Ketamine ameliorates depressive-like behaviors in mice through increasing glucose uptake regulated by the ERK/GLUT3 signaling pathway

To investigate the effects of ketamine on glucose uptake and glucose transporter (GLUT) expression in depressive-like mice. After HA1800 cells were treated with ketamine, 2-[N-(7-Nitrobenz-2-oxa-1,3-diazol-4-yl)Amino]-2-Deoxyglucose (2-NBDG) was added to the cells to test the effects of ketamine on glucose uptake, production of lactate, and expression levels of GLUT, ERK1/2, AKT, and AMPK. Adult female C57BL/6 mice were subjected to chronic unpredictable mild stress (CUMS), 27 CUMS mice were randomly divided into the depression, ketamine (i.p.10 mg/kg), and FR180204 (ERK1/2 inhibitor, i.p.100 mg/kg) + ketamine group. Three mice randomly selected from each group were injected with 18F-FDG at 6 h after treatment. The brain tissue was collected at 6 h after treatment for p-ERK1/2 and GLUTs. Treatment with ketamine significantly increased glucose uptake, extracellular lactic-acid content, expression levels of GLUT3 and p-ERK in astrocytes and glucose uptake in the prefrontal cortex (P < 0.05), and the immobility time was significantly shortened in depressive-like mice (P < 0.01). An ERK1/2 inhibitor significantly inhibited ketamine-induced increases in the glucose uptake in depressive-like mice (P < 0.05), as well as prolonged the immobility time (P < 0.01). The expression levels of p-ERK1/2 and GLUT3 in depressive-like mice were significantly lower than those in normal control mice (P < 0.01). Ketamine treatment in depressive-like mice significantly increased the expression levels of p-ERK1/2 and GLUT3 in the prefrontal cortex (P < 0.01), whereas an ERK1/2 inhibitor significantly inhibited ketamine-induced increases (P < 0.01).Our present findings demonstrate that ketamine mitigated depressive-like behaviors in female mice by activating the ERK/GLUT3 signal pathway, which further increased glucose uptake in the prefrontal cortex.

In the FST (Fig. 2C), a significant increase in immobility time was found in depressive-like mice compared to that of control mice (depressed vs control: 114.83 ± 16.03 s vs. 60.67 ± 16.02 s, P < 0.01). The immobility times of depressive-like mice treated with ketamine were significantly reduced when compared with those of depressivelike mice treated with saline (58.33 ± 12.67 s, P < 0.01). In contrast, addition of an ERK1/2 inhibitor increased the immobility times of depressive-like mice treated with ketamine compared with those of depressive-like mice only treated with ketamine (92.67 ± 8.65 s, P < 0.05).
In the NSF test (Fig. 2D), depressive-like mice presented a significant increase in the latency to feed compared to that of the control group (depressed vs. control: 108.06 ± 39.42 s vs. 46.02 ± 8.12 s, P < 0.01). Ketamine treatment significantly decreased the latency to feed of depressive-like mice to 44.32 ± 15.21 s (P < 0.01), which was increased in depressive-like mice when they were additionally treated with an ERK1/2 inhibitor combined with ketamine (82.04 ± 23.30 s, P < 0.05).
Ketamine increases glucose uptake in the prefrontal cortices of depressive-like mice via the ERK1/2 signaling pathway. As shown in Fig. 3, compared with that of the control group, the decrease of glucose uptake was found in the prefrontal cortices of depressive-like mice, as demonstrated by significantly lower normalized SUV values (P < 0.01). Ketamine treatment significantly increased glucose uptake, as demonstrated by higher normalized SUV values in depressive-like mice compared with those of depressive-like mice without ketamine treatment (P < 0.01). Inhibition of ERK1/2 by FR180204 in ketamine-treated depressive-like mice decreased glucose uptake, as demonstrated by a significant decrease in normalized SUV values (P < 0.05).
Ketamine upregulates the protein levels of P-ERK1/2 and GLUT3 in the prefrontal cortices of depressed mice via the ERK1/2 signaling pathway. As shown in Fig. 4, the protein levels of P-ERK1/2 (P < 0.05) and GLUT3 (P < 0.05) in the prefrontal cortex were decreased in depressed mice compared with those in control mice. Ketamine significantly increased the expression of P-ERK1/2 (P < 0.01) and GLUT3 (P < 0.01) in the prefrontal cortex of depressed mice. Treatment with an ERK1/2 inhibitor (FR180204) decreased these protein levels in the prefrontal cortices of depressed mice treated with ketamine (P < 0.01).
Astrocytes, but not neurons, are involved in the effects of ketamine mitigating depressive-like behaviors in mice via the ERK1/2 signaling pathway. We also measured the expression levels and spatial distributions of P-ERK1/2 and GLUT3 in the prefrontal cortices of mice. As shown in Fig. 5, upregula- (B) The expression of GLUT1, GLUT3 and GLUT4 were examined by Western blots at 6 h post ketamine treatment with different concentration (0, 10, 25, 50 and 100 μM); (C) Western blots analysis was used to test the levels of P-ERK1/2, P-AKT and P-AMPK at 6 h post ketamine treatment with different concentration (0, 10, 25, 50 and 100 μM); (D) The subcellular location of P-ERK1/2 response to ketamine treatment was detected by immunofluorescent staining; (E) The production of lactate was tested by a lactate detection kit. Vs control, *P < 0.05 and **P < 0.01. Immunoblotting signals were quantified via Image J software. www.nature.com/scientificreports/ tion of P-ERK1/2 and GLUT3 in the prefrontal cortices of mice in the ketamine group was mainly distributed in astrocytes; the upregulation of P-ERK1/2 and GLUT3 in the prefrontal cortex in the ketamine group was less distributed in neurons. Furthermore, P-ERK1/2 in the prefrontal cortices of mice in the ketamine group was aggregated in the nuclei of astrocytes.

Discussion
Although several studies have reported that ketamine exerts rapid and sustained antidepressant effects in patients with depression, the partial underlying mechanisms have been elucidated [19][20][21][22][23][24][25] . Additionally, ketamine-induced antidepressant effects are accompanied by an increase in the cerebral metabolic rate of glucose (CMRGlc) 26 . Therefore, it is possible that ketamine-enhanced CMRGlc is due to an increase in glucose uptake within the brain. Astrocytes play numerous complex functions in the central nervous system 17 , and a reduction in astrocytes and their related markers are associated with the pathology of major depressive disorder 27,28 . Hence, in the present study, we assessed the effects of glucose uptake induced by ketamine in human-derived astrocytes (HA1800). Our results indicated that treatment with 50 μM (equivalent to subanesthetic dose 29 ) of ketamine for 6 h remarkably enhanced glucose uptake in astrocytes. However, the mechanism of glucose uptake induced by ketamine in astrocytes remains unclear.
Relevant studies have shown that GLUTs play a key role in the glucose uptake of astrocytes 30 , among which GLUT1, GLUT3, and GLUT4 play important roles 31 . Hence, in the current study, we detected the expression levels of GLUT1, GLUT3, and GLUT4 induced by ketamine in astrocytes. Our findings indicated that ketamine induced glucose absorption in astrocytes through GLUT3 instead of GLUT1 or GLUT4. Under physiological conditions, the expression levels of GLUT1 and GLUT3 are detected at relatively low levels, and astrocytes exhibit increased expression levels of GLUT3 under some stress conditions, such as mild ischemia 32 . Additionally, the endotoxin lipopolysaccharide (LPS) 33 can lead to an increase in glucose uptake. GLUT4 has been identified as an insulin-responsive transporter, and its regulatory mechanisms differ from those of GLUT1 or GLUT3 34 . Since GLUT3 transports extracellular glucose into cells at a rate of approximately seven-fold that of GLUT1 35 , we hypothesize that ketamine-induced increases in GLUT3 levels in HA1800 cells in our present study may www.nature.com/scientificreports/ lead to the significant increase in glucose uptake. The results of the present study are consistent with those of Tomioka et al. in which they reported the effects of ketamine on glucose uptake by GLUT3 expressed in Xenopus oocytes 36 . In addition, Iasevoli et al. found that ketamine increases GLUT3 mRNA expression in rats 37 . However, it is still unclear which signaling pathway may be associated with ketamine-induced increases in GLUT3 levels in astrocytes. Several studies have shown that active ERK1/2, AKT, and AMPK signaling pathways accelerate glucose uptake in astrocytes 38,39 . The findings of our present study indicated that ketamine treatment markedly increased the expression levels of p-ERK1/2, while it did not influence the expression levels of p-AKT or p-AMPK. Furthermore, immunostaining assays showed that p-ERK1/2 translocated into the nucleus following ketamine treatment,  18 F-FDG PET CT brain imaging; the more red, the more glucose uptake. (B) The SUV was prefrontal-cortex SUV (normalized to that of the whole-brain SUV). NC normal control group, D depression group, K ketamine group, FR ERK1/2 inhibitor (FR180204) group; n = 3; *P < 0.05, **P < 0.01. Images were analyzed using ASI Pro VM software (Siemens Medical Solutions, USA). ; NC normal control group, D depression group, K ketamine group, FR: ERK1/2 inhibitor (FR180204) group; n = 6; *P < 0.05, **P < 0.01. Protein levels of P-ERK1/2, T-ERK1/2, and GLUT3 in the prefrontal cortices were quantified by Image J software. www.nature.com/scientificreports/ which may generate physiological actions such as proliferation and differentiation 40,41 . However, whether p-ERK1/2 further regulates the expression of GLUT3 after entering the nucleus requires further investigation. Our present findings revealed that ketamine increased glucose uptake in astrocytes. However, we did not explore which metabolic process participated in this glucose uptake. An ANLS model ( Fig. 6) showed that glucose is taken up by astrocytes located around blood vessels and is converted to lactate, the latter of which is transferred into neurons as energy substrates 42 . Although we demonstrated an increased lactate level in astrocytes, whether lactate provides energy to neurons requires further investigation.
On the basis of the above results in vitro, we further explored the related phenomena and mechanisms by establishing a mouse model of depression. CUMS, which induces long-lasting depressive-like behavior for up to several months, is a well-established model of depression in rodents and can be reversed by chronic treatments with traditional antidepressant agents [43][44][45] . Compared to those of males, females have higher incidence rates of major depressive disorder 46,47 . Experimental studies have also reported a sex difference in depressive-like behaviors, with more remarkable depressive-like behaviors in female rather male LPS-exposed or CUMS mice 48,49 . Therefore, we established a mouse model of depression in female mice in the current study. When we designed the experiment, we deliberately chose the stimulation factors that would not cause damage to the limbs or affect the motor ability of mice, all kinds of stress involved in the experiment were mild and non-violent as we previously reported 5 , therefore, in the study, we did not again detect the influences of the treatment administered on the motor ability of mice before performing the OFT. Male rats have been previously reported to exhibit depressivelike behavior after 4-8 weeks of CUMS [50][51][52] . In the present study, we successfully established depressed-like female mice after 21 days of CUMS, as demonstrated by results of the SPT and FST. Therefore, this provides a model to explore potential targets for the treatment of depression.
In the current study, results of behavioral tests in depressed-like mice indicated that ketamine treatment dramatically ameliorated depressive-like behaviors. This finding is consistent with previous experiments performed by Berman et al. 53 and Zarate et al. 54 . In addition, many authors report behavioral effects in mice submitted to www.nature.com/scientificreports/ the CUMS or corticosterone model with ketamine at 3 or 5 mg/kg. We use ketamine at 10 mg/kg, which do not cause schizophrenia, because the subclinical doses of ketamine (10 mg/kg) used in the study of antidepressant effect of ketamine in mice has been recorded in the literature 55 .
Early studies have reported a relationship of ketamine with phosphorylation levels of the ERK signaling pathway and anti-depressant effects 23,56 . In our present study, we found that inhibition of ERK signaling significantly eliminated the improved effects of ketamine on depressive-like behaviors of depressed mice, further confirming the essential roles of ERK signaling in the anti-depressant effects of ketamine.
The prefrontal cortex, considered as a large part of the neural system crucial for normal socio-emotional and executive function in humans and other primates, has been suggested to be tightly associated with depression 57,58 . Animal studies have documented a dramatic decrease in glucose uptake within the prefrontal cortex by 18 F-FDG micro-PET/CT, which is increased after anti-depressive treatment 59 . A Clinical PET study reported lower glucose metabolism in the prefrontal cortices of depressed patients at rest, as compared with that of control subjects 6 . Another randomized controlled study also found that low-dose ketamine increased glucose uptake in the prefrontal cortex in treatment-resistant depressed patients 7 . Together with these previous studies, our present results further demonstrated an association of antidepressant effects of ketamine with glucose metabolism in the prefrontal cortex. Both astrocytes and neurons are located in the prefrontal cortex. However, it has remained unclear whether astrocytes and/or neurons participate in the antidepressant effects of ketamine on glucose metabolism in the prefrontal cortex. Furthermore, few studies have elucidated the mechanisms of ketamine on glucose uptake involving the ERK/GLUT3 signaling pathway in the prefrontal cortex. In the current study, we found that the protein levels of GLUT3 and P-ERK1/2 were significantly decreased in the prefrontal cortices of depressed mice compared to those in control mice. Low-dose ketamine dramatically increased these protein levels in the prefrontal cortex. Such antidepressant effects of ketamine were subsequently inhibited by the additional treatment of an ERK1/2 inhibitor. Our results are consistent with previous findings by Lee et al. 12 and Iasevoli et al. 37 . Immunofluorescent staining further confirmed that astrocytes, but not neurons, exhibited enhanced expression levels of GLUT3 and P-ERK1/2 in the prefrontal cortex. Previous studies have revealed that nuclear translocation of P-ERK1/2 plays an important role in the initiation of proliferation and differentiation 40,41 . Furthermore, Watanabe et al. 60 reported that GLUT3 expression in cancer cells by DNA-damaging agents was dependent on the ERK pathway. Therefore, in our present study, we hypothesize that nuclear localization of P-ERK1/2 in astrocytes subsequently enhanced glucose metabolism in the prefrontal cortex via promoting the expression of GLUT3. However, the potential mechanisms of P-ERK1/2 regulating the expression of GLUT3, the involvement of glucose uptake in astrocytes, and energy metabolism in peripheral neurons remain unclear. Hence, further studies are warranted to elucidate these mechanisms.
In conclusion, our present study revealed that ketamine ameliorated depressive-like behaviors in mice by enhancing glucose uptake in the prefrontal cortex via activating the ERK/GLUT3 signaling pathway.

Immunostaining analysis.
Immunostaining analysis was carried out to detect localization of total-ERK1/2 and phospho-ERK1/2 in astrocytes. HA1800 cells were seeded in 24-well plates overnight. After treatment with ketamine, cells were immediately fixed with 4% PFA for 20 min and were then washed three times with cold PBS. After blocking with 1% bovine serum albumin (BSA) for 1 h at room temperature, the cells were incubated with specific primary antibodies at 4 °C overnight. After washing with PBS, the cells were incubated with fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG (Bioss, ID bs-0295G-FITC, dilution, 1:500) for 1 h at room temperature. Next, cellular nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). After addition of antifade mounting medium, fluorescent signals were immediately examined under a fluorescent microscope. The fluorescent signals were quantified via Image J software 1.47v (National Institutes of Health, USA).
Lactate production assays. HA1800 cells were seeded in 10-cm dishes and maintained for 24 h, after which they were treated with ketamine and/or an ERK1/2 inhibitor for 6 h. Thereafter, the culture medium from each sample was transferred to an Eppendorf (EP) tube, and lactate production was then measured by a lactate detection kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer's protocol.
Animals. Female C57BL/6 mice (18-22 g, 6-8 weeks, n = 90) were obtained from Chengdu Dossy Experimental Animals Company of China, and were maintained under a controlled temperature (22 °C ± 2 °C) and humidity (50% ± 20%) with a 12/12-h light/dark cycle (lights on at 7:00 a.m.). All animals were provided food and water ad libitum, except when they were 24-h food-deprived prior to the novelty suppressed feeding (NSF) test and 24-h water-deprived prior to the sucrose preference test (SPT). Mice were acclimatized for 7 days prior to the start of experiments. The experimental flowchart is presented in Fig. 7. After baseline assessment of depressive-like behaviors (excluding 9 mice with abnormal baseline data), 81 mice were then randomly assigned to the www.nature.com/scientificreports/ following two groups: (1) normal control group (NC, n = 9) without any treatment; and (2) test group (n = 72) with chronic unpredictable mild stress (CUMS). Successful establishment of our CUMS model was determined by significant differences in depressive-like behaviors acquired on experimental day 21 when compared with those of the control group. Ultimately, 27 CUMS mice were randomly assigned to the following three groups: (1) depressive-like group (D group, n = 9), administered intraperitoneally (i.p.) with 0.9% saline; (2) ketamine group (K group, n = 9), administered i.p. with 10 mg/kg of ketamine (this dose is appropriate for producing antidepressive effects in rats, and the rats were still conscious for flowing injection of this dose 61  Chronic unpredictable mild stress (CUMS). CUMS, as described by Wang et al. 62 , was performed with some modifications. The experimental stressors were as follows: horizontal oscillation for 1 h; physical restraint for 1 h; tail clipping for 5 min; wetting litter for 12 h; cage tilting for 12 h; swimming in 10 °C water for 3 min; swimming in 45 °C water for 3 min; continuous illumination during the dark phase; and water or food deprivation. Mice were administered these stressors in a random order for 21 days, such that we avoided using the same stress for two consecutive days, and a given stress was not used more than two times in a week.

Sucrose preference test (SPT).
The SPT was carried out on experimental day 8, 31, and 34 in each mouses home cage, as previously reported 12 . At the beginning of the experiment, mice were trained to habituate to sucrose solution by placing two bottles of either 2.5% sucrose solution or tap water in each cage for 24 h, with the positions of these bottles changing every 12 h. All mice were then water-deprived for 24 h prior to the SPT. On day 3, each mice was given free access to same two bottles of either 2.5% sucrose solution or tap water for 1 h, with the positions of these bottles changing every 30 min. The sucrose preference rate (%) was calculated as follows: sucrose consumption mL/(total water + sucrose consumption ml) × 100%.
Novelty suppressed feeding (NSF) test. The NSF test was performed on experimental days 8, 31, and 34, as previously described 63 . The latency of each mouse to approach and eat food was measured in a novel environment following an extended period (up to 24 h) of food deprivation. Approximately 24 h after the removal of the food, mice were placed in an illuminated and sound-proofed box (40 × 40 × 40 cm) with a small piece of mouse chow placed in the center of the box. Each mouse was placed in the corner of the testing arena, and the time until the first feeding episode was recorded within 5 min.
Tail suspension test (TST). The tail suspension test (TST) was performed on experimental days 9, 32, and 34, according to a previous protocol 64 . Mice were individually suspended by their tails using adhesive tape placed approximately 1 cm from the tip of the tail, keeping mice positioned 15 cm above the tabletop. The entire pro- Brain tissue collection. Mice were anesthetized by administration with 1% sodium pentobarbital (50 mg/ kg) after behavioral tests. All mice were transcardially perfused with normal saline and were decapitated immediately. Whole brains were rapidly separated on ice. One side of the prefrontal cortex was quickly dissected and snap-frozen with liquid nitrogen prior to storage at − 80 °C until later use. The other side of the prefrontal cortex was promptly immersed in 4% paraformaldehyde overnight, and was then stored at − 80 °C for future use following dehydration in 20% and 30% sucrose solutions.
Western blotting. Prefrontal cortices were mechanically homogenized in RIPA buffer combined with protease and phosphatase inhibitors (Solarbio, Beijing, China). Lysates were centrifuged (12,000 rpm for 15 min, at 4 °C), and total protein concentrations in the supernatants were quantified using BCA protein assay kits (Beyotime, Shanghai, China). The samples were separated by SDS-PAGE using a 10% gel, and the proteins were transferred to PVDF membranes (Millipore, Massachusetts, USA) using a semi-dry blotting apparatus (1.2 mA/cm 2 ; 1.5 h). Subsequently, the membranes were blocked with 5% BSA in TBS (20 mM Tris-Cl, 150 mM NaCl, pH 8.0). The protein levels of GLUT3 (1:100; Santa Cruz, Dallas, Texas, USA), T-ERK1/2 (1:1000; Santa Cruz, Dallas, Texas, USA) P-ERK1/2 (1:1000; Beyotime, Shanghai, China), and β-actin (1:5000, loading control; Proteintech, Wuhan, Hubei, China) were determined via overnight incubation in corresponding primary antibodies diluted in TBS-T (20 mM Tris-Cl, 150 mM NaCl, 0.05% Tween-20, pH 8.0). The membranes were then incubated with horseradish-peroxidase-conjugated secondary antibodies (anti-mouse antibody, 1:5000, Proteintech, Wuhan, Hubei, China; or anti-rabbit antibody, 1:2500, Proteintech, Wuhan, Hubei, China) for 60 min. The immunoreactive bands were developed using a chemiluminescence kit (Beyotime, Shanghai, China). All blocking and incubation steps were followed by three washes (5 min each wash) of the membranes with TBS-T. The optical densities (ODs) of the protein blots were quantified using the Image J software (National Institutes of Health, USA). The phosphorylation levels of ERK1/2 were determined as a ratio of the OD of the phosphorylated protein band to the OD of the total protein band. GLUT3 protein levels were determined as a ratio of the OD of the GLUT3 band to the OD of the β-actin band.
Immunofluorescent assays. The prefrontal cortices of mice were analyzed via immunofluorescent assays.
Statistical analysis. Statistical analysis was carried out using the GraphPad Prism software 8.0.2 (https:// www. graph pad. com/). The data are presented as the mean ± standard error of the mean (SEM). Differences among experimental groups were determined by one-way analysis of variance (ANOVA) followed by Bonferroni tests for post-hoc comparisons. P < 0.05 was considered statistically significant.

ARRIVE Guidelines Statement
Here we state that all methods in this study are reported in accordance with ARRIVE guidelines.