Neuroprotective effect of chronic lithium ... - Wiley Online Library

Copyright ª Blackwell Munksgaard 2008 Bipolar Disorders 2008: 10: 360–368

BIPOLAR DISORDERS

Original Article

Neuroprotective effect of chronic lithium treatment against hypoxia in specific brain regions with upregulation of cAMP response element binding protein and brain-derived neurotrophic factor but not nerve growth factor: comparison with acute lithium treatment Omata N, Murata T, Takamatsu S, Maruoka N, Mitsuya H, Yonekura Y, Fujibayashi Y, Wada Y. Neuroprotective effect of chronic lithium treatment against hypoxia in specific brain regions with upregulation of cAMP response element binding protein and brain-derived neurotrophic factor but not nerve growth factor: comparison with acute lithium treatment. Bipolar Disord 2008: 10: 360–368. ª Blackwell Munksgaard, 2008

N Omataa, T Murataa, S Takamatsub, N Maruokaa, H Mitsuyaa, Y Yonekurab, Y Fujibayashib and Y Wadaa a

Department of Neuropsychiatry, bBiomedical Imaging Research Center, University of Fukui, Fukui, Japan

Objectives: We evaluated the neuroprotective effect of chronically or acutely administered lithium against hypoxia in several brain regions. Furthermore, we investigated the contribution of brain-derived neurotrophic factor (BDNF), nerve growth factor (NGF), and cAMP response element binding protein (CREB) to the neuroprotective effect of lithium. Methods: Brain slices were prepared from rats that had been treated chronically or acutely with lithium. The cerebral glucose metabolic rate (CMRglc) before and after hypoxia loading to brain slices was measured using the dynamic positron autoradiography technique with [18F]2fluoro-2-deoxy-d-glucose. The changes of expression of proteins were investigated using Western blot analysis. Results: Before hypoxia loading, the CMRglc did not differ between the lithium-treated and untreated groups. After hypoxia loading, the CMRglc of the untreated group was significantly lower than that before hypoxia loading. However, the CMRglc of the chronic lithium treatment group recovered in the frontal cortex, caudate putamen, hippocampus and cerebellum, but not in the thalamus. In contrast, the CMRglc of the acute lithium treatment group did not recover in any analyzed brain regions. After chronic lithium treatment, the levels of expression of BDNF and phospho-CREB were higher than those of untreated rats in the frontal cortex, but not in the thalamus. However, the expression of NGF did not change in the frontal cortex and thalamus. Conclusions: These results demonstrated that lithium was neuroprotective against hypoxia only after chronic treatment and only in specific brain regions, and that CREB and BDNF might contribute to this effect.

The authors of this paper do not have any commercial associations that might pose a conflict of interest in connection with this manuscript.

360

Key words: BDNF – brain slice – chronic treatment – CREB – lithium – neuroprotection – positron – regional differences Received 9 August 2006; revised and accepted for publication 19 March 2007 Corresponding author: Tetsuhito Murata, Department of Neuropsychiatry, University of Fukui, Fukui 910-1193, Japan. Fax: +81 776 61 8136; e-mail: [email protected]

The efficacy of lithium against bipolar disorder (BD) was demonstrated more than 50 years ago; even now, it is often used clinically (1). Lithium shows superior effects in the treatment and

Region-specific neuroprotection by lithium prophylaxis of manic and depressive episodes (2). However, in spite of its frequent use, the cellular and molecular bases of its therapeutic effect remain poorly understood. Recently, it has been reported that lithium activated intracellular signaling pathways that are neuroprotective (3, 4) and inhibited cascades that promote neuronal cell death (5, 6). Furthermore, lithium administration increases the gray matter volume of bipolar patients (7, 8). Those studies imply that the neuroprotective effects of lithium are important for its therapeutic actions. Some drugs show neuroprotective effects against several stresses after acute treatment, whereas some drugs require chronic treatment. For example, acute treatment with N-methyl-d-aspartate (NMDA) receptor antagonists is neuroprotective against hypoxia (9), whereas chronic antioxidant treatment is required to prevent ischemic neuronal damage (10). Clinically, acute lithium administration is hardly effective against BD; continuous administration is necessary to obtain its therapeutic effect. Therefore, it is important to compare the neuroprotective effects of chronically or acutely administered lithium to clarify the relationship between the length of treatment and therapeutic actions. Furthermore, the neuroprotective effect of lithium has often been studied using cultures or in specific brain regions such as the hippocampus (11, 12), but it is not likely that the effect of a certain drug occurs equally throughout the entire brain. Recent studies suggest that the plasticity and the response against drugs are not uniform among different brain regions, and that these regional differences are related to symptoms of stressinduced disorders (13, 14). Therefore, the neuroprotective effect of lithium should be studied in several brain regions. It has been suggested that the changes of expression of several proteins such as neurotrophins and their modulators are related to the neuroprotective effect of lithium (4, 15). Continuous lithium administration is necessary until a patientÕs recovery, and the therapeutic effect continues for some time even after stopping its use. These clinical findings imply that changes of expression of genes or proteins are related to the therapeutic effect of lithium. Therefore, it is important to examine changes of expression of proteins, especially neurotrophins and their modulators, after lithium treatment. We developed a dynamic positron autoradiography technique (dPAT) to image metabolic changes in rodent brain slices (9, 16). Because of the high specific radioactivity of the radiotracers, serial two-dimensional images of radioactivity in the slices can be constructed quantitatively while the brain tissue remains alive in the incubation solution. In

dPAT, we can control the extracellular environment precisely before and after loading drugs or stress. Moreover, it is possible to examine the neuroprotective effects in each brain region under identical conditions regarding the maintenance of cerebral glucose metabolic rate (CMRglc) as a parameter of viability using dPAT with [18F]2fluoro-2-deoxy-d-glucose ([18F]FDG). In previous studies, the neuroprotective effect of lithium against ischemia (insufficient supply of blood flow) was examined thoroughly (17–19), but no reports in the literature have described the effect of lithium against hypoxia (insufficient supply of oxygen). Neuronal damage caused by hypoxia differs from that caused by ischemia (9). In this study, we evaluated the neuroprotective effect of chronically or acutely administered lithium against hypoxia in several brain regions using dPAT. Lithium promotes the insulin pathway (20), and this effect might have some influence on CMRglc; therefore, we also evaluated the influence of insulin on CMRglc using dPAT. In addition, using Western blot analysis, we investigated the changes of expression of brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) as neurotrophins, and cAMP response element binding protein (CREB) as possible modulators after lithium treatment.

Methods

Animals and lithium treatment protocol

All protocols were consistent with the NIH policy on the use of animals in experimental research. The experiments were approved by the Institutional Animal Care Committee at the University of Fukui. Six-week-old male Wistar rats were used. The animals were bred under standard conditions, housed at 24 ± 1C with a light ⁄ dark cycle set to 12 h ⁄ 12 h, and allowed free access to food and water. We gave lithium-containing water (lithium treatment group) or water containing no lithium (untreated group) orally to animals for 1 day (acute lithium treatment), 7 days (subacute lithium treatment) or 14 days (chronic lithium treatment) prior to starting dPAT or Western blot analysis. A solution of 12 mm was used as lithium-containing water. Lithium chloride was purchased from Wako Pure Chemical Industries Ltd. (Osaka, Japan). These protocols of lithium treatment for 1, 7 and 14 days resulted in mean (SEM) serum lithium concentrations of 0.85 (0.07), 0.80 (0.11) and 0.84 (0.14) mEq ⁄ L, respectively. These serum levels are also close to the range of concentrations that are achieved and maintained during lithium therapy in human subjects (0.4–1.0 mEq ⁄ L).

361

Omata et al. Dynamic positron autoradiography technique

Dynamic positron autoradiography technique (dPAT) was performed as described previously (9, 16). Briefly, under diethyl ether anesthesia, rats of lithium-treated and untreated groups were decapitated; the brains were removed quickly. Sagittal brain slices (300 lm thick) were prepared using a Microslicer (DTK-2000; Dosaka EM, Kyoto, Japan). After one hour of pre-incubation, they were incubated in Krebs Ringer solution (having the following composition in mm concentration: NaCl 124, KCl 5, MgCl2 1, CaCl2 2, KH2PO4 1.2, NaHCO3 26, glucose 10; 36C, pH 7.3–7.4 and bubbled with 95% O2 ⁄ 5% CO2 gas) containing [18F]FDG diluted to 150 kBq ⁄ mL. The exposed imaging plate (BAS-MP 2040S; Fuji Photo Film Co. Ltd., Tokyo, Japan; serially replaced with another plate every 10 min) was scanned using a BAS-1500 (Fuji Photo Film Co. Ltd.). Loading and relief of hypoxia were conducted by exchanging the solution for another solution bubbled in advance with either 95% N2 ⁄ 5% CO2 gas or 95% O2 ⁄ 5% CO2 gas with the same amount of [18F]FDG. A three-compartment model using the Gjedde– Patlak graphical method was applied to the image data for determination of the fractional rate constant for phosphorylation of [18F]FDG (=k3*, proportional to the CMRglc) (9, 16, 21, 22). The value of k3* is estimated from the slope of the linear portion of the graph, Ci*(t) ⁄ Cp*(t) expressed in terms of the radioactivity signal ratio on the imaging plate (y-axis) versus time in minutes (x-axis), where Ci*(t) is the total brain tissue radioactivity and Cp*(t) is the input function (9, 16, 23, 24). As the hypoxia-loading time, 20 min, the minimum time after which k3* did not recover to the preloading level, was selected (9, 16). A neuroprotective effect was considered to have been confirmed when k3* of the lithium-treated group after re-oxygenation was significantly higher than that of the untreated group. Data were analyzed using a Macintosh computer (Apple Computer Inc., Cupertino, CA, USA) and MacBAS software version 2 (Fuji Photo Film Co. Ltd.). The presented values are mean (SEM) of eight slices obtained from four rats in two experiments. The Mann– Whitney U-test was used to evaluate the significance of differences, and a p-value of less than 0.05 was considered significant. Investigation of influence of insulin on CMRglc using dPAT

Six-week-old male Wistar rats were used. dPAT was performed as in the examination of the

362

neuroprotective effect of lithium. Briefly, brain slices were prepared and incubated in Krebs Ringer solution containing [18F]FDG. The brain slices were then treated with 5 lg ⁄ mL insulin. Insulin was purchased from Sigma Chemical Co. (St. Louis, MO, USA). A three-compartment model using the Gjedde–Patlak graphical method was applied to the image data for determination of the k3* value. The presented values are the mean (SEM) of eight slices obtained from four rats in two experiments. Western blot analysis

Target brain regions were isolated from lithiumtreated and untreated rats, and homogenized in CelLytic MT (Sigma Chemical Co.) as lysis buffer. Protein concentrations were determined and aliquots of 50 lg of the total proteins were separated using electrophoresis on SDS-polyacrylamide gels (10%). Proteins were subsequently transferred to polyvinylidene difluoride membranes; the membranes were incubated with primary antibody (1:2,000). Antibodies for BDNF and NGF were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA); antibodies for phosphoCREB and total CREB were purchased from Sigma Chemical Co. Membranes were washed using Tris-buffered saline ⁄ Tween. The membranes were then incubated with secondary antibody. Western blots were visualized using an ECL Advanced Western Blotting Detection Kit (Amersham Biosciences, Piscataway, NJ, USA) and Kodak BioMax MR film (Eastman Kodak Co., Rochester, NY, USA). We performed Western blot analysis four times using different samples for each target protein, and quantified the Western blots using an FLA-7000 and Multi Gauge software version 3.0 (Fuji Photo Film Co. Ltd.). Immunoreactivity compared to that of the untreated control is presented as mean (SEM). StudentÕs t-test was used to evaluate the significance of differences, and a p-value of less than 0.05 was considered significant.

Results

Investigation of neuroprotective effect of lithium using dPAT

Figure 1 shows the images of [18F]FDG uptake of two typical slices of the chronic lithium treatment group and the untreated group at two representative times, 120 min (before hypoxia loading) and 270 min (after hypoxia loading). Gjedde–Patlak plots in the frontal cortex and thalamus with 20-min hypoxia loading to the chronic lithium

Region-specific neuroprotection by lithium

Fig. 1. Time-resolved pseudo-color images of [18F]2-fluro-2deoxy-D-glucose ([18F]FDG) uptake in sagittally sectioned rat brain slices. Time zero (t = 0) is when [18F]FDG was introduced into the bathing solution containing brain slices. Two typical slices of the chronic lithium treatment group (A) and the untreated group (B) focusing on two representative times, 120 min (before hypoxia loading) and 270 min (after hypoxia loading). Filled circles in the diagram represent the five brain regions examined in the present study. Fr = frontal cortex; CPu = caudate putamen; T = thalamus; H = hippocampus; Cb = cerebellum.

Fig. 2. Effect of chronic lithium treatment against hypoxia as shown on the Gjedde–Patlak plots of [18F]2-fluro-2-deoxy-Dglucose uptake in the frontal cortex and thalamus. Ordinate: Ci*(t) ⁄ Cp*(t) expressed in terms of the radioactivity signal ratio on the imaging plate, versus abscissa: time (min). The time period of hypoxia loading is shown in the figure. Values are means (SEM) obtained in eight slices (SEM is shown only for the uppermost and lowermost lines).

treatment and untreated groups are indicated in Fig. 2. The k3* value of the chronic lithium treatment group before hypoxia loading was not significantly different from that of the untreated group. Therefore, chronic lithium treatment was inferred to have little influence on the basic brain glucose metabolism. The k3* value was markedly increased by hypoxia loading over the value before hypoxia loading, suggesting the inhibition of oxidative phosphorylation by hypoxia at the respiratory chain level and enhanced anaerobic glycolysis, compensating for the resultant decrease in aerobic glucose metabolism (inverse Pasteur effect). In the untreated group, the k3* value after re-oxygenation was significantly lower than that before hypoxia loading; therefore, 20-min hypoxia loading was inferred to be lethal. In the frontal cortex of the chronic lithium treatment group, the k3* value after re-oxygenation was significantly higher than that of the untreated group, indicating the maintenance of CMRglc after hypoxic loading and suggesting the neuroprotective effect of chronic lithium treatment. Similar findings were obtained in the caudate putamen, hippocampus, and cerebellum (Table 1). However, in the thalamus of the chronic lithium treatment group, the k3* value after re-oxygenation did not differ significantly from that of the untreated group. Consequently, a neuroprotective effect of chronic lithium treatment was not confirmed. Next, Fig. 3 shows Gjedde– Patlak plots in the frontal cortex and thalamus with 20-min hypoxia loading to the acute lithium treatment and untreated groups. In the frontal cortex and thalamus of the acute lithium treatment group, the k3* values after re-oxygenation did not differ significantly from those of the untreated group. Therefore, a neuroprotective effect of acute lithium treatment was not confirmed. Similar findings were obtained in the caudate putamen, hippocampus, and cerebellum (Table 2). Figure 4

Table 1. Effect of chronic lithium treatment against hypoxia on the fractional rate constant of [18F]2-fluro-2-deoxy-D-glucose ([l8F]PDG) Untreated group

Frontal cortex Caudate putamen Thalamus Hippocampus Cerebellum

Chronic lithium treatment group

Before hypoxia loading

After hypoxia loading

5.48 5.99 6.20 5.92 6.02

0.18 0.68 0.71 0.42 0.87

(1.02) (1.11) (0.87) (0.78) (0.76)

(0.40)a (0.52)a (0.45)a (0.87)a (0.56)a



5.42 5.51 6.34 5.85 5.87

5.18 4.94 0.77 5.38 5.44

(0.67) (0.30) (0.85) (0.74) (0.98)

(0.89)b (0.80)b (0.56) (0.74)b (0.83)b

Values are means (SEM). The k3* (·1000), indicating the fractional rate constant of [l8F]PDG, was obtained from the slope of the regression equation fitted to Gjedde–Patlak plots using the linear regression analysis. a p < 0.05; significantly different from that before hypoxia loading of the untreated group. b p < 0.05; significantly different from that after hypoxia loading of the untreated group.

363

Omata et al.

lithium treatment, and in the frontal cortex but not in the thalamus. Therefore, as the next experiment, the changes of expression of BDNF, NGF, phospho-CREB and total CREB after chronic lithium treatment were investigated in the frontal cortex and thalamus. The expression levels of BDNF and phospho-CREB, compared to those of the untreated group, were elevated in the frontal cortex but not in the thalamus. In contrast, the expression of NGF and total CREB did not change in the frontal cortex or thalamus (Fig. 6). Discussion Fig. 3. Effect of acute lithium treatment against hypoxia as shown on the Gjedde–Patlak plots of [18F]2-fluro-2-deoxy-Dglucose uptake in the frontal cortex and thalamus.

shows Gjedde–Patlak plots in the frontal cortex and thalamus with 20-min hypoxia loading to the subacute lithium treatment and untreated groups. A neuroprotective effect of subacute lithium treatment was not confirmed in any analyzed brain regions (Table 3). Investigation of influence of insulin on CMRglc using dPAT

Figure 5 shows Gjedde–Patlak plots in the frontal cortex and thalamus with insulin. The k3* value after insulin treatment was not significantly different from that before insulin treatment in the frontal cortex and thalamus. Similar findings were obtained in the caudate putamen, hippocampus, and cerebellum (Table 4). Investigation of protein expression changes after chronic lithium treatment using Western blot analysis

In the experiment using dPAT, a neuroprotective effect of lithium was confirmed only after chronic

In studies using dPAT, it is very important to confirm the relationship between the maintenance of CMRglc and neuronal viability. Alamar Blue (AccuMed International Inc., Westklake, OH, USA) is an indicator of metabolic activity, and this dye is reduced by mitochondrial respiration (25). Previously we evaluated the brain slice neuronal viability after hypoxia loading using Alamar Blue (26). When CMRglc showed a significant decrease relative to control values, Alamar Blue reduction similarly showed corresponding significant changes, and when CMRglc did not show significant changes, Alamar Blue reduction did not show significant changes either. Therefore, neuronal viability using maintenance of CMRglc as an index and neuronal viability using Alamar Blue reduction as an index were consistent. In this study using dPAT, a neuroprotective effect of chronically administered lithium against hypoxia was confirmed in specific brain regions, but neuroprotective effects of acutely or subacutely administered lithium were not confirmed in any analyzed brain regions. In previous reports studying neuroprotective effects of lithium, the distinction between the effects of chronically administered and acutely administered lithium was

Table 2. Effect of acute lithium treatment against hypoxia on the fractional rate constant of [18F]2-fluro-2-deoxy-D-glucose ([18F]FDG) Untreated group


Acute lithium treatment group



5.74 5.99 6.78 6.52 6.10

0.49 0.38 0.67 0.53 0.72

(0.81) (0.73) (0.74) (0.90) (0.67)

(0.46)a (0.42)a (0.57)a (0.52)a (0.47)a



5.45 6.01 5.71 6.20 5.94

0.42 0.34 0.55 0.47 0.64

(0.98) (1.01) (0.86) (0.87) (0.93)

(0.72) (0.74) (0.87) (0.82) (0.71)

Values are means (SEM). The k3* (·1000), indicating the fractional rate constant of [l8F]PDG, was obtained from the slope of the regression equation fitted to Gjedde–Patlak plots using the linear regression analysis. a p < 0.05; significantly different from that before hypoxia loading of the untreated group. There was no significant difference in after hypoxia loading between the untreated and acute lithium treatment groups.

364

Region-specific neuroprotection by lithium

Fig. 4. Effect of subacute lithium treatment against hypoxia as shown on the Gjedde–Patlak plots of [18F]2-fluro-2-deoxy-Dglucose uptake in the frontal cortex and thalamus.

not complete. However, the present study revealed that chronic treatment, but not acute treatment, was necessary to show a neuroprotective effect. The mechanisms of acute responses against stress are different from those of chronic responses in neuronal tissues (27). Transient alteration of receptor or ion channel functions on cell membranes contributes to the acute response, and NMDA receptor antagonists induce neuroprotection against hypoxia after their acute administration (9). On the other hand, changes of the expression of genes or proteins are thought to contribute to the chronic response (28). Reportedly, expression of some neurotrophins or their modulators changed after chronic lithium treatment (15, 29). Furthermore, Fukumoto et al. (4) demonstrated that the expression of BDNF, which is a representative neurotrophin, increased after 14 days but not after 1 or 7 days of lithium treatment in the rat brain, so the stage in which the expression of BDNF increased in their study and

the stage in which a neuroprotective effect of lithium was confirmed in our study were consistent. Therefore, the neuroprotective effect of lithium, which was observed only after chronic administration in this study, might be dependent upon changes of the expression of these proteins. Ren et al. (19) demonstrated that post-insult treatment with lithium reduced brain damage induced by cerebral ischemia. However, in this study, we examined the neuroprotective effect of lithium against hypoxia. Both hypoxia and ischemia cause neuronal damage, but their mechanisms differ (9). It has also been reported that DNA degradation is provoked by hypoxia but not ischemia (30), suggesting that hypoxia is more neurotoxic than ischemia. These factors might have contributed to the different neuroprotective effects of acute lithium treatment in this study and the previous report. Several reports have thoroughly investigated the neuroprotective effect of lithium against ischemia (17–19), but no reports have described the effect of lithium against hypoxia. Results of this study showed that chronic administration of lithium was neuroprotective against hypoxia. To our knowledge, this is the first demonstration of the neuroprotective effect of lithium against hypoxia. We confirmed the neuroprotective effect of chronic lithium treatment in the frontal cortex, caudate putamen, hippocampus, and cerebellum, but not in the thalamus. Previous studies have often examined the neuroprotective effect of lithium using cultures or in specific brain regions such as the hippocampus (11, 12, 29), but little examination of regional differences has been undertaken. The findings of the present study suggested that lithium does not exert neuroprotection in all brain regions uniformly. Plasticity and the response to drugs are not uniform in different brain regions;

Table 3. Effect of subacute lithium treatment against hypoxia on the fractional rate constant of [18F]2-fluro-2-deoxy-D-glucose ([18F]FDG) Untreated group


Subacute lithium treatment group



5.64 5.73 6.65 6.43 6.14

0.37 0.62 0.62 0.48 0.62

(0.87) (0.70) (0.92) (0.91) (0.71)

(0.52)a (0.75)a (0.52)a (0.56)a (0.64)a



5.57 5.89 6.02 6.23 5.98

0.50 0.43 0.51 0.48 0.70

(0.93) (1.02) (0.69) (0.82) (0.96)

(0.58) (0.72) (0.89) (0.75) (0.73)

Values are means (SEM). The k3* (·1000), indicating the fractional rate constant of [l8F]PDG, was obtained from the slope of the regression equation fitted to Gjedde–Patlak plots using the linear regression analysis. a p < 0.05; significantly different from that before hypoxia loading of the untreated group. There was no significant difference in after hypoxia loading between the untreated and subacute lithium treatment groups.

365

Omata et al.

Fig. 5. Effect of insulin as shown on the Gjedde–Patlak plots of [18F]FDG uptake in the frontal cortex and thalamus. The point at which insulin was administered is indicated in the figure. Values are the means (SEM) obtained in eight slices (SEM is shown only for the uppermost and lowermost lines).

18

Table 4. Effect of insulin on the fractional rate constant of [ F]2-fluro-2deoxy-D-glucose ([18F]FDG)


Before insulin administration

After insulin administration

5.51 5.87 5.78 5.56 6.11

5.34 5.38 5.62 5.29 6.03

(0.95) (0.88) (0.82) (0.73) (0.91)

(1.12) (0.82) (1.01) (0.90) (0.96)

Values are means (SEM). The k3* (·1000), indicating the fractional rate constant of [l8F]PDG, was obtained from the slope of the regression equation fitted to Gjedde–Patlak plots using the linear regression analysis. There was no significant difference between before and after lithium administration.

different protein expression changes are thought to contribute to these regional differences (13, 14). Chronic lithium treatment increased the expression of BDNF in the rat cerebral cortical neurons (29) and in the rat hippocampus (4). On the other hand, it was suggested that the protein expression might change less markedly in the thalamus after chronic lithium treatment compared to the frontal cortex or hippocampus (31). Therefore, regional differences of the neuroprotective effect of lithium found in this study might depend on the different patterns of expression of proteins that are related to neuroprotection. Lithium promotes the insulin pathway (20), and insulin promotes glycolysis. Therefore, this effect might have some influence on CMRglc. However, in dPAT, CMRglc after insulin treatment was not significantly different from that before

366

Fig. 6. Changes of expression of proteins after chronic lithium treatment in the frontal cortex and thalamus using Western blot analysis. (A) Representative Western blots of BDNF, NGF, phospho-CREB and total CREB. (B) Relative BDNF, NGF, phospho-CREB and total CREB content. Results are expressed as a percentage of untreated control and presented as means (SEM) from four independent experiments. *p < 0.05. BDNF = brain-derived neurotrophic factor; NGF = nerve growth factor; CREB = cAMP response element binding protein.

insulin treatment. Therefore, insulin treatment was inferred to have little influence on CMRglc in this study. As described above, the region-specific neuroprotective effect of chronically administered lithium confirmed in this study might depend on the region-specific changes of expression of proteins like neurotrophins or their modulators. Therefore, as the next experiment, we investigated the changes of expression of BDNF and NGF as representative neurotrophins and CREB as a modulator after chronic lithium treatment using Western blot analysis in the frontal cortex, where a neuroprotective effect was confirmed, and the thalamus, where it was not confirmed. Compared to the levels of the untreated group, the expression of BDNF and phospho-CREB was elevated only in the frontal cortex without an elevation of the expression of total CREB. CREB exerts its transcriptional efficacy after its phosphorylation (32). In addition, BDNF transcription is regulated by phospho-CREB (33), and BDNF is neuroprotective against several stresses (34, 35). Results of these reports suggest that the activation of CREB followed by the elevation of BDNF expression might contribute to the neuroprotective effect of lithium confirmed in this study. On the other hand,

Region-specific neuroprotection by lithium the expression of NGF after chronic lithium treatment did not change in the frontal cortex and thalamus. Therefore, NGF might not contribute strongly to the neuroprotective effect of lithium. The spatial distribution pattern of BDNF differs from that of NGF (36). Under physiological conditions, NGF is synthesized by peripheral target tissues, whereas BDNF synthesis is higher in the central nervous system (37). In spite of high structural homology, these neurotrophins operate via different receptors. The neurotrophins differ in target specificity and responses to injury of the neuronal tissue (36). For example, BDNF protected cerebellar granule neurons from glucose deprivation-induced cell death, but NGF did not (38). Furthermore, it was reported that the continuous intramedullary infusion of BDNF, but not NGF, provided neuroprotection and enhanced some regenerative activity after spinal cord injury in adult rats (39). These findings might be related to the different contributions of BDNF and NGF to the neuroprotective effect of lithium found in this study. In this study, the serum lithium concentration after chronic lithium treatment was 0.84 ± 0.14 mEq ⁄ L, which was within therapeutic levels in human subjects. Continuous lithium administration is necessary to exert its therapeutic effect, and a neuroprotective effect of lithium was confirmed in this study only after chronic lithium treatment. Abnormal glucose metabolism and blood flow in the frontal cortex or hippocampus of mood disorder patients have been reported in studies using positron emission tomography or functional magnetic resonance imaging (40, 41). It is interesting that these regions are the same brain regions in which neuroprotective effects of lithium were confirmed in this study. We showed elevation of BDNF expression in these brain regions; moreover, it was reported previously that BDNF infused in the rat hippocampus caused an antidepressant effect (42). These observations and considerations suggest that the neuroprotective effect of lithium is related to its clinical action. However, the mechanism of the therapeutic effects of lithium cannot be explained only by the results obtained in this study. The relation between the neuroprotective effect of lithium and its clinical action is not fully clarified. Therefore, further studies are necessary to elucidate the precise therapeutic action of lithium. Acknowledgements This work was supported in part by the 21st Century COE program ÔBiomedical Imaging Technology Integration

ProgramÕ of the Japan Society for the Promotion of Science (JSPS) and a Research and Development Project Aimed at Economic Revitalization (Leading Project), ÔResearch and Development of Technology for Measuring Vital Functions Merged with Optical TechnologyÕ from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.

References 1. Kessing LV, Sondergard L, Kvist K, Andersen PK. Suicide risk in patients treated with lithium. Arch Gen Psychiatry 2005; 62: 860–866. 2. Hartong EG, Moleman P, Hoogduin CA, Broekman TG, Nolen WA; LitCar Group. Prophylactic efficacy of lithium versus carbamazepine in treatment-naive bipolar patients. J Clin Psychiatry 2003; 64: 144–151. 3. Chen RW, Chuang DM. Long term lithium treatment suppresses p53 and Bax expression but increases Bcl-2 expression. A prominent role in neuroprotection against excitotoxicity. J Biol Chem 1999; 274: 6039–6042. 4. Fukumoto T, Morinobu S, Okamoto Y, Kagaya A, Yamawaki S. Chronic lithium treatment increases the expression of brain-derived neurotrophic factor in the rat brain. Psychopharmacology (Berl) 2001; 158: 100–106. 5. Mora A, Sabio G, Risco AM et al. Lithium blocks the PKB and GSK3 dephosphorylation induced by ceramide through protein phosphatase-2A. Cell Signal 2002; 14: 557–562. 6. Ghribi O, Herman MM, Spaulding NK, Savory J. Lithium inhibits aluminum-induced apoptosis in rabbit hippocampus, by preventing cytochrome c translocation, Bcl-2 decrease, Bax elevation and caspase-3 activation. J Neurochem 2002; 82: 137–145. 7. Manji HK, Drevets WC, Charney DS. The cellular neurobiology of depression. Nat Med 2001; 7: 541–547. 8. Glitz DA, Manji HK, Moore GJ. Mood disorders: treatment-induced changes in brain neurochemistry and structure. Semin Clin Neuropsychiatry 2002; 7: 269–280. 9. Omata N, Murata T, Fujibayashi Y et al. Hypoxic but not ischemic neurotoxicity of free radicals revealed by dynamic changes in glucose metabolism of fresh rat brain slices on positron autoradiography. J Cereb Blood Flow Metab 2000; 20: 350–358. 10. Cao DH, Xu JF, Xue RH, Zheng WF, Liu ZL. Protective effect of chronic ethyl docosahexaenoate administration on brain injury in ischemic gerbils. Pharmacol Biochem Behav 2004; 79: 651–659. 11. Hashimoto R, Hough C, Nakazawa T, Yamamoto T, Chuang DM. Lithium protection against glutamate excitotoxicity in rat cerebral cortical neurons: involvement of NMDA receptor inhibition possibly by decreasing NR2B tyrosine phosphorylation. J Neurochem 2002; 80: 589–597. 12. Cadotte DW, Xu B, Racine RJ et al. Chronic lithium treatment inhibits pilocarpine-induced mossy fiber sprouting in rat hippocampus. Neuropsychopharmacology 2003; 28: 1448–1453. 13. Shen CP, Tsimberg Y, Salvadore C, Meller E. Activation of Erk and JNK MAPK pathways by acute swim stress in rat brain regions. BMC Neurosci 2004; 5: 36. 14. Kitchener P, Di Blasi F, Borrelli E, Piazza PV. Differences between brain structures in nuclear translocation and DNA binding of the glucocorticoid receptor during stress and the circadian cycle. Eur J Neurosci 2004; 19: 1837– 1846.

367

Omata et al. 15. Kopnisky KL, Chalecka-Franaszek E, Gonzalez-Zulueta M, Chuang DM. Chronic lithium treatment antagonizes glutamate-induced decrease of phosphorylated CREB in neurons via reducing protein phosphatase 1 and increasing MEK activities. Neuroscience 2003; 116: 425–435. 16. Murata T, Omata N, Fujibayashi Y et al. Dynamic changes in glucose metabolism of living rat brain slices induced by hypoxia and neurotoxic chemical-loading revealed by positron autoradiography. J Neural Transm 1999; 106: 1075–1087. 17. Xu J, Culman J, Blume A, Brecht S, Gohlke P. Chronic treatment with a low dose of lithium protects the brain against ischemic injury by reducing apoptotic death. Stroke 2003; 34: 1287–1292. 18. Nonaka S, Chuang DM. Neuroprotective effects of chronic lithium on focal cerebral ischemia in rats. Neuroreport 1998; 9: 2081–2084. 19. Ren M, Senatorov VV, Chen RW, Chuang DM. Postinsult treatment with lithium reduces brain damage and facilitates neurological recovery in a rat ischemia ⁄ reperfusion model. Proc Natl Acad Sci U S A 2003; 100: 6210–6215. 20. Mora A, Sabio G, Gonzalez-Polo RA et al. Lithium inhibits caspase 3 activation and dephosphorylation of PKB and GSK3 induced by K+ deprivation in cerebellar granule cells. J Neurochem 2001; 78: 199–206. 21. Gjedde A. High- and low-affinity transport of d-glucose from blood to brain. J Neurochem 1981; 36: 1463–1471. 22. Patlak CS, Blasberg RG, Fenstermacher JD. Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data. J Cereb Blood Flow Metab 1983; 3: 1–7. 23. Sokoloff L, Reivich M, Kennedy C et al. The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 1977; 28: 897–916. 24. Newman GC, Hospod FE, Patlak CS. Brain slice glucose utilization. J Neurochem 1988; 51: 1783–1796. 25. Yang L, Zhang B, Toku K, Maeda N, Sakanaka M, Tanaka J. Improvement of the viability of cultured rat neurons by the non-essential amino acids L-serine and glycine that upregulates expression of the anti-apoptotic gene product Bcl-w. Neurosci Lett 2000; 295: 97–100. 26. Murata T, Omata N, Fujibayashi Y et al. Posthypoxic reoxygenation-induced neurotoxicity prevented by free radical scavenger and NMDA ⁄ non-NMDA antagonist in tandem as revealed by dynamic changes in glucose metabolism with positron autoradiography. Exp Neurol 2000; 164: 269–279. 27. Meldrum DR, Cleveland JC Jr, Rowland RT, Banerjee A, Harken AH, Meng X. Early and delayed preconditioning: differential mechanisms and additive protection. Am J Physiol 1997; 273: H725–H733. 28. Luo C, Xu H, Li XM. Post-stress changes in BDNF and Bcl-2 immunoreactivities in hippocampal neurons: effect of chronic administration of olanzapine. Brain Res 2004; 1025: 194–202.

368

29. Hashimoto R, Takei N, Shimazu K, Christ L, Lu B, Chuang DM. Lithium induces brain-derived neurotrophic factor and activates TrkB in rodent cortical neurons: an essential step for neuroprotection against glutamate excitotoxicity. Neuropharmacology 2002; 43: 1173–1179. 30. Copin JC, Reola LF, Chan TY, Li Y, Epstein CJ, Chan PH. Oxygen deprivation but not a combination of oxygen, glucose, and serum deprivation induces DNA degradation in mouse cortical neurons in vitro: attenuation by transgenic overexpression of CuZn-superoxide dismutase. JNeurotrauma 1996; 13: 233–244. 31. Wang JF, Chen B, Young LT. Identification of a novel lithium regulated gene in rat brain. Brain Res Mol Brain Res 1999; 70: 66–73. 32. Yamamoto KK, Gonzalez GA, Biggs WH III, Montminy MR. Phosphorylation-induced binding and transcriptional efficacy of nuclear factor CREB. Nature 1988; 334: 494–498. 33. Tao X, Finkbeiner S, Arnold DB, Shaywitz AJ, Greenberg ME. Ca2+ influx regulates BDNF transcription by a CREB family transcription factor-dependent mechanism. Neuron 1998; 20: 709–726. 34. Sharma HS. Neuroprotective effects of neurotrophins and melanocortins in spinal cord injury: an experimental study in the rat using pharmacological and morphological approaches. Ann N Y Acad Sci 2005; 1053: 407–421. 35. Beck T, Lindholm D, Castren E, Wree A. Brain-derived neurotrophic factor protects against ischemic cell damage in rat hippocampus. J Cereb Blood Flow Metab 1994; 14: 689–692. 36. Skup MH. BDNF and NT-3 widen the scope of neurotrophin activity: pharmacological implications. Acta Neurobiol Exp (Wars) 1994; 54: 81–94. 37. Meyer M, Matsuoka I, Wetmore C, Olson L, Thoenen H. Enhanced synthesis of brain-derived neurotrophic factor in the lesioned peripheral nerve: different mechanisms are responsible for the regulation of BDNF and NGF mRNA. J Cell Biol 1992; 119: 45–54. 38. Tong L, Perez-Polo R. Brain-derived neurotrophic factor (BDNF) protects cultured rat cerebellar granule neurons against glucose deprivation-induced apoptosis. J Neural Transm 1998; 105: 905–914. 39. Namiki J, Kojima A, Tator CH. Effect of brain-derived neurotrophic factor, nerve growth factor, and neurotrophin-3 on functional recovery and regeneration after spinal cord injury in adult rats. J Neurotrauma 2000; 17: 1219– 1231. 40. Videbech P. PET measurements of brain glucose metabolism and blood flow in major depressive disorder: a critical review. Acta Psychiatr Scand 2000; 101: 11–20. 41. Deckersbach T, Dougherty DD, Savage C et al. Impaired recruitment of the dorsolateral prefrontal cortex and hippocampus during encoding in bipolar disorder. Biol Psychiatry 2006; 59: 138–146. 42. Shirayama Y, Chen AC, Nakagawa S, Russell DS, Duman RS. Brain-derived neurotrophic factor produces antidepressant effects in behavioral models of depression. J Neurosci 2002; 22: 3251–3261.