STZ was from Sigma (Sigma, St. Louis, MO, USA). Antibodies employed in this study are listed in Supplementary Table 1.

Two-month-old male (n = 24) and female (n = 24) Sprague-Dawley (SD) rats were provided by the Experiment Animal Center of Tongji Medical College, Huazhong University of Science and Technology. All animal experiments were performed according to the “Policies on the Use of Animals and Humans in Neuroscience Research” approved by Society for Neuroscience in 1995 and approved by the Experiment Animal Center of Tongji Medical College, Huazhong University of Science and Technology. The animals were individually housed in cages (house temperature 24°C, controlled humidity 40% and 12/12 h inverted light cycle) with free access to water and food.

STZ, soluble in artificial cerebrospinal fluid (aCSF), was injected slowly (1 μl/min) into the ventriculus lateralis cerebri of rats (10 μl, 3 mg/kg). Control animals were identically treated with the same volume of aCSF. After 30 days, morris water maze was employed to train and test spatial learning and memory. After this procedure which lasted for 7 days, mice were sacrificed and other tests were proceeded (Supplementary Figure 1).

The water maze used was a circular, steel pool (1.6 m in diameter) that was filled with black water (temperature 25°C) that was non-toxic and contrast to rat. A black-colored, circular platform (12 cm in diameter) was placed below the water surface at a specific location. Distinctive visual cues were stuck to the wall. For spatial training, rats were subjected to 4 trials each day from 2:00 to 5:00 p.m. The training was lasted 6 days and 24 trials were given to every rat. For each trial, the rat was placed at different starting position spaced equally around the perimeter of the pool. Rats were allowed to find the submerged platform within 60 s. If the rat could not find the hidden platform, it was then gently guided to the platform and allowed to stay there for 30 s. The time that each rat took to reach the platform was recorded as the escape latency. For the probe trial test, rats were submitted to the same pool with the platform removed and a probe trial of 60 s was given. The number of crossings and the time in the target quadrant were recorded.

The protocol was performed as previously described (Xu et al., 2014). Four left hippocampus per group for Western blotting. Hippocampi were rapidly dissected out and homogenized in a buffer containing NaF 50 mM, Tris-Cl (pH 7.6) 10 mM, 1 mM EDTA, 1 mM Na₃VO₄, 1 mM benzamidine, and 1 mM phenylmethylsulfonylfluoride (PMSF), 10 g/ml leupeptin, and 2 g/ml each of pepstatin A and aprotinin. The homogenates were added to one-third of sample buffer containing 200 mM Tris-HCl (pH 7.6), 8% sodium dodecyl sulfate, and 40% glycerol, boiled in a water bath for 10 min, and then centrifuged at 14,510 r for 10 min. Protein concentration of the supernatants were measured by the bicinchoninic acid Protein Assay Kit (Pierce, Rockford, IL, USA). Ten micrograms of protein for DMIA and pS396 antibodies, 20 μg protein for other antibodies, were loaded and separated by SDS-polyacrylamide gel electrophoresis (10% gel), and then transferred to a nitrocellulose membrane. After blocking in 3% non-fat milk for 1 h, the nitrocellulose membranes were incubated with primary antibodies at 4°C overnight. The membranes were then incubated with secondary antibodies conjugated to IRDye (800CW) for 1–2 h and visualized using the Odyssey Infrared Imaging System (LI-Cor Biosciences, Lincoln, NE, USA). Image J software was employed for the quantitative analysis of the western blots.

The Golgi staining protocol was performed as previously described (Morest, 1960). Three per group were used for Golgi Staining. The rats were anesthetized with 6% chloral hydrate and perfused with 300 ml of normal saline containing 0.5% sodium nitrite, followed by 400 ml of 4% formaldehyde solution and further by ~400 ml dying solution (4% formaldehyde, 5% potassium dichromate, and 5% chloral hydrate) for 5 h in the dark. The brains were removed and incubated in the same fixative in the dark. After 3 days, the brains were transferred to a solution containing 1% silver nitrate for 3 days in the dark. The silver solution was changed each day. Thirty-five micrometers of thick coronal brain sections were cut using a vibrating microtome (Leica, VT1000S, Germany).

Three per group were used for Immunofluorescence Staining. The anesthetized rats were immediately perfused through the aorta with 300 ml normal saline, followed by a 300 ml solution containing 4% paraformaldehyde. The brains were dissected and post-fixed in 4% paraformaldehyde for another 48 h. Coronal sections (30 μm thick) were cut using a vibrating microtome. After incubation in 0.3% Triton-X100-PBS for 30 min at room temperature, free floating sections were blocked with 5% goat serum in PBS for 45 min at room temperature. Sections were then incubated overnight at 4°C with primary antibodies: polyclonal anti-MAP2 antibodies obtained from Abcam, (dilution 1:200, Cambridge, MA, USA). This was followed by incubation with secondary antibodies for 2 h at room temperature. The antibody staining was semi-quantitated by mean fluorescence intensities (MFIs) with Image J software.

The protocol was performed as previously described (Qi et al., 2016). Five right hippocampus per group for the assay. Beta-secretase activity was monitored using a commercial kit, from Abnova (Neihu District, Taipei City 114 Taiwan) according to the manufacturer instructions and using a multi-well fluorescence plate reader capable of Ex = 335–355 nm and Em = 495–510 nm. In briefly, 50 μl of 4 μg/μl hippocampus lysate was added to a 96-well plate. Fifty microliters of 2×reaction buffer were added, followed by 2 μl of β-secretase substrate. The reaction mixtures were incubating for 1 h in the dark. Fluorescence was monitored at excitation wave (wavelength = 334–355 nm) and emission wave (emission wavelength = 490–510 nm). β-secretase activity can be expressed as the Relative Fluorescence Units (RFU) per μg of protein sample.

The protocol was performed as previously described (Zhang et al., 2015). Five right hippocampus per group for the Elisa assay. To detect the concentration of Aβ in hippocampi lysates, the rat hippocampi were homogenized in buffer (PBS with 5% BSA and 0.03% Tween-20, supplemented with protease inhibitor cocktail), and centrifuged at 16,000 g for 20 min. Aβ1-40 or Aβ1-42 was quantified using the rats Aβ1-40 or Aβ1-42 ELISA Kit (Elabscience, Wuhan, China) in accordance with the manufacturer's instructions. The Aβ concentrations were determined by comparison with the standard curve.

Three right hippocampus per group for the assay. To measure the levels of estradiol in hippocampi lysates and serum, the rat hippocampi were homogenized in 1× PBS and blood is obtained from the orbital vessels, then centrifuged at 1,500 r for 20 min. Estradiol was quantified using the rat estradiol ELISA Kit (CZVV, Nanjing, China). The results were expressed in ng/L.

Data are descriptively presented as means ± SD and analyzed by SPSS 17.0. Statistical analysis was performed using either Student's t-test (two-group comparison) for behavior test, dendritic plasticity, Western blot, enzymatic activity. For the levels of estradiol in serum and hippocampus, we firstly performed a descriptive analysis in Supplementary Tables 2, 3, and then a Shapiro–wilk test for a normal distribution of the samples from four group, finally a general linear model to be used for two-way ANOVA followed by post-hoc comparison, and differences with P < 0.05 were considered significant.

A study showed that a significant cognitive impairment was evoked at the 2nd week onwards, which persisted up to the 14th week with ICV-STZ (3 mg/kg) in rats (Mehla et al., 2013). To investigate whether sex differences exist in cognitive deficits induced by STZ, in the present study, we performed morris water maze to evaluate the memory and learning abilities of rats 30 days after ICV-STZ treatment. For male rats, we found that the escape latency to find a hidden platform dramatically increased while the traversing times and the time in the target quadrant were significantly decreased at the 7th day in ICV-STZ rats when compared to vehicle control (Figures 1A–C). This confirmed that ICV-STZ induced learning and memory deficits in male rats. For female rats, to our surprise, we failed to observe any learning and/or memory deficits. The latency to find the hidden platform, the crossing numbers and time spent in the target quadrant also did not change in female rats (Figures 1E–G). Both groups in male or female rats exhibited comparable swimming speed (Figures 1D,H), indicating that motor function was not affected. Altogether, the findings suggest that ICV-STZ injection induces cognitive impairments in male but not female rats.

Dendrite complexity (Li et al., 2008) and synaptic plasticity (Kasai et al., 2010) are neurobiological basis for learning and memory. We determined the effect of ICV-STZ on neuronal integrity, by examining levels of the dendritic marker MAP2. For male rats, the semi quantitative results showed a strongly reduced mean fluorescence intensities (MFIs) of MAP2 immunoreactivity in the pyramidal neurons of CA1 region of the hippocampus in ICV-STZ rats compared to vehicle control (Figures 2A,B). However, in female rats, MAP2 immunoreactivity showed that ICV-STZ had no effect on dendritic number compared to control (Figures 2A,C). We also examined alterations in dendritic spines using Golgi staining. Mushroom-type spines in the CA1 of ICV-STZ treated male rats decreased remarkably compared to control (Figures 2D,E), but the number of mushroom-type dendritic spines were not altered in the ICV-STZ treated female rats compared to control (Figures 2D,F).

Normal synaptic function is contingent upon the stable expression of synaptic proteins. Therefore, we evaluated several key synapse-associated proteins using Western blotting. ICV-STZ treatment remarkably suppressed the expressions of presynaptic synapsin I, synaptagmin and postsynaptic PSD95, PSD93, NR2A, and NR2B in male rats (Figures 3A,B). Nonetheless, there is no any significant difference between vehicle and ICV-STZ treated female rats (Figures 3C,D). These data demonstrate that ICV-STZ induces loss of dendritic and synaptic plasticity in male, but not in female rats.

Abnormal hyperphosphorylation and accumulation of Tau play a key role in AD pathology (Wang and Liu, 2008), and hyperphosphorylated tau causes dendritic loss and neurodegeneration (Wang et al., 2010). In addition, ICV-STZ treatment induces tau hyperphosphorylation in rats (Zhou et al., 2013). In this study, we also explored whether ICV-STZ induces tau hyperphosphorylation in male or female rats, respectively. We detected a significantly increasing tau phosphorylation at the Ser199/202(AT8), Ser262, Ser396, and Ser404 sites in ICV-STZ treated male rats (Figures 4A,B). Conversely, in the female rats, ICV-STZ did not induce tau hyperphosphorylation (Figures 4C,D).

GSK-3β is the first identified and critical tau kinase (Singh et al., 1995), therefore we evaluated the total level and the activity-dependent phosphorylation of GSK-3β. In male rats, we found that the p-GSK-3β (Ser9) (the inactive form) level was remarkably decreased, while the level of total GSK-3β and p-GSK-3β (Tyr216) (the active form) didn't change (Figures 5A,B). In female rats, no significant difference was observed between ICV-STZ treated and vehicle control (Figures 5C,D). Taken together, these findings suggest that ICV-STZ activates GSK-3β and consequently leads to hyperphosphorylation of tau protein in male rats, but does not elicit these demonstrable pathological alterations in female rats.

Another characterized histology of AD is extracellular senile plaques, which are composed of aggregated protein Aβ initiated by β-secretase (BACE1) (Alafuzoff et al., 1987; Vassar et al., 2009). We employed β-secretase Activity Assay Kit and Aβ40/42 Elisa Kit to detect BACE1 activity and Aβ levels of hippocampus. In ICV-STZ treated male rats, both BACE1 activity and Aβ40 level were significantly increased compared to control rats, while Aβ42 showed ascendant trend without significant difference (Figures 6A,C,D). However, in female rats, BACE1 activity and Aβ40/42 levels were not altered in both ICV-STZ and vehicle treated rats (Figures 6B,E,F). Thus, these data strongly support that ICV-STZ increases BACE1 activity and augments Aβ production in male rats, while did not exhibit these toxic effects in female rats.

Previous studies have shown that estradiol reduces Aβ production via reducing total BACE1 activity, and decreases tau hyperphosphorylation by mediated GSK-3β activity (Singh et al., 1999; Zhang et al., 2008). To investigate whether estradiol influences the generation of the sporadic AD animal model induced by ICV-STZ, we measured estradiol levels in serum and hippocampus of ICV-STZ treated male and female rats. Shapiro–wilk test showed that all of the p-values were > 0.05, indicating estradiol levels in serum (Supplementary Table 4) and hippocampus (Supplementary Table 5) according with normal distribution from each group. And then, a general linear model was used for two-way ANOVA (Supplementary Tables 6, 7) followed by post-hoc comparison. We found that estradiol levels of serum (Figure 7A) and hippocampus (Figure 7B) of female rats were much higher than that of male rats, while no difference was observed between the groups with same sex. The data suggest that high estradiol might protect from STZ induced neurotoxic effects in female rats.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.