Neurochemical Research

, Volume 36, Issue 1, pp 117–128

The Chronological Characteristics of SOD1 Activity and Inflammatory Response in the Hippocampi of STZ-Induced Type 1 Diabetic Rats


  • Sun Shin Yi
    • Department of Anatomy and Cell Biology, College of Veterinary Medicine and Research Institute for Veterinary ScienceSeoul National University
    • Department of Biomedical Sciences, College of Health SciencesMarquette University
  • In Koo Hwang
    • Department of Anatomy and Cell Biology, College of Veterinary Medicine and Research Institute for Veterinary ScienceSeoul National University
  • Dae Won Kim
    • Department of Biomedical Sciences, Division of Life SciencesHallym University
  • Jae Hoon Shin
    • Department of Anatomy and Cell Biology, College of Veterinary Medicine and Research Institute for Veterinary ScienceSeoul National University
  • Sung Min Nam
    • Department of Anatomy and Cell Biology, College of Veterinary Medicine and Research Institute for Veterinary ScienceSeoul National University
  • Jung Hoon Choi
    • Department of Anatomy, College of Veterinary MedicineKangwon National University
  • Choong Hyun Lee
    • Department of Neurobiology, School of MedicineKangwon National University
  • Moo-Ho Won
    • Department of Neurobiology, School of MedicineKangwon National University
  • Je Kyung Seong
    • Department of Anatomy and Cell Biology, College of Veterinary Medicine and Research Institute for Veterinary ScienceSeoul National University
    • Department of Anatomy and Cell Biology, College of Veterinary Medicine and Research Institute for Veterinary ScienceSeoul National University

DOI: 10.1007/s11064-010-0280-6

Cite this article as:
Yi, S.S., Hwang, I.K., Kim, D.W. et al. Neurochem Res (2011) 36: 117. doi:10.1007/s11064-010-0280-6


Because it appears that oxidative stress and inflammation are implicated with disease pathogenesis in the diabetic brain, many researchers have used streptozotocin (STZ)-induced diabetic animals to study superoxide production and the effects of superoxide scavengers like Cu,Zn-superoxide dismutase (SOD1). However, many studies have been conducted without considering temporal changes after STZ injection. Interestingly, though SOD activities were not significantly different among the groups, SOD1 and 4-hydroxy-2-nonenal (4-HNE) immunoreactivities were significantly enhanced at 3 weeks after an STZ injection (STZ3w) versus only marginal levels in sham controls, whereas microglial activity was remarkably reduced in injected rats at this time. However, SOD1 immunoreactivity and microglial activities were only at the sham level at STZ4w. The present study provides important information concerning cell damage by ROS generated by STZ. Microglial response was found to be inactivated at STZ3w and neuronal cells (NeuN) showed a non-significant tendency to be reduced in number at STZ4w except in the dentate gyrus. We speculated that the above oxidative stress-related events should be accomplished at STZ3w in the brains of STZ-induced diabetes animal models. Therefore, the aim of the present study was to investigate chronological changes in SOD1 immunoreactivity associated with lipid peroxidation and inflammatory responses in the hippocampi of STZ-induced type I diabetic rats.


Oxidative stressDiabetesSOD14-HNEMicroglial activityNeuronal cells


Diabetes mellitus is a complex disease associated with systemic and neural abnormalities [13]. In particular, diabetes mellitus, if left untreated, may initiate degenerative processes in the central nervous system [4, 5] and peripheral organs [2, 6], due to excessive reactive oxygen species (ROS) production [7]. ROS can initiate a series of events in tissues that might ultimately lead to the development of many complications. In diabetes mellitus all tissues are affected by ROS, but greatest effect is observed in the brain [3]. Thus, chronic hyperglycemia causes an imbalance in the oxidative status of nervous tissue, and the resulting free radicals damage brain tissues, such as, the hippocampus, hypothalamus, and cortex, via Fenton-like reactions [8]. Furthermore, accumulating evidence suggests that ROS play pivotal roles in diabetes and that the formation of free radicals induced by hyperglycemia plays a principal role in the pathogenesis of diabetic neuropathy [9]. Moreover, these radicals can increase the expressions of tumor necrosis factor-α (TNF-α) and of nuclear transcription factor [10].

Various antioxidant mechanisms in brain neutralize the harmful effects of the free radical excesses associated with diabetes [3]. Among these, Cu,Zn-superoxide dismutase (SOD1) detoxifies intracellular free radicals, and thus, protects them from oxidative damage [11]. SOD1 is expressed in the cytoplasm and nuclei of all eukaryotic cells, where it acts as a bulk scavenger of superoxide [12, 13], and its expression has been studied in streptozotocin (STZ)-induced type I diabetes models. However, although Kawamura et al. found that the percentage of glycated SOD1 in diabetic patients was significantly higher than in controls [14], results concerning SOD1 expression in diabetes have been conflicting [3, 15].

The aim of the present study was to investigate chronological changes in SOD1 immunoreactivity associated with lipid peroxidation and inflammatory responses in the hippocampi of streptozotocin (STZ)-induced type I diabetic rats.

Experimental Procedure

Animals and Experimental Design

Experiments were performed on adult Wistar male rats (n = 45; b.w. 190 and 240 g; 8 weeks old). Animals were housed at room humidity 60% and 23°C under a 12-h light/12-h dark cycle with free access to food and water. A Purina 5008 rodent diet (7.5% fat) was provided as recommended by the manufacturer (Ralston Purina, St. Louis, MO). The procedures used for the handling and care of animals were in accord with currently accepted guidelines (NIH Guide for the Care and Use of Laboratory Animals, NIH Publication No. 85-23, 1985, revised 1996), and were approved by the Institutional Animal Care and Use Committee (IACUC) at Seoul National University, College of Veterinary Medicine (approval no. SNU-080128-6). All experiments were conducted so as to minimize the number of animals used and suffering caused by the procedures used.

Diabetes mellitus was induced by a single intraperitoneal (i.p.) injection of freshly prepared 70 mg/kg/5 ml streptozotocin (STZ) (Sigma–Aldrich, USA) in 0.1 M sodium citrate buffer (pH 4.3). After 72 h, fasting blood glucose levels were monitored and rats with blood glucose levels of >145 mg/dl were utilized for the study (n = 40). Sham controls were injected i.p. with the same volume of 0.1 M sodium citrate buffer (pH 4.3).

Physiological Data

Blood was collected on a weekly basis from tail veins and blood glucose levels were measured using a validated one-touch basic glucose measurement system (SureStepTM blood glucose meter, Lifescan, CA, USA). Body weights were also measured weekly from baseline. Daily food intake and water consumptions were recorded for 72 h in Sham controls and STZ-treated groups.

Plasma Corticosterone Levels

Plasma (50 μl) was added to 5 ml of methylene chloride and incubated at room temperature for 10 min. After filtration through a cheesecloth, the mixture was combined with 2.5 ml of fluorescence reagent (7:3, sulfuric acid/absolute ethanol), vortexed vigorously, and incubated for 30 min at room temperature. After centrifugation, lower layer absorbance was measured using a spectrophotometer (UV-1601, Shimadzu Co., Japan; excitation wavelength, 475 nm; emission wavelength, 530 nm).

Antioxidant (SOD1) Activity

SOD1 activity was measured by determining ability of samples to inhibit the reduction of ferricytochrome c by xanthine/xanthine oxidase, as described by McCord and Fridovich [15]. Cu,Zn-SOD was separated by electrophoresis in 10% native polyacrylamide gels and visualized as described by Beauchamp and Fridovich [16]. Briefly, gels were soaked in 2.45 mM nitroblue tetrazolium solution for 15 min, and then in 0.36 mM potassium phosphate buffer containing 28 mM N,N,N′,N′-tetramethylethylene diamine and 28 μM riboflavin at pH 7.8 for 30 min. Gels were then exposed to a fluorescence light source until maximum band resolution was achieved.

Tissue Processing for Histology

For histological analyses, five animals at STZ2, 3, and 4w were anesthetized with sodium pentobarbital and perfused transcardially with 0.1 M phosphate-buffered saline (PBS, pH 7.4) followed by 4% paraformaldehyde in 0.1 M phosphate-buffer (PB, pH 7.4). Brains were then removed and postfixed in the same fixative for 6 h. Brain tissues so obtained were then cryoprotected by infiltration with 30% sucrose overnight, and frozen tissues were serially sectioned (30-μm) in the coronal plane using a cryostat (Leica, Wetzlar, Germany). Sections were then placed into six-well plates containing PBS.

Immunohistochemistry for SOD1, 4-hydroxy-2-nonenal (4-HNE), ionized calcium binding adaptor molecule-1 (Iba-1), and neuronal specific nuclear protein (NeuN).

Freshly prepared sections were sequentially treated with 0.3% hydrogen peroxide (H2O2) in PBS for 30 min and then with 10% normal goat serum or normal rabbit serum in 0.05 M PBS for 30 min. They were then incubated with diluted sheep anti-SOD1 antibody (1:1,000, Calbiochem, San Diego, CA), diluted mouse anti-4-HNE (1:1,000, Alexis Biochemicals, San Diego, CA), diluted rabbit anti-Iba-1 (1:500, Wako, Osaka, Japan), or diluted mouse anti-NeuN (1:1,000, Chemicon, Temecula, CA) overnight at 4°C, and exposed to biotinylated rabbit anti-sheep IgG, goat anti-rabbit, or goat anti-mouse IgG and streptavidin peroxidase complex (1:200, Vector, Burlingame, CA). Sections were then visualized using 3, 3′-diaminobenzidine tetrachloride (Sigma, St Louis, MO) in 0.1 M Tris–HCl buffer and mounted on gelatin-coated slides. After dehydration, they were mounted using Canada balsam (Sigma, St Louis, MO). Negative control tests were carried out using pre-immune serum instead of primary antibodies in order to establish the specificity of immunostaining. Negative controls showed no immunoreactivity whatsoever.

Western Blot Analysis for SOD1, IFN-γ, and IL-1β

To confirm changes in SOD1 levels in hippocampal regions at each time point, 5 animals were sacrificed at each time point for Western blot analysis. Brains were removed and hippocampal regions were then dissected with a surgical blade. Tissues were homogenized in 50 mM PBS (pH 7.4) containing 1 mM EGTA (pH 8.0), 0.2% NP-40, 10 mM EDTA (pH 8.0), 15 mM sodium pyrophosphate, 100 mM β-glycerophosphate, 50 mM NaF, 150 mM NaCl, 2 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 1 mM dithiothreitol (DTT). After centrifugation, protein levels were determined in supernatants using Micro BCA protein assay kits and bovine serum albumin as the standard (Pierce Chemical, USA). Briefly, aliquots containing 50 μg of total protein were boiled in loading buffer containing 150 mM Tris (pH6.8), 3 mM DTT, 6% SDS, 0.3% bromophenol blue, and 30% glycerol. After electrophoresis, gels were transferred to nitrocellulose membranes (Pall Crop, East Hills, NY, USA). To reduce background staining, membranes were incubated with 5% non-fat dry milk in PBS containing 0.1% Tween 20 for 45 min, and then incubated with diluted sheep anti-SOD1 antibody (1:1,000; Calbiochem, San Diego, CA), goat anti-IFN-γ (1:500, Santa Cruz Biotechnology, Santa Cruz, CA) and mouse anti-IL-1β (Santa Cruz Biotechnology), peroxidase-conjugated donkey anti-sheep (Sigma, St Louis, MO), rabbit anti-goat IgG (Sigma, St Louis, MO), or goat anti-mouse IgG (Sigma). Blots were developed using an ECL kit (Pierce Chemical, Rockford, IL).

Data Analysis

All measurements were performed under blind conditions by two observers per experiment using the same conditions.

In order to analyze quantitatively the immunoreactivities of SOD1, 4-HNE, Iba-1 and NeuN, hippocampal regions were measured in 10 sections per animal. Images of all NeuN immunoreactive structures were taken using a BX51 light microscope (Olympus, Japan) equipped with a digital camera (DP71, Olympus) connected to a PC monitor. NeuN-positive cells were counted. The staining intensities of all 4-HNE and SOD1 immunoreactive structures were evaluated based on their optical densities (OD), which were obtained by transforming mean gray levels using the formula: OD = log (256/mean gray level). The ODs of backgrounds were subtracted from those of areas adjacent to measured areas. OD ratios were then expressed as percentages (relative optical densities, ROD) using Adobe Photoshop version 8.0 and then analyzed using NIH Image 1.59 software. Western blots were scanned and quantified using Scion Image Software (Scion Corp., Frederick, MD).
Fig. 1

Body weights and blood glucose levels of STZ-treated rats. Body weights of the sham controls increased exponentially, but those of STZ-treated rats reduced slowly (a).The blood glucose levels of rats at 1-, 2-, 3-, 4-weeks post-STZ (STZ1, 2, 3, 4w) were significantly different from those of sham controls. Blood glucose levels were continuously higher than 380 mg/dl from STZ1w (b). Food intakes and water consumptions are also obviously higher from 1-week post-STZ (c, d). (P < 0.001)

The data shown represent means ± SEs (Standard Errors). Differences between means were analyzed for significance by one-way analysis of variance followed by Bonferroni post test and Duncan’s new multiple methods to determine differences between experimental groups. Statistical significance was accepted for P values of < 0.05.


Physiological Data

Mean body weight of STZ-treated rats fell over the 4-week assay period whereas that of sham controls increased (Fig. 1a). STZ rats had blood glucose levels that were about 4-times greater than those of controls (Fig. 1b), and they consumed more food (Fig. 1c) and water (Fig. 1d) than controls.

Plasma Corticosterone Levels

Plasma corticosterone levels values at STZ2w and STZ4w were significantly higher than those of sham controls, but levels at STZ3w were not significantly different (Fig. 2).
Fig. 2

Plasma corticosterone levels in STZ-treated rats. Plasma corticosterone levels at 2-, 4-weeks post-STZ (STZ2, 4w) were significantly different from those of sham controls (each n = 5, § P < 0.001, P < 0.05). Plasma corticosterone levels were reduced at STZ3w and at this time were not significantly different from that of sham control. Bars indicate mean ± SE

SOD1 Activity

SOD activities in brain tissues (n = 5) are presented in Fig. 3. SOD activity levels were similar in the two groups (P > 0.05).
Fig. 3

Antioxidant enzyme (SOD) activities in brain tissues of all groups. SOD values are indicated as Unit/mg of protein in brain. Values are expressed as means ± SEs (n = 5). Statistical analysis was performed using one-way ANOVA or the Bonferroni test (n = 5 per group). Bars indicate means ± SEs

Immunohistochemistry for SOD1, 4-HNE, Iba-1, and NeuN

SOD1, 4-hydroxy-2-nonenal (4-HNE), Iba-1, and NeuN immunoreactivities were detected in the hippocampus (CA1, CA3, Dentate gyrus). However, SOD1 was rarely detected in any area of the hippocampus until STZ2w, and it was highly expressed in CA1 (P < 0.001), CA3 (P < 0.05), and DG (P < 0.005) at STZ3w versus sham controls (Fig. 4). However, these immunoreactivities were rarely detected in CA1 or CA3 (except DG) at STZ4w (P < 0.01). SOD1 immunoreactivity was detected in the DG at STZ4w, but its immunoreactivity was 25% lower than at STZ3w.
Fig. 4

Immunoreactivities of SOD1 in the CA1, CA3, and DG hippocampal regions of sham controls and STZ treated rats at STZ2, 3, and 4w and their relative optical densities (ROD) in hippocampus expressed as percentages of SOD1 immunoreactivity in sham controls. Differences between means were analyzed using the Bonferroni test (n = 5 per group, § P < 0.001, **P < 0.005,  P < 0.05, significantly different from the sham controls), SO stratum oriens, SP stratum pyramidale, SR stratum radiatum, ML molecular layer, GCL granule cell layer, PoL polymorphic layer. Bars indicate mean ± SE. Bar = 100 μm

4-HNE was rarely detected any hippocampal area until STZ2w, but it was highly expressed in CA1 (P < 0.005), CA3 (P < 0.001), and DG (P < 0.005) at STZ3w as compared with sham controls (Fig. 5). However, 4-HNE immunoreactivities were drastically decreased in all hippocampal areas at STZ4w as compared with STZ3w. Nevertheless, its immunoreactivities in CA3 and DG at STZ4w were significantly greater than in sham controls (P < 0.05).
Fig. 5

Immunoreactivities of 4-HNE in the hippocampus (CA1, CA3, and DG) of sham and STZ treated rats at STZ2, 3, and 4w. Relative optical densities (ROD) in these hippocampal regions are quoted as percentages of 4-HNE immunoreactivity in sham controls. Differences between means were analyzed using the Bonferroni test (n = 5 per group, § P < 0.001, ** P < 0.005, P < 0.05 vs. the sham controls), SO stratum oriens, SP stratum pyramidale, SR stratum radiatum, ML molecular layer, GCL granule cell layer, PoL polymorphic layer. Bars indicate mean ± SE. Bar = 100 μm

Iba-1 immunoreactivities were elevated in CA1, CA3, and DG at STZ2w, but were markedly lower at STZ3w (Fig. 6). Highly magnified images of microglia are shown in the squares on the top right of Fig. 6. More ramified microglia were observed at STZ3w than in the other groups. In particular, microglia had round and swollen cell bodies and were more highly activated in at STZ2w and STZ4w than in at STZ3w (Fig. 6). We counted all microglia around hippocampus in the black rectangle (1,800 × 1,000 μm), as shown in Fig. 6, in count extent. After counting microglia, we calculated the ratio of activated/inactivated microglia and plotted results as a bar graph (Fig. 6). The STZ2w ratio showed a significantly high activity (P < 0.001), but STZ3w (P = 0.051) and STZ4w (P = 0.059) ratios were not statistically significant. However, the values may be very meaningful. Microglia showed resting at STZ3w and activity again at STZ4w compared with that of Sham.
Fig. 6

Immunohistochemical staining of Iba-1 in the CA1, CA3 and DG hippocampal regions of sham controls and STZ2, 3, and 4w rats. Microglia in STZ-treated rats were highly activated and had round, swollen cell bodies at STZ2w and STZ4w, but not at STZ3w. Microglia were inactivate in the STZ3w group as shown by the top right hand square. Ratio of microglial activities (active/inactive microglia) are presented as a bar graph. The area of the hippocampus used in calculation is indicated by the black rectangle (1,800 × 1,000 μm) at ‘Microglia count extent’. The thick lines represent averages and the thin lines represent median values. Error bars indicate maximum and minimum values, respectively. Differences between means were analyzed using the Bonferroni test (n = 5 per group, § P < 0.001 vs. the sham controls), SO stratum oriens, SP stratum pyramidale, SR stratum radiatum, ML molecular layer, GCL granule cell layer, PoL polymorphic layer. The bar represents 100 μm

Numbers of NeuN immunoreactive neurons were similar in sham and STZ treated animals, and the only significant difference between STZ and sham treated rats was for DG immunoreactivity at STZ4w (P < 0.05) (Fig. 7).
Fig. 7

Immunohistochemical staining of NeuN in the CA1, CA3, and DG hippocampal regions in sham controls and in STZ-treated rats at STZ2, 3, and 4w. NeuN positive cells numbers are presented as bar graphs. Differences between means were analyzed using the Bonferroni test (n = 5 per group, P < 0.05, significantly different from the sham controls), SO stratum oriens, SP stratum pyramidale, SR stratum radiatum, ML molecular layer, GCL granule cell layer, PoL polymorphic layer. Bars indicate mean ± SE. Bar = 100 μm

Changes in SOD1, IFN-γ and IL-1β Protein Levels

Western blot results for SOD1 protein in the hippocampus paralleled its observed immunohistochemical changes (Fig. 8a). SOD1 protein levels significantly increased at STZ3w (1.42 times higher than in the sham control; P < 0.01), and even though its expression level was lower at STZ4w, this was still significantly greater than in the sham controls (1.31 times higher; P < 0.05). However, Western blot results for IFN-γ and IL-1β levels showed that both were significantly increased at STZ3w and STZ4w compared with those of sham controls and STZ2w (P < 0.05) (Fig. 8b). Both pro-inflammatory cytokines (IFN-γ, IL-1β) at STZ4w were significantly higher in STZ-treated rats than in Sham controls in hippocampus STZ2w and STZ3w (P < 0.01) (Fig. 8b).
Fig. 8

Western Blot analysis of (A) SOD1 and (B) pro-inflammatory cytokines (IFN-γ and IL-1β) in the hippocampus region of sham controls and STZ-treated rats at STZ2w, STZ3w, and STZ4w. Relative optical densities (ROD) are quoted as percentages versus SOD1 protein in the sham controls (n = 3 per group, (A) ** P < 0.01, * P < 0.05 versus the sham controls; (B) a and b (P < 0.05) versus sham and STZ2w, respectively, c (P < 0.001) versus STZ3w). Bars indicate means ± SEs


Hyperglycemia causes oxidative stress by inducing the overproduction of reactive oxygen species (ROS), which results in an imbalance between free radical production and the ability of the antioxidant defense system to remove free radicals [17]. It has been reported that ROS concentrations in brain are significantly increased in diabetic rats [1719]. Neurons and axons are particularly sensitive to ROS damage because of their high polyunsaturated lipid contents; lipid peroxidation has been suggested to increase cell membrane rigidity and impair cellular function [20]. However, studies on STZ-induced diabetes differ in terms of the time-points used after STZ-injections [2124], and thus, in the present study, temporal SOD1 expression patterns were determined after STZ injection.

SOD1 scavenges superoxide radicals (O2) by catalyzing the conversion of two of these free radicals into H2O2 and O2 [22]. However, because the time-points used to measure SOD1 levels in previous studies differed [7, 21, 25], we sought to determine the temporal profile of SOD activation post-STZ. In the present study, although SOD1 activity assays showed no distinctive differences between the STZ and sham controls, and SOD1 immunoreactivities were not detected until STZ2w and were highest at STZ3w. Based on our results, there appears to be delay between ROS increases post-STZ and SOD1 activation. At STZ3w, 4-HNE immunoreactivity in all hippocampus areas was markedly elevated. It is notable that 4-HNE immunolabeling appears to prominently label neuronal layers in CA1-3 and the dentate gyrus. 4-HNE is an aldehyde that is generated during lipid peroxidation, and can impair cellular function [26], block neurite out growth [27], and disrupt neuronal microtubules [28, 29]; furthermore, its elevation implies cellular damage by ROS [30]. In addition, 4-HNE also induces apoptotic cell death in PC12 cells and primary rat hippocampal neurons [31, 32]. Taken together, neurons are most easily influenced by insult such as 4-HNE. Therefore, preventing 4-HNE insults may protect neurons during neurodegenerative progression [33]. In addition, 4-HNE has a considerably longer half-life than free radical species and is capable of inhibiting the function of key enzymes such as glyceraldehyde-3-phosphate dehydrogenase [34], which explains the presence of 4-HNE immunoreactivity in the hippocampus at STZ4w. Furthermore, 4-HNE is the most toxic and most abundant aldehyde produced by lipid peroxidation, and its expression has been associated with a variety of cytotoxic and genotoxic consequence [35].

The SOD1 increases observed in the present study, may have been induced by 4-HNE, which is known to form pathological adducts. Some reports have reported SOD1 activities in vitro, following the isolation of the protein under non-physiological conditions, and in the absence of a correlation with in vivo activity [36, 37]. Oxidative stress plays a pivotal role in the damage of neuronal structures and their functions. Furthermore, it has been reported that at 72 h after STZ treatment (50 mg/kg) rats exhibited increased SOD and catalase activities but no alteration in glutathione peroxidase activity in brain. On the other hand, in one study, increased catalase levels were found to reduce glutathione peroxidase activity but not to alter SOD activity in rats at 4 weeks post-STZ [38], whereas in another study, SOD1 (at the mRNA and protein levels) in the brains of rats 2 weeks post-STZ (60 mg/kg) was increased, except in a region of the cerebellum, in which mRNA levels of Cu,Zn-SOD were reduced in STZ-diabetic rats [39].

Microglia are immunocompetent cells [40], and there exists a strong association between free radical accumulation in brain and the evolutions of inflammation and inflammatory-related responses [21]. Although the functions of resting microglia have not been determined, they are known to activate rapidly in response to brain injury or infection [4143]. In addition, it is known that activated microglia are both phagocytic and potent sources of reactive oxygen and nitrogen intermediates [4447], and that reactive microglia are involved in the removal of debris resulting from neuronal degeneration [48, 49]. In the present study, microglia were highly activated until STZ2w, but at STZ3w this was activation was dramatically reduced. This result is in accord with the findings of Chang et al., who found that transient SOD1 overexpression in microglial cells reduces cytokine release, which suggests that SOD1 could act to attenuate or control microglial reactivity [50]. Based on such results, it was suggested that SOD could control or attenuate destructive inflammatory processes [40]. In the present study, 4-HNE expression was much attenuated after SOD1 expression increased at STZ4w. However, SOD1 expression seemed to fade away before 4-HNE was entirely removed. 4-HNE expression was significantly increased in CA3 and dentate gyrus, and SOD1 expression was significantly increased in the dentate gyrus, but tendency to increase in CA3 at STZ4w. Thus, we can estimate that resting microglia at STZ3w were re-activated at STZ4w. In addition, we found that plasma corticosterone levels were significantly higher at STZ2 and 4w in STZ treated rats than in sham controls. However, only a non-significant increase was observed in corticosterone at STZ3w. Since plasma corticosterone levels are known to be controlled by the hypothalamus-pituitary-adrenal axis [5156], it appears possible that antioxidant expressions and reduced inflammatory responses have a moderating influence on the stresses induced by STZ at 3 weeks post injection. Based on our results, a chain of protective and homeostasis maintaining events appears to have been initiated in hippocampus at STZ3w. We also examined IFN-γ and IL-1β expressions to confirm inflammatory response in the hippocampus, and the protein levels of both were found to increase gradually after STZ injection. Microglial responses were once decreased after increase of SOD1, but the expressions of pro-inflammatory cytokines increased continuously with significance. As was previously reported, endogenous protective regulatory signals in the brain have been identified that inhibit microglial overactivation [57, 58]. However, it has been proposed that when the ability to activate these protective mechanisms fails, or when they are overwhelmed by an excessive inflammatory response, microglia initiate neuronal death [58, 59]. Accordingly, we propose that STZ insult exhibits the endogenous protective mechanisms failed even though anti-oxidative stress mechanisms were operated such as SOD1, thus, the proinflammatory cytokines are cumulatively increased by the maintaining toxicity of STZ.

It is known that in the long-term diabetes causes hippocampal dysfunction, and recently, it was found that NeuN-positive cell numbers are reduced during the early stage of diabetes development [60]. However, in the present study, NeuN-positive cell numbers were not significantly decreased in any area, except the DG, at STZ2, 3 or 4w (in the DG significant reductions were observed at STZ4w vs. sham controls). Unfortunately, few reports have been issued on neuronal maturation in type I diabetes. However, neuronal maturation in the hippocampus appears to be affected by diabetes, which suggests that the development of clinically relevant cognitive deficits in diabetic patients might be progressive.

According to Biessels et al. progressive change, along with defects in the central and peripheral nervous systems may become clear at various times, some later than the time intervals in the present study by STZ-induced type 1 diabetes [61]. Therefore, a 4-week duration of STZ-induced diabetes appears to be a relatively short time, although dynamic changes in the brain are more easily detected during early diabetes development. However, if STZ-treated animals are aged, such as, in Biessels et al.’s study, it would not be possible to differentiate between the effects of STZ-induced oxidative stress and aging.

In conclusion, many reports have been issued on free radical formation and SOD1 reactions in STZ-induced type I diabetes animal models. However, the times at which SOD1 is activated post-injection have not been determined. In the present study, we examined the activation of SOD1 at three time-points after STZ injection, and the temporal patterns observed were in-line with those of microglial marker (Iba-1) and matured neuronal marker (NeuN). Observed microglial activity post-STZ concurred with previous reports, however, the matured neuron generally showed a decline tendency. Based on these results, we suggest that the augmentation of STZ-induced cell damage by ROS in the brain leads to an increase in SOD1 activity at 3 weeks post STZ and to a concurrent decrease in microglial response. We anticipate that these oxidative stress-related events in the brains of our STZ-induced diabetes animal models are accomplished at 3 weeks post-STZ administration.


The authors would like to thank Mr. Seung Uk Lee and Ms. Hyun Sook Kim for their technical help. This work was supported by a National Research Foundation of Grant funded by the Korean Government [NRF-2009-352-E00051].

Copyright information

© Springer Science+Business Media, LLC 2010