Probable involvement of p11 with interferon alpha induced depression

Depression is one of the major side effects of interferon alpha (IFN-α) treatment, but the molecular mechanism underlying IFN-α-induced depression remains unclear. Several studies have shown that the serotonin receptors 5-HTR1b and 5-HTR4 play key roles in the anti-depression effects associated with p11 (S100A10). We investigated the effects of IFN-α on the regulation of p11, 5-HTR1b and 5-HTR4 in mice and human neuroblastoma cells (SH-sy5y). We found that intraperitoneal injection with IFN-α in Balb/c mice resulted in an increased immobility in FST and TST, and potently lowered the protein levels of p11, 5-HTR1b and 5-HTR4 in the hippocampus or cingulate gyrus. IFN-α significantly down-regulated the protein levels of p11, 5-HTR1b and 5-HTR4 in SH-sy5y cells, in a time- and dose-dependent manner. Our study revealed that over-expression of p11 could prevent the IFN-α-induced down-regulation of 5-HTR1b and 5-HTR4. The results indicated that IFN-α treatment resulted in p11 down-regulation, which subsequently decreased 5-HTR1b and 5-HTR4 in vitro or in vivo. Our findings suggested that p11 might be a potential regulator on 5-HTR1b and 5-HTR4 as well as a predictor of or a therapeutic target for IFN-α-induced depression.

with psychiatric side effects in cancer patients treated with IFN-α 15 . In addition, there is no association between the variants in the IOD gene polymorphisms and the diagnosis of the IFN-α -induced depression 16 . p11, or S100A10, a member of the S100 family of proteins with 2 EF-hand calcium-binding motifs, is recently reported to play a crucial role in the onset of depression by interacting with the 5-hydroxytryptamine receptor 1b (5-HTR1b) or 5-hydroxytryptamine receptor 4 (5-HTR4) 17,18 . The levels of the p11 protein are significantly increased in some rat brain regions after treatment with anti-depressants 19 , whereas p11 mRNA levels are significantly decreased in the depressive rodent model and in patients with severe depression 20,21 . p11 knockout mice show spontaneous depressive behavior and a low response to 5-HT1b-stimulating agents 17,22,23 . However, it remains unknown whether p11 is involved in IFN-α -induced depression.
5-HTR1b and 5-HTR4 are reportedly critical receptors involving in the development of depression. 5-HTR1b is previously identified to be an autoreceptor that inhibits the release of serotonin in the presynaptic membrane 24 . Recently, the activation of 5-HTR1b in the post-synapse was found to facilitate excitatory synaptic transmissions associated with depression 25 . 5-HTR1b agonists are found to be effective anti-depressants in animal models 26 . 5-HTR4 plays roles in both the peripheral and central nervous systems to modulate the release of various neurotransmitters 27 . 5-HTR4 agonists have rapid effects on depressive behaviors, and its activation is necessary for these effects of serotonin-specific reuptake inhibitors (SSRIs) 28 . 5-HTR1b and 5-HTR4 have become novel targets for the pathophysiology of depression and for antidepressants therapies 29,30 . Even though the anti-depression effects of p11 have been well documented and an uncovered fundamental mechanism suggests that p11 alters the distribution of 5-HTR1b and 5-HTR4 in cells 17,18 , the effect of p11 on the expression of 5-HTR1b and 5-HTR4 is still unknown. Furthermore, the regulating effects of IFN-α on levels of 5-HTR1b and 5-HTR4, and its dependence on p11 still need to be elucidated.
In the present study, we hypothesized that IFN-α treatment resulted in depression through the inhibition of p11 and subsequent reduction of the density of 5-HTR1b/4 in nerve synapses of brain's certain areas which were associated with depression, thereby causing nervous system dysfunction and, ultimately, depression. Therefore, we investigated the behavioral and histological responses of Balb/c mice after IFN-α treatment and the relationship between p11 and 5-HTR1b/4 in a human neuroblastoma cell line. Our study revealed that IFN-α treatment decreased p11 protein expression. 5-HTR1b and 5-HTR4 were also significantly declined after IFN-α treatment. Over-expression of p11 is sufficient to prevent the IFN-α -induced down-regulation of 5-HTR1b and 5-HTR4.

IFN-α remarkably
The protein levels of p11, 5-HTR1b and 5-HTR4 in the cytomembrane of SH-sy5y cells were significantly reduced after IFN-α treatment. We further investigated the effects of IFN-α treatment on the protein levels of p11, 5-HTR1b and 5-HTR4 in the cytomembrane of SH-sy5y cells. SH-sy5y cells were treated with 1000 IU/mL of hIFN-α -2b or equal volume of PBS. At 24 h, the cells were collected, and then membranal proteins were extracted for western blot analysis to determine the p11 and 5-HTR1b/4 protein levels. The results showed that there were no tublin protein bands in membrane protein samples (Fig. 8A). It was indicated that there was no cross-contamination between membranal and cytoplasmic protein extracts. We found that the protein levels of p11, 5-HTR1b and 5-HTR4 In parallel, similar changes in the 5-HTR1b and 5-HTR4 protein levels were observed in the p11-pcDNA3.0 groups or in the p11-miRNA groups. After hIFN-α -2b treatment, the 5-HTR1b or 5-HTR4 protein levels in the p11-pcDNA3.0 groups increased by 141.9% (P = 0.0081, F = 4.324) or 148.9% (P = 0.0064, F = 4.618), respectively, compared with the controls (Fig. 9C,D). On the contrary, even if in the absence of hIFN-α -2b, the 5-HTR1b or 5-HTR4 protein levels in the p11-miRNA groups were reduced by 41.6% (P = 0.0302, F = 3.684) or 49.3% (P = 0.0224, F = 6.372), respectively, compared with the controls (Fig. 9G,H). Such results indicated that 5-HTR1b and 5-HTR4 expressions were stringently correlated with p11, and more importantly over-expression of p11 was sufficient to prevent the IFN-α -induced inhibition on 5-HTR1b and 5-HTR4.
In addition, when mRNA levels of 5-HTR1b and 5-HTR4 were assayed using real-time PCR, the results showed that neither 5-HTR1b nor 5-HTR4 mRNA levels in the p11-pcDNA3.0 groups had significant difference compared with the controls (Fig. 9I). Similar results were observed in the p11-miRNA groups compared with the controls (Fig. 9J).
The statistical power analysis showed the powers are greater than 90% for each experiment of all, indicating that the number of subjects met the demand of statistic analysis.

Discussion
It is not uncommon for IFN-α treatment to induce the onset of a severe mental disorder, particularly major depression 5 . The mechanisms underlying this process remain unclear, but several hypotheses suggest that IFN-α -induced depression may be related to the effects of IFN-α on tryptophan metabolism, the synthesis of 5-hydroxytryptamine, and/or the activity of serotonin reuptake receptors 6,31,32 . However, previous studies are still controversial in their explanations of the molecular mechanisms responsible for IFN-α -induced depression. It has been reported that the deficiency of p11 plays key roles in the onset of depression via impairing the functions of its downstream chaperonins 5-HTR1b/4 17,33,34 . In the present study, we therefore determined whether the deficiency of p11 was involved in IFN-α -induced depression, whether p11 altered the levels of 5-HTR1b and 5-HTR4, and whether the regulation effects of IFN-α on levels of 5-HTR1b and 5-HTR4 were dependent upon p11.
First of all, we assayed depression-like behavior in mice after IFN-α injection using forced swimming tests and tail suspension tests. We found treatment with IFN-α could increase the immobility of the Balb/c mice in FST and TST (Fig. 1A,B). The results indicated that IFN-α treatment could cause depression-like behavior in mice. Animals' responses in physiology and behavioral psychology to IFN-α treatment are different, it may be due to the source of IFN-α , administration routes, the dosage, injection duration and genetic backgrounds. It is reported that, when administrated with murine IFN-α (60,000 IU/kg for 8 consecutive days, intraperitoneal injection), the mice exhibit subtle behavioral changes in the FST, but not TST 35 . hIFN-α -2b (400-1600 IU/g/day for 5-15 days, subcutaneously) significantly increases the immobility in the FST 36 . A more recent study finds that treatment with murine IFN-α (250 IU/day) for 14 days results in a reproducible depression-like state that can be characterized by increased immobility of C57BL6/J mice in TST and FST 37 . Therefore, we evaluated the depression-associated behavior of Balb/c mice that were treated with hIFN-α -2b (1000IU/g/day, intraperitoneal injection) for 10 days. In addition, in this study we chose human instead of murine IFN-α to treat mice, based on the consideration that clinical side effects were the result of human IFN-α treatment, though their functions in mice could be different 38,39 . Studies have shown that p11, 5-HTR1b and 5-HTR4 proteins in certain areas of rodent brains (e.g., hippocampus 27,34,40 , cortex 21,41 , amygdala 21,42 and nucleus raphe 29 ) play key roles in the onset of depression. Using IHC, we evaluated these protein levels in mice brains after IFN-α treatment. We found that continuous intraperitoneal injections of IFN-α into Balb/c mice led to lower protein levels of p11, 5-HTR1b and 5-HTR4 in the hippocampus or cingulate gyrus (Fig. 2), and the results were further confirmed with western blots (Fig. 3). Both the hippocampus and cingulate gyrus belong to the limbic system, which is involved in emotion formation and processing, learning, and memory. The combination of these functions makes the hippocampus and cingulate gyrus highly influential in linking behavioral outcomes to motivation 43,44 , and the roles make them highly important in mental disorders, particularly depression 45,46 . The hippocampus and cingulate gyrus contain abundant serotonergic nerve fibers expressing 5-HTR1b and 5-HTR4 47 . It is found that 5-HTR1b and 5-HTR4 can co-localize with p11 in hippocampus and play crucial roles in anti-depression effects in mice treated with SSRIs 48 . Results from our study and previous ones indicated that lower levels of these proteins in hippocampus or cingulate gyrus could be critical in IFN-α -induced depression. In addition, the western blot data suggested some more dramatic hippocampal down-regulation of p11 and 5-HTR1b/4 protein levels by IFN-α treatment than the IHC quantification (Figs 2 and 3). This was probably due to the fact that western blotting was more sensitive than IHC in the proteins detection 49 .
Next, we investigated the effects of IFN-α on the expressions of p11 and 5-HTR1b/4 in SH-sy5y cell line. Our tests showed that the p11, 5-HTR1b and 5-HTR4 protein-but not mRNA-levels were significantly reduced in a dose-and time-dependent manner after treatment with IFN-α (Figs 4-7). Furthermore, we also found that the levels of p11 and 5-HTR1b/4 proteins in cytomembrane were remarkably declined after IFN-α treatment (Fig. 8). Several studies have shown that 5-HTR1b and 5-HTR4 proteins play key roles in the anti-depression effects associated with p11 18,50 . Both 5-HTR1b and 5-HTR4 are G protein-coupled transmembrane receptors that produce different functions in response to 5-HT 51 . The functions in both the peripheral and central nervous systems are to modulate the release of various neurotransmitters 29,52 . It was not difficult to speculate that the decreased 5-HTR1b/4 protein levels in the membrane of neural synapses after IFN-α treatment would affect the functions in serotonergic neurons. We noted a recovery of protein levels for 5-HTR1b and 5-HTR4 at 48 h after IFN-α treatment (Fig. 7B,C), and such single dose curves shown in Fig. 7B,C were commonly seen in other researches 53,54 . Generally, a single dose treatment tended to affect the expression in protein level or mRNA level in a certain period of time and the effect would decrease over time. The protein levels of 5-HTR1b and 5-HTR4 should be fully recovered after the IFN-α was degraded. We did not test for a longer period of time because of the limitations of the cell culture procedure-the medium had to be changed after 48 h. On the other hand, the IFN-α is usually clinically administered on a daily basis 55 .
To investigate whether the down-regulating effects of IFN-α on 5-HTR1b or 5-HTR4 were mediated by p11, we analyzed 5-HTR1b or 5-HTR4 protein levels under the condition of higher or lower levels of p11 protein in SH-sy5y cells. We found that the 5-HTR1b and 5-HTR4 protein levels greatly increased in the p11-pcDNA3.0 transfection group (Fig. 9A-D), and they significantly declined in the p11-miRNA transfection group (Fig. 9E-H), compared with the controls. The data indicated that the 5-HTR1b and 5-HTR4 protein levels were positively correlated with the p11 protein levels even if treated with IFN-α . Such results indicated that the down-regulating effects of IFN-α on 5-HTR1b and 5-HTR4 were controlled by p11.
Although Svenningsson and Warner-Schmidt had reported that p11 can alter the distributions of 5-HTR1b/4 in cells by transferring them from the cytoplasm to the cytomembrane and thus increasing the density of 5-HTR1b/4 in the cytomembrane, it is still unknown whether p11 has the ability to modulate 5-HTR1b/4 protein levels 17,18 . Our studies suggested that p11 could affect 5-HTR1b/4 protein levels. In addition, we found decreased levels of p11 protein and depressive behaviors in mice treated with IFN-α , indicating a correlation between lower levels of p11 and depression-like behaviors. Several studies have demonstrated that both 5-HTR1b and 5-HTR4 are critical for the initiation of depression. In the pre-synapses, 5-HTR1b acts as an autoreceptor to inhibit the release of serotonin 56 . Cai et al. demonstrate that the activation of 5-HTR1b in post-synapse facilitates excitatory synaptic transmission, which is associated with depression 25 . It has been suggested that 5-HT1b antagonists may be effective adjunctive therapies for depression 57 . 5-HTR4 was originally identified as a mediator of 5-HT. Further studies indicated that the activation of 5-HTR4 increases the adenylate cyclase activity in mouse    59 . These evidences demonstrated that lower levels of 5-HTR1b or 5-HTR4 protein might cause the onset of depression.
We found that IFN-α administration resulted in depression-like behavior in mice and inhibited the protein levels of p11, 5-HTR1b and 5-HTR4 in SH-sy5y cells and the brains of mice. Furthermore, our results also suggested that the reduction in the protein levels of 5-HTR1b and 5-HTR4 were dependent upon p11 after IFN-α treatment. Together, these results illustrated that p11, 5-HTR1b, and 5-HTR4 play key roles in IFN-α -induced depression. These receptors might be involved in the mechanism underlying the inhibitory effects of IFN-α on the protein levels of p11 in brain areas (e.g., hippocampus, cingulate gyrus, and other regions) related to depression. Lower p11 levels probably led to a decline in the protein levels of 5-HTR1b and 5-HTR4 in local nerve synapses, subsequently disturbing the transmission of neurotransmitters in the synapses, and ultimately causing the body to experience depression. Also, there might be other molecules besides p11 involved in IFN-α -induced depression, which we did not examine. Further experiments using p11 knockout mice might help elucidate this issue. If the mice developed depression deferentially after IFN-α injection compared to the wild-type mice, IFN-α -induced depression might involve additional mechanisms besides p11. It is noteworthy that only a subsection of the individuals become depressed following IFN-α treatment, which could be explained by the variation of genetic makeup of an individual [60][61][62] . Further studies would allow us to clarify the molecular mechanism of IFN-α -induced depression.
Furthermore, p11 can be a potential biomarker in depression, as studies suggest that p11 mRNA levels in peripheral blood mononuclear cells correlate with suicide risk in mental disorders 33 and can distinguish post-traumatic stress disorder from bipolar disorder or major depression 63 . p11 is considered to be an up-stream regulator in the translocation and signal transduction of 5-HTR1b and 5-HTR4 17,18,29 . Our study found that IFN-α dramatically down-regulated p11 protein levels in a dose-and time-dependent manner, and that p11 controlled 5-HTR1b/4 protein levels. We suspected that p11 protein could be a potential biomarker in monitoring IFN-α -induced depression. In addition, we found IFN-α reduced p11 protein but not mRNA levels in cells, which were obviously different from the p11 mRNA studies in other types of depression 33,63 . Our findings illustrated that p11 protein-but not mRNA-levels could be a potential biomarker in IFN-α -induced depression, although clinical investigation is needed

Methods
Animals and drug treatments. The Balb/c mice were purchased from Vital River Laboratory Animal Technology Company (Beijing), and housed in specific-pathogen-free conditions. This study was approved by and performed in accordance with guidelines established by the Xinxiang Medical University Committee on the Use and Care of Animals. Adult male wild-type Balb/c mice were used in the animal tests. All mice were maintained with a 12-h light/dark cycle, and food and water were provided ad libitum. For drug treatments, three or four mice were housed per cage and hIFN-α -2b was injected into mice for 10 uninterrupted days (1000 IU/g/day, intraperitoneal injection, total volume = 0.2 mL) 36 . hIFN-α -2b was purchased from Sigma and diluted with PBS. Mice were injected with equal volumes and frequencies of PBS as the controls. The mice were injected at 9 a.m. each day.
Mice behavior tests. FST were carried out 1-2 days after the last IFN-α treatment on mice according to the method described in literature 64 . A video tracking system (Yishu, China) equipped with two 4 × 4 photo beam arrays allowing for the monitoring of swimming behaviors was used. Two swim sessions were conducted: the day before the start of the test, the mice were placed into the water to acclimatize them to FST. For testing, mice were individually and gently placed in a glass cylinder (30 cm high, 10 cm in diameter) filled with water at 24-26 °C to a depth of 18 cm, and allowed to swim for 6 min. The data of immobility within 2-6 min were collected to measure the behaviors.
TST was performed one day after the last IFN-α treatment on mice. Briefly, mice were suspended in the testing equipment (Bioseb, USA) by taping the distal part of the tail (1-1.5 cm) to a flat metallic surface 40 cm above the floor. Escape movements were recorded for 6 min. The time spent in an immobile posture within 2-6 min was measured as an index of depression-like behavior.
Immunohistochemistry. Fifty hours after the last IFN-α injection, the mice were anesthetized with ether. The brains were perfused with 4% ice-cold paraformaldehyde and treated with a series of processes to achieve dehydration, clearing, and wax immersion. The tissues were cut in a coronal plane for 5 μ m-thick sections, which included areas of the hippocampus and cingulate gyrus. The sections were subjected to a series of processes for dewaxing, rehydration, inactivation of endogenous enzymes (immersion in 3% H 2 O 2 for 30 min at room temperature), and antigen repair (immersion in citrate buffer at 99 °C for 30 min, followed by cooling to room temperature). The immune staining steps used streptavidin-biotin complex (SABC) immunohistochemistry kits (Biosail, China). First, the sections were blocked with 5% BSA for 30 min at 37 °C, followed by incubation with the primary antibody (anti-p11 1:300, Proteintech; anti-5-HTR1b 1:500, Abcam; anti-5-HTR4 1:500, Abcam; 1/1000 rabbit IgG as a control, Biosail, China). Second, after three washes with PBS, the sections were incubated with the secondary antibody (biotinylated goat anti-rabbit IgG, 1:200) at 37 °C for 30 min. Third, after three washes with PBS, the sections were incubated with 1/100 SABC reagents (Biosail, China) at 37 °C for 30 min. Fourth, the sections were stained with hematoxylin (Solarbio, USA) using a DAB chromogenic kit (Bostail, China). Finally, the sections were dried, sealed with a slide cover, and observed using an optical microscope (Olympus BX53, Japan).
Preparation of hippocampus proteins. Fifty hours after the last IFN-α injection, 5 mice in each group were anesthetized with ether, and the brains were removed and dissected on ice-cold PBS. Hippocampi were removed and pulverized completely in liquid nitrogen. The samples were mixed with 10 folds volume (μ L/μ g) of lysis buffer (Beyotime, Beijing) that contained 1% Triton X-100, 1% deoxycholate, and 0.1% sodium dodecyl sulfate (SDS), and supplemented with protease inhibitor cocktail (Abcam, USA). After being centrifuged at 12,000 rpm for 5 min at 4 °C, the supernatants were collected and stored at − 80 °C for western blotting analysis.
Cell culture. Human neuroblastoma cell line (SH-sy5y cell line, Jiahe Biotech Company, Shanghai, China) was cultured in Dulbecco's modified Eagle medium (DMEM) containing 10% fetal bovine serum, 50 U/mL penicillin, and 50 mg/mL streptomycin at 37 °C, with 5% CO 2 . All of the experiments were performed when the cells were 80-90% confluent.
Plasmid transfection and drug treatment. SH-sy5y cells were seeded at a density of 3 × 10 5 cells per well in six-well plates and incubated in DMEM at 37 °C for 24 h. Before transfection, the cells were washed with antibiotic-free DMEM containing 10% fetal bovine serum. The plasmids were transfected into cells with Liposome Transfast 2000 (Invitrogen, USA). The transfected plasmids contained p11-pcDNA3.0, pcDNA3.0 (control), P11-miRNA, and miRNA-control (control). At 5 h after transfection, the cells were washed with PBS and supplemented with complete DMEM. After incubation for another 24 h, 1000 IU/ mL of hIFNα -2b or PBS was added to the medium. The cells were collected at 48 h and the total proteins were extracted using RIPA reagents (Beyotime, China).
Cell membrane proteins extraction. SH-sy5y cells were treated with 1000IU/mL hIFN-α -2b and harvested at 24 h. Cytomembrane proteins were extracted with the Mem-PER eukaryotic membrane protein extraction kit (Thermo Fisher Scientific, USA) according to the manufacturer's recommendations 66 . Anti-pan-cadherin (Abcam, USA) and anti-tublin (Abcam, USA) antibodies were applied to test the levels and the purity of membranal proteins 67 .
Western blots. Sample preparation and electrophoresis. Cells were collected with trypsin (EPET, Biofluids) at specific time points. After three washes with cold PBS, the cells were re-suspended in RIPA lysis buffer, which contained 1% triton X-100, 1% deoxycholate, and 0.1% SDS (Beyotime, Beijing). The samples were vortexed and kept on ice for 30 min, followed by centrifugation (12000 rpm, 4 °C) for 5 min. The supernatants were collected in fresh tubes. The protein concentrations was measured using a BCA assay (Pierce, USA) according to the manufacturer's instructions 68 .
Scientific RepoRts | 6:17029 | DOI: 10.1038/srep17029 Immunofluorescence. The cells were plated onto sheet glasses for the immunofluorescence staining tests. At specific time points, the cells were washed with cold BSA-TBS (1 g/L BSA, 2.42 g/L Tris-base, 8 g/L NaCl, pH 7.6) and then fixed using − 20 °C pre-cooled 70% ethanol at room temperature for 10 min. The sections were washed twice with TBS-T (0.25% Triton X-100, Sigma) for 15 min each time. Subsequently, the samples were blocked using 5% BSA (diluted with TBS-T) for 1 h at room temperature and then incubated overnight with the primary antibody (anti-p11 1:300, Proteintech; 1/1000 rabbit IgG as negative control, Beyotime, China) at 4 °C. After five washes with TBS-T, the sections were incubated with an FITC-labeled secondary antibody (goat anti-rabbit IgG-FITC 1:2000, Santa Cruz) for 1 h at room temperature, followed by three washes with TBS-T. Next, the cells were stained with DAPI (4,6-diamidino-2-phenylindole, Invitrogen) for 20 min and sealed with 20 μ L antifade mounting medium. Finally, the sections were observed using a laser confocal microscope (Olympus FV1000, Japan).

Statistical analysis.
The P values in all experiments were determined with Prism software (GraphPad Software Inc.) for nonparametric unpaired T tests. F tests were used to compare variances. Regression analyses were performed with SPSS19.0. We considered P-values < 0.05 to be significant and the degree of significance is indicated as follows: *P < 0.05, **P < 0.01. In addition, to ensure the number of subjects meet the demand of statistics, the statistic power was calculated (alpha = 0.05, two tails).