Basu-Modak S., Braissant O., Escher P., Desvergne B., Honegger P. and Wahli W. (1999) Peroxisome proliferator-activated receptor beta regulates acyl-CoA synthetase 2 in reaggregated rat brain cell cultures. J. Biol. Chem. 274, 3588135888.
  • Becker C. M. and Harris R. A. (1983) Influence of valproic acid on hepatic carbohydrate and lipid metabolism. Arch. Biochem. Biophys. 223, 381392.
  • Bennett C. N. and Horrobin D. F. (2000) Gene targets related to phospholipid and fatty acid metabolism in schizophrenia and other psychiatric disorders: an update. Prostaglandins Leukot. Essent. Fatty Acids 63, 4759.
  • Bernard N., Sacquet J., Benzoni D. and Sassard J. (2000) Cyclooxygenases 1 and 2 and thromboxane synthase in kidneys of Lyon hypertensive rats. Am. J. Hypertens. 13, 404409.
  • Bosetti F., Rintala J., Seemann R., Rosenberger T. A., Contreras M. A., Rapoport S. I. and Chang M. C. (2002a) Chronic lithium downregulates cyclooxygenase-2 activity and prostaglandin E2 concentration in rat brain. Mol. Psychiatry 7, 844849.
  • Bosetti F., Seemann R., Bell J. M., Zahorchak R., Friedman E., Rapoport S. I. and Manickam P. (2002b) Analysis of gene expression with cDNA microarrays in rat brain after 7 and 42 days of oral lithium administration. Brain Res. Bull. 57, 205209.
  • Bowden C. L., Janicak P. G., Orsulak P., Swann A. C., Davis J. M., Calabrese J. R., Goodnick P., Small J. G., Rush A. J., Kimmel S. E., Risch S. C. and Morris D. D. (1996) Relation of serum valproate concentration to response in mania. Am. J. Psychiatry 153, 765770.
  • Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal. Biochem. 72, 248254.
  • Calabrese J. R., Skwerer R. G., Barna B., Gulledge A. D., Valenzuela R., Butkus A., Subichin S. and Krupp N. E. (1986) Depression, immunocompetence, and prostaglandins of the E series. Psychiatry Res. 17, 4147.
  • Chang M. C. and Jones C. R. (1998) Chronic lithium treatment decreases brain phospholipase A2 activity. Neurochem. Res. 23, 887892.
  • Chang M. C., Grange E., Rabin O., Bell J. M., Allen D. D. and Rapoport S. I. (1996) Lithium decreases turnover of arachidonate in several brain phospholipids. Neurosci. Lett. 220, 171174.
  • Chang M. C., Contreras M. A., Rosenberger T. A., Rintala J. J., Bell J. M. and Rapoport S. I. (2001) Chronic valproate treatment decreases the in vivo turnover of arachidonic acid in brain phospholipids: a possible common effect of mood stabilizers. J. Neurochem. 77, 796803.
  • Corey E. J., Shih C. and Cashman J. R. (1983) Docosahexaenoic acid is a strong inhibitor of prostaglandin but not leukotriene biosynthesis. Proc. Natl Acad. Sci. USA 80, 35813584.
  • Deutsch J., Rapoport S. I. and Rosenberger T. A. (2003)Valproyl-CoA and esterified valproate are not found in brains of rats treated with valproic acid, but the brain concentrations of CoA and acetyl-CoA are altered. Neurochem. Res.28, 861866.
  • Faour W. H., He Y., He Q. W., De Ladurantaye M., Quintero M., Mancini A. and Di Battista J. A. (2001) Prostaglandin E2 regulates the level and stability of cyclooxygenase-2 mRNA through activation of p38 mitogen-activated protein kinase in interleukin-1 beta-treated human synovial fibroblasts. J. Biol. Chem. 276, 3172031731.
  • Feng L., Xia Y., Garcia G. E., Hwang D. and Wilson C. B. (1995) Involvement of reactive oxygen intermediates in cyclooxygenase-2 expression induced by interleukin-1, tumor necrosis factor-alpha, and lipopolysaccharide. J. Clin. Invest. 95, 16691675.
  • Fitzpatrick F. A. and Soberman R. (2001) Regulated formation of eicosanoids. J. Clin. Invest. 107, 13471351.
  • Frank E., Swartz H. A. and Kupfer D. J. (2000) Interpersonal and social rhythm therapy: managing the chaos of bipolar disorder. Biol. Psychiatry 48, 593604.
  • Guay D. R. (1995) The emerging role of valproate in bipolar disorder and other psychiatric disorders. Pharmacotherapy 15, 631647.
  • Hayaishi O. (1999) Prostaglandin D2 and sleep – a molecular genetic approach. J. Sleep Res. 8, 6064.
  • Herschman H. R. (1996) Prostaglandin synthase 2. Biochim. Biophys. Acta 1299, 125140.
  • Iyer V. G., Reid K. H., Young C., Miller J. and Schurr A. (1995) Early, but not late, antiepileptic treatment reduces relapse of sound-induced seizures in the post-ischemic rat. Brain Res. 689, 159162.
  • Kam P. C. and See A. U. (2000) Cyclo-oxygenase isoenzymes: physiological and pharmacological role. Anaesthesia 55, 442449.
  • Kaufmann W. E., Worley P. F., Pegg J., Bremer M. and Isakson P. (1996) COX-2, a synaptically induced enzyme, is expressed by excitatory neurons at postsynaptic sites in rat cerebral cortex. Proc. Natl Acad. Sci. USA 93, 23172321.
  • Ketterer M. W., Brymer J., Rhoads K., Kraft P. and Lovallo W. R. (1996) Is aspirin, as used for antithrombosis, an emotion-modulating agent? J. Psychosom. Res. 40, 5358.
  • Kliewer S. A., Umesono K., Noonan D. J., Heyman R. A. and Evans R. M. (1992) Convergence of 9-cis retinoic acid and peroxisome proliferator signalling pathways through heterodimer formation of their receptors. Nature 358, 771774.
  • Lampen A., Siehler S., Ellerbeck U., Gottlicher M. and Nau H. (1999) New molecular bioassays for the estimation of the teratogenic potency of valproic acid derivatives in vitro: activation of the peroxisomal proliferator-activated receptor (PPARdelta). Toxicol. Appl. Pharmacol. 160, 238249.
  • Leonard B. E. (2001) The immune system, depression and the action of antidepressants. Prog. Neuropsychopharmacol. Biol. Psychiatry 25, 767780.
  • Leslie J. B. and Watkins W. D. (1985) Eicosanoids in the central nervous system. J. Neurosurg. 63, 659668.
  • Linnoila M., Whorton A. R., Rubinow D. R., Cowdry R. W., Ninan P. T. and Waters R. N. (1983) CSF prostaglandin levels in depressed and schizophrenic patients. Arch. Gen. Psychiatry 40, 405406.
  • Loscher W. and Honack D. (1995) Comparison of anticonvulsant efficacy of valproate during prolonged treatment with one and three daily doses or continuous (‘controlled release’) administration in a model of generalized seizures in rats. Epilepsia 36, 929937.
  • Matsumura K., Watanabe Y., Imai-Matsumura K., Connolly M., Koyama Y. and Onoe H. (1992) Mapping of prostaglandin E2 binding sites in rat brain using quantitative autoradiography. Brain Res. 581, 292298.
  • Mirnikjoo B., Brown S. E., Kim H. F., Marangell L. B., Sweatt J. D. and Weeber E. J. (2001) Protein kinase inhibition by omega-3 fatty acids. J. Biol. Chem. 276, 1088810896.
  • Nishino S., Ueno R., Ohishi K., Sakai T. and Hayaishi O. (1989) Salivary prostaglandin concentrations: possible state indicators for major depression. Am. J. Psychiatry 146, 365368.
  • Niwa K., Araki E., Morham S. G., Ross M. E. and Iadecola C. (2000) Cyclooxygenase-2 contributes to functional hyperemia in whisker-barrel cortex. J. Neurosci. 20, 763770.
  • Noaghiul S. and Hibbeln J. R. (2003) Cross-national comparisons of seafood consumption and rates of bipolar disorder. Am. J. Psychiatry (in press).
  • O'Banion M. K. (1999) Cyclooxygenase-2: molecular biology, pharmacology, and neurobiology. Crit. Rev. Neurobiol. 13, 4582.
  • Parfenova H., Balabanova L. and Leffler C. W. (1998) Posttranslational regulation of cyclooxygenase by tyrosine phosphorylation in cerebral endothelial cells. Am. J. Physiol. 274, C72C81.
  • Post R. M., Ketter T. A., Denicoff K., Pazzaglia P. J., Leverich G. S., Marangell L. B., Callahan A. M., George M. S. and Frye M. A. (1996) The place of anticonvulsant therapy in bipolar illness. Psychopharmacology (Berl) 128, 115129.
  • Powell W. S. (1985) Reversed-phase high-pressure liquid chromatography of arachidonic acid metabolites formed by cyclooxygenase and lipoxygenases. Anal. Biochem. 148, 5969.
  • Rapoport S. I. and Bosetti F. (2002) Do lithium and anticonvulsants target the brain arachidonic acid cascade in bipolar disorder? Arch. Gen. Psychiatry 59, 592596.
  • Rapoport S. I., Purdon D., Shetty H. U., Grange E., Smith Q., Jones C. and Chang M. C. (1997) In vivo imaging of fatty acid incorporation into brain to examine signal transduction and neuroplasticity involving phospholipids. Ann. N. Y. Acad. Sci. 820, 5673; discussion 73–54.
  • Reynolds L. J., Hughes L. L., Yu L. and Dennis E. A. (1994) 1-Hexadecyl-2-arachidonoylthio-2-deoxy-sn-glycero-3-phosphorylcholine as a substrate for the microtiterplate assay of human cytosolic phospholipase A2. Anal. Biochem. 217, 2532.
  • Ricote M., Li A. C., Willson T. M., Kelly C. J. and Glass C. K. (1998) The peroxisome proliferator-activated receptor-gamma is a negative regulator of macrophage activation. Nature 391, 7982.
  • Rintala J., Seemann R., Chandrasekaran K., Rosenberger T. A., Chang L., Contreras M. A., Rapoport S. I. and Chang M. C. (1999) 85 kDa cytosolic phospholipase A2 is a target for chronic lithium in rat brain. Neuroreport 10, 38873890.
  • Rubin D. and Laposata M. (1992) Cellular interactions between n-6 and n-3 fatty acids: a mass analysis of fatty acid elongation/desaturation, distribution among complex lipids, and conversion to eicosanoids. J. Lipid Res. 33, 14311440.
  • Shimizu T. and Wolfe L. S. (1990) Arachidonic acid cascade and signal transduction. J. Neurochem. 55, 115.
  • Siafaka-Kapadai A., Patiris M., Bowden C. and Javors M. (1998) Incorporation of [3H]valproic acid into lipids in GT1-7 neurons. Biochem. Pharmacol. 56, 207212.
  • Smith W. L., DeWitt D. L. and Garavito R. M. (2000) Cyclooxygenases: structural, cellular, and molecular biology. Annu. Rev. Biochem. 69, 145182.
  • Stoll A. L., Severus W. E., Freeman M. P., Rueter S., Zboyan H. A., Diamond E., Cress K. K. and Marangell L. B. (1999) Omega 3 fatty acids in bipolar disorder: a preliminary double-blind, placebo-controlled trial. Arch. Gen. Psychiatry 56, 407412.
  • Swinney D. C., Mak A. Y., Barnett J. and Ramesha C. S. (1997) Differential allosteric regulation of prostaglandin H synthase 1 and 2 by arachidonic acid. J. Biol. Chem. 272, 1239312398.
  • Szupera Z., Mezei Z., Kis B., Gecse A., Vecsei L. and Telegdy G. (2000) The effects of valproate on the arachidonic acid metabolism of rat brain microvessels and of platelets. Eur. J. Pharmacol. 387, 205210.
  • Wolfe L. S. and Horrocks L. A. (1994) Eicosanoids, in Basic Neurochemistry (SiegelG. J., AgranoffB. W., AlbersR. W. and MolinoffP. B., eds), 5th edn, pp. 475490. Raven Press, New York.
  • Yamagata K., Andreasson K. I., Kaufmann W. E., Barnes C. A. and Worley P. F. (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron 11, 371386.
  • Yamamoto T. and Nozaki-Taguchi N. (1996) Analysis of the effects of cyclooxygenase (COX)-1 and COX-2 in spinal nociceptive transmission using indomethacin, a non-selective COX inhibitor, and NS-398, a COX-2 selective inhibitor. Brain Res. 739, 104110.
  • Zarate J. C. A. and Toehn M. (1996) Epidemiology of mood disorders throughout the life cycle, in Mood Disorders Across the Life Span (Shulman K. I., Toehn M. and Kutcher S. P., eds), pp. 1733. Wiley–Liss, New York.