Protective effect of curcumin on tumor necrosis factor-alpha-induced neuronal damage in the rat hippocampus A relationship to the inhibition of neuronal Ca2+ influx****★
Luyan Guo1, 2, Rongbo Tu1, 2, Min Lin3, Jun Dong1, 2

1Department of Pathophysiology, Key Laboratory of State Administration of Traditional Chinese Medicine, Medical College of Jinan University, Guangzhou  510632, Guangdong Province, China
2Institute of Brain Research, Medical College of Jinan University, Guangzhou  510632, Guangdong Province, China
3Department of Ultrasound, First Affiliated Hospital, Medical College of Jinan University, Guangzhou  510632, Guangdong Province, China

Luyan Guo★, Master, Department of Pathophysiology, Key Laboratory of State Administration of Traditional Chinese Medicine, Medical College of Jinan University, Guangzhou  510632, Guangdong Province, China; Institute of Brain Research, Medical College of Jinan University, Guangzhou  510632, Guangdong Province, China

Corresponding author: Jun Dong, Professor, Department of Pathophysiology, Key Laboratory of State Administration of Traditional Chinese Medicine, Medical College of Jinan University, Guangzhou  510632, Guangdong Province, China; Institute of Brain Research, Medical College of Jinan University, Guangzhou  510632, Guangdong Province, China
dongjunbox@163.com

Supported by: the grant from Medical Science Foundation of Guangdong Province, No. A2006334*; the Natural Science Foundation of Guangdong Province, No.06105246*, No.9151040701000008*; the Science and Technology Project of Guangzhou City, No.2007J1-C0041*

Abstract
BACKGROUND: Previous studies of curcumin have focused mainly on its cytotoxic properties for antitumor therapy. There are few studies addressing the application of curcumin in the prevention and treatment of nervous system diseases.
OBJECTIVE: To observe the protective effect of curcumin against tumor necrosis factor-alpha (TNF-α)-induced neuronal damage in the rat hippocampus and to explore the intervention effect of curcumin on Ca2+ influx following neuronal damage.
DESIGN, TIME AND SETTING: A cell morphological and physiological study was performed at the Institute of Brain Research, Medical College of Jinan University, China, from December 2006 to June 2007.
MATERIALS: Curcumin (Sigma, USA) and TNF-α (Sigma, USA) were used in this study.
METHODS: Hippocampal neurons were isolated from one-day neonatal rats and primarily cultured for 5 days. Following this they received 1 μmol/L curcumin and 100 ng/mL TNF-α pre-treatment. Dynamic morphological changes were observed for 1 hour by inverted microscopy. At 48 hours post-treatment, static morphological characteristics of the neurons were observed using inverted microscopy. Subsequently, hippocampal neurons were primarily cultured for 7 days, after receiving 1 μmol/L curcumin and 4.5 ng/mL TNF-α pre-treatment. Intracellular free Ca2+ was measured using Fluo 3/acetoxymethyl ester.
MAIN OUTCOME MEASURES: Effects of curcumin on TNF-α-induced neuronal damage and Ca2+ influx in the rat hippocampus were measured.
RESULTS: Following curcumin treatment, TNF-α-induced neurons grew as normal. TNF-α induced a rapid Ca2+ influx into the neuronal cytoplasm; however, Ca2+ fluorescence intensity only slightly increased when neurons were co-perfused with curcumin and TNF-α.
CONCLUSION: Curcumin has a protective effect on rat hippocampal neurons possibly by reducing the TNF-α-induced rapid Ca2+ influx into neuronal cytoplasm and by maintaining the Ca2+ homeo-stasis.
Key Words: curcumin; tumor necrosis factor-alpha; primary culture; Ca2+; human immunodeficiency virus type 1-associated dementia

INTRODUCTION
   
Human immunodeficiency virus type 1 (HIV-1)-associated dementia (HAD) is one of the most common and severe neurological complications of acquired immunodeficiency syndrome (AIDS). However, the viral load in the brains of the HAD patients is low, which does not explain the severe pathological changes[1]. HIV-1 and its viral proteins (gp120, gp41, Tat) can directly damage neurons[1]. However, the immune response and the inflammation induced by cytokines, such as tumor necrosis factor-α (TNF-α), interleukin-1β and interleukin-6, which are secreted from HIV-1 activated macrophages and microglia, are believed to play an important role in the development of HAD. Among the cytokines involved in the immune response, TNF-α is the most neurotoxic. It can cause increases in extracellular glutamate concentrations, resulting in the neuronal overexcitation and damage.
Many studies have encouraged investigations on the biological activity of curcumin, demonstrating a plethora of pharmacological properties, including anti-inflammatory[2-3] and anti-HIV-1[4]. However, the application of curcumin in  
HAD treatment has not been reported. Our previous study showed that curcumin could improve learning and memory ability in memory disorders induced by gp120 in rats[5]. The present study sought to observe the protective effect and mechanism of curcumin against TNF-α-induced neuronal damage in the rat hippocampus.

MATERIALS AND METHODS

Design
A cell morphological and physiological study.
Time and setting
Experiments were performed at the Institute of Brain Research, Medical College of Jinan University, China from December 2006 to June 2007.
Materials
Experimental animals
A total of 30 Sprague-Dawley rats, aged 1 day, were supplied by the Animal Experimental Center of Southern Medical University of China (license No. SCXK (yue) 2006-0015). Protocols were performed in accordance with the Guidance Suggestions for the Care and Use of Laboratory Animals, formulated by the Ministry of Science and Technology of the People’s Republic of China[6].
Major drug, reagents, and instrument
Curcumin (Sigma, USA), a phenolic compound isolated from the dry rhizome of curcumae longae, usually contains 2%–8% curcumin. The molecular weight of curcumin is 368.37; the chemical formula of which is C21H20O6[7] (Figure 1).

Reagents and instrument are listed as follows:

Methods
Primary cell cultures
Hippocampal neurons were cultured using previously described procedures[8] with some modifications. Briefly, neonatal Sprague-Dawley rats, aged 1 day, were sacrificed and sterilized using 60%–80% alcohol. The scalp and skull were incised to expose and obtain both cerebral hemispheres. The hippocampus was washed in phosphate buffered saline (PBS) (4 °C, 0.01 mol/L) and the blood vessels were removed with straight iris forceps. Fresh hippocampal tissues were then placed in a 10 mL centrifuge tube, containing 6 mL culture media (90% DMEM/F12+10% fetal calf serum), and sectioned into approximately 1-mm slices with a forceps. Enzymatic treatment was not used. The sectioned tissues were dispersed mechanically by blowing lightly with fire-polished pipettes of large- and small-diameters. The solution was then precipitated for 2–3 minutes. The supernatant was obtained and filtrated into a new 10 mL centrifuge tube using a stainless steel screen cloth. The precipitation and filtration steps were then repeated once. Cells were seeded at 2 × 105 cells/cm2 on poly-L-lysine (0.1 mg/mL)- coated Petri dishes, supplemented with 90% DMEM and 10% fetal calf serum. Cells were incubated in a saturated humidified 5% CO2 and 95% air at 37 °C. After 24 hours in vitro, the culture medium was fully replaced with fresh media. Cytosine arabinoside    (10 μmol/L) was added to the dishes to inhibit the growth of non-neuronal cells on day 3 and the media was replaced fully after another 48 hours. Subsequently, half the medium was replaced twice a week. Cultures were monitored to ensure that neurons constituted ≥95% of the total population. All experiments were performed using rat hippocampal neurons cultured for 5 and 7 days.
Morphological characteristics of hippocampal neurons under inverted microscopy
Experimental grouping: After hippocampal neurons were cultured for 5 days, the experiment was randomly divided into four groups: control, TNF-α, curcumin, and curcumin-TNF-α groups; with five Petri dishes in each group. Prior to drug treatment, each dish was filled with 1 mL 90% DMEM and 10% fetal calf serum to replace the extracellular fluid. The control group did not receive drug treatment. The TNF-α group received 10 μg/mL TNF-α (10 μL, final concentration: 100 ng/mL); the curcumin group received 1 mmol/L curcumin (1 μL, final concentration: 1 μmol/L); and the curcumin-TNF-α group received 1 mmol/L curcumin (1 μL, final concentration:  1 μmol/L), immediately followed by 10 μg/mL TNF-α  (10 μL, final concentration: 100 ng/mL).
Following drug treatment, Petri dishes were immediately placed onto the microstat of the inverted microscope; neurons were selected with the 20-fold objective lens and held stable. A series of neurons were photographed for 1 hour using Olympus Image-Pro Plus 6.0 software. The time interval between pictures was set at 2.5 minutes. Finally, the photo series of neurons was analyzed and exported with Olympus Image-Pro Plus 6.0 software.  
At 48 hours post-drug treatment, single photos were used to observe the static morphological characteristics of neurons from four groups under the inverted microscope.
Calcium ion fluorescence imaging
Experimental grouping: after culturing hippocampal neurons for 7 days, the experiment was divided into four groups: control, TNF-α, curcumin, and curcumin-TNF-α group, with five Petri dishes in each group. Prior to drug treatment, each dish was filled with 1 mL HEPES-buffered Krebs-Ringer solution to replace the extracellular fluid. Cells were selected under a 20-fold objective lens, where the Ca2+ fluorescence intensity remained stable for 60 seconds, following which drugs were added into HEPES-buffered Krebs-Ringer solution. The control group received 10 μL PBS; the TNF-α group received 1 μg/mL TNF-α (4.5 μL, final concentration:  4.5 ng/mL); the curcumin group received 1 mmol/L curcumin (1 μL, final concentration: 1 μmol/L); and finally the curcumin-TNF-α group received 1 mmol/L curcumin (1 μL, final concentration: 1 μmol/L), immediately followed by 1 μg/mL TNF-α (4.5 μL, final concentration: 4.5 ng/mL).
Single cell fluo 3 fluorescence imaging: Intracellular free Ca2+ level was indicated by single cell fluo 3 fluorescence imaging, with slight modifications on a previously described method[9]. Briefly, cells were washed twice with Mg2+-free D-Hank’s solution, and then incubated at 37 °C, 5% CO2, away from light for  30 minutes in the KRH’s solution supplemented with  10 μmol/L Fluo 3/acetoxymethyl ester (Fluo 3/AM). This was then followed by washing twice again with Mg2+ free D-Hank’s solution to remove extracellular superfluous Fluo 3/AM.
Calcium imaging was recorded by a confocal laser scanning microscope. Successive images of 512 × 512 pixels were collected with a convert 20 times, 0.5 number aperture oil objective lens. The scanning speed was adjusted to 1 frame/second (fps) by Timeseries program. Intracellular Ca2+ level ([Ca2+]i) was represented by the fluo 3 fluorescence intensity. No dye bleaching was observed during the testing time, to normalize to the baseline fluorescence intensity. Ca2+ fluorescence intensity was recorded for 60 seconds prior to drug treatment and 250 seconds after that. Quantification of the fluorescence intensity was performed using Zeiss AIM 4.0 software.
Main outcome measures
Effects of curcumin on TNF-α-induced neuronal damage in the rat hippocampus and on Ca2+ influx.
Statistical analysis
AIM 4.0 and SPSS 13.0 were used to analyze Ca2+ fluorescence intensity-time curve and measure the increased fluorescence intensity. The maximum of increased Ca2+ fluorescence intensity, which was expressed by Mean ± SD, in TNF-α group and TNF-α plus curcumin group was analyzed by two independent sample t test. A value of P < 0.05 was considered statistically significant (n1 = n2 = 5).

RESULTS

Effect of curcumin on structural damage of neurons induced by TNF-α
Neurons were photographed at five time points of 0, 15, 30, 45, and 60 minutes to form a photo series. By analyzing and comparing the photo series, we found that the cell bodies of the neurons appeared normal without any obvious changes. The axons and secondary branches extended out from cell bodies in the control and curcumin groups. However, after TNF-α treatment, the cell bodies of neurons swelled, axons and secondary branches gradually retracted, eventually disappearing in the TNF-α group. However, cell bodies of neurons remained normal in the curcumin-TNF-α group and axons and secondary branches extended as usual (Figure 2).

At 48 hours after drug treatment, the inverted microscope showed that neurons from the control and curcumin groups were growing closer together. Their cell bodies were comparatively large and plentiful, pyramidal or elliptical shaped, with conspicuous halos surrounding them. There were many axons, dendrites and secondary branches stretching out from the cell bodies which formed a network structure (Figures 3A, B). However, neurons from the TNF-α group grew sparsely. Their cell bodies were small and spindle shaped, and there were no conspicuous halos. There were rare axons, dendrites and secondary branches growing from the cell bodies which did not form a network structure (Figure 3C). However, neurons from the curcumin-TNF-α group were not significantly different from the control group. Cell bodies recovered back to the normal large and plump, pyramidal and elliptical shapes, with conspicuous halos surrounding them. Many axons, dendrites and secondary branches stretched out from the cell bodies and expanded to form a network structure (Figure 3D).

Effect of curcumin on TNF-α-induced Ca2+ influx
After neurons were loaded with Fluo 3/AM, Ca2+ in the cytoplasm was marked green by the fluorescent probe using laser confocal microscopy. In the control and curcumin groups, Ca2+ could be seen uniformly distributed in the neuronal cell bodies, and following the addition of PBS or curcumin, the fluorescence did not increase (Figures 4A, B). In the TNF-α group, there was an obvious increase in fluorescence intensity in the neuronal cell bodies, stems and distal small branches of neurites. On the fluorescence intensity-time curve, we can see that the curve increased quickly after the addition of TNF-α. This reached a peak 15 seconds later and gradually returned to stable state another 75 seconds later. Nevertheless, the stable status was still higher than the basal intensity. The maximal Ca2+ fluorescence intensity increased by 53.40 ± 1.58 (a.u.) over the basal intensity (Figure 4C). However, in the curcumin-TNF-α group, there was a slight increase in the fluorescence intensity, mainly in the neuronal cell body. The intensity of the neurites did not increase obviously. The fluorescence intensity-time curve increased slowly after the addition of curcumin and TNF-α, reached a peak after 15 seconds and gradually decreased to a stable state after another 75 seconds. The stable state was the same as the basal level. The maximal Ca2+ fluorescence intensity increased by 25.36 ± 0.76 (a.u.) over the basal level and had significant difference compared with the TNF-α group (P < 0.05, n1 = n2 = 5) (Figure 4D).

DISCUSSION

TNF-α is one of the main cytokines involved in the HIV-1 invasion of the central nervous system (CNS) and the immune response of HAD. In vitro, the exposure of macrophages and microglia to either gp120 or Tat results in the up-regulation of TNF-α expression[10-12]. By increasing the permeability of the blood-brain barrier, TNF-α opens a paracellular route for HIV-1 invasion of the CNS[13]. Furthermore, TNF-α induces the expression of the adhesion molecules, intercellular adhesion molecule-1, vascular cell adhesion molecule-1 and E-selectin, on astrocytes and endothelial cells, which allows HIV-1-infected monocytes/macrophages to transmigrate into the CNS[14-15]. In addition, TNF-α up-regulates the expression and release of various chemokines in the CNS, such as monocyte chemoattractant protein-1, a potent chemoattractant for monocytes and macrophages[16]. Therefore, by increasing the permeability of the blood-brain barrier and inducing adhesion molecule and chemokine expression, TNF-α plays an important role in the entry of HIV-infected cells into the brain. It has also been demonstrated that TNF-α is toxic to human neurons in vitro. TNF-α induces its neurotoxic effects through the over-stimulation of glutamate receptors. Ca2+ channels on the cell membrane remain open[17] causing excessive Ca2+ flow into the cytoplasm of neurons, ultimately leading to neuronal death by apoptosis or necrosis.
Curcumin has been shown to be a potent scavenger of a variety of reactive oxygen species, including superoxide anion radicals, hydroxyl radicals and nitrogen dioxide radicals[18]. Curcumin prevents the induction of cyclooxygenase-2 and inducible nitric oxide synthase to decrease oxidative damage, inhibiting the production of inflammatory factors, such as TNF-α, interleukin-1β and interleukin-8. Another critical characteristic of curcumin is its capacity for crossing the blood-brain barrier to provide direct neuroprotection[19]. The present study confirmed that 100 ng/mL TNF-α can directly inhibit the growth of primarily cultured rat hippocampal neurons in vitro. In the 1-hour dynamic observation experiments using an inverted microscope, cell bodies of neurons swelled, axons and secondary branches gradually contracted and finally disappeared, as demonstrated by the photo series. At 48 hours after TNF-α treatment, neurons were observed by static photography, showing sparse growth and decreased neuronal cell bodies. There were few axons and secondary branches stretching out from the cell bodies and did not form a network structure. After pre-treating with 1 μmol/L curcumin, neurons grew as normal and cell bodies of neurons were comparatively large and plentiful. There were a lot of axons and secondary branches stretching out from cell bodies which formed a network structure. These indicated that curcumin protects neurons from growth inhibition induced by TNF-α. Pre-treatment with curcumin promotes healthy neuronal growth. During Ca2+ fluorescence imaging, we found that TNF-α induced a rapid Ca2+ influx into the neuronal cytoplasm. However, Ca2+ fluorescence intensity only increased slightly when neurons were co-perfused with curcumin and TNF-α and there was only a low peak in the fluorescence intensity-time curve. The results demonstrated that curcumin reduced the TNF-α-induced Ca2+ influx and maintained the Ca2+ homeostasis in neuronal cytoplasm. Overall, curcumin has a protective effect on hippocampal neurons of rats, possibly by reducing the rapid Ca2+ influx into neuronal cytoplasm induced by TNF-α and by maintaining Ca2+ homeostasis. However, we only investigated one aspect of the complicated mechanisms underlying protective effects of curcumin on hippocampal neurons in rats and the complete set of mechanisms requires further study.

REFERENCES

[1] Glass JD, Fedor H, Wesselingh SL, et al. Immunocytochemical quantitation of human immunodeficiency virus in the brain: correlations with dementia. Ann Neurol. 1995;38(5):755-762.
[2] Abe Y, Hashimoto S, Horie T. Curcumin inhibition of inflammatory cytokine production by human peripheral blood monocytes and alveolar macrophages. Pharmacol Res. 1999; 39(1):41-47.
[3] Aggarwal BB, Harikumar KB. Potential therapeutic effects of curcumin, the anti-inflammatory agent, against neurodegenerative, cardiovascular, pulmonary, metabolic, autoimmune and neoplastic diseases. Int J Biochem Cell Biol. 2009;41(1):40-59.
[4] Li CJ, Zhang LJ, Dezube BJ, et al. Three inhibitors of type 1 human immunodeficiency virus long terminal repeat-directed gene expression and virus replication. Proc Natl Acad Sci U S A. 1993;90(5):1839-1842.
[5] Dong J, Lu DX, Pan R, et al. Effect and mechanism of curcumin on learning and memory dysfunction induced by gp120 in rats. Xi Bao Yu Fen Zi Mian Yi Xue Za Zhi. 2008;24(4):328-331.
[6] The Ministry of Science and Technology of the People’s Republic of China. Guidance Suggestions for the Care and Use of Laboratory Animals. 2006-09-30.
[7] Chinese Medicinal Herb Information Centre of State Pharmaceutical Administration. Handbook of Active Ingredients of Phytomedicines. Beijing: People’s Medical Publishing House. 1984:281-282.
[8] Liu BF, Ma J, Gao EJ, et al. Development of an artificial neuronal network with post-mitotic rat fetal hippocampal cells by polyethylenimine. Biosens Bioelectron. 2008;23(8):1221-1228.
[9] Zhao P, Huang YL, Cheng JS. Taurine antagonizes calcium overload induced by glutamate or chemical hypoxia in cultured rat hippocampal neurons. Neurosci Lett. 1999;268(1):25-28.
[10] Nicolini A, Ajmone-Cat MA, Bernardo A, et al. Human immunodeficiency virus type-1 Tat protein induces nuclear factor (NF)-kappaB activation and oxidative stress in microglial cultures by independent mechanisms. J Neurochem. 2001;79(3):713-716.
[11] Lee C, Tomkowicz B, Freedman BD, et al. HIV-1 gp120-induced TNF-alpha production by primary human macrophages is mediated by phosphatidylinositol-3 (PI-3) kinase and mitogen- activated protein (MAP) kinase pathways. J Leukoc Biol. 2005; 78(4):1016-1023.
[12] Buscemi L, Ramonet D, Geiger JD. Human immunodeficiency virus type-1 protein Tat induces tumor necrosis factor-alpha- mediated neurotoxicity. Neurobiol Dis. 2007;26(3):661-670.
[13] Fiala M, Looney DJ, Stins M, et al. TNF-alpha opens a paracellular route for HIV-1 invasion across the blood-brain barrier. Mol Med. 1997;3(8):553-564.
[14] Winkler MK, Beveniste EN. Transforming growth factor-beta inhibition of cytokine-induced vascular cell adhesion molecule-1 expression in human astrocytes. Glia. 1998;22(2):171-179.
[15] Konsman JP, Drukarch B, Van Dam AM. (Peri)vascular production and action of pro-inflammatory cytokines in brain pathology. Clin Sci (Lond). 2007;112(1):1-25.
[16] Ding T, Sun J, Zhang P. Study on MCP-1 related to inflammation induced by biomaterials. Biomed Mater. 2009;4(3):35005.
[17] Lipton SA. Neuronal injury associated with HIV-1 and potential treatment with calcium-channel and NMDA antagonists. Dev Neurosci. 1994;16(3-4):145-151.
[18] Gu JL, Yang KY, Yang JY, et al. Effect of curcumin on nitric oxide synthase expression in Iipopolysaccharide-activated microglia cells and the anti-oxidative effect of curcumin. Neural Regen Res. 2007;2(12):709-712.
[19] Yang F, Lim GP, Begum AN, et al. Curcumin inhibits formation of amyloid beta oligomers and fibrils, binds plaques, and reduces amyloid in vivo. J Biol Chem. 2005;280(7):5892-5901.
 (Edited by Tu QY/Qiu Y/Song LP)