HSP60 critically regulates endogenous IL-1β production in activated microglia by stimulating NLRP3 inflammasome pathway

All the animal experiments were performed after obtaining approval from the Institutional Animal Ethics Committee of the National Brain Research Centre (NBRC) (NBRC/IAEC/2016/115 and NBRC/IAEC/2017/028). For in vivo experiments, postnatal day 8–10 (P08-P10) BALB/c mice were used, irrespective of their sex. The animals were handled in strict accordance with good animal practice as per the guidelines of the Committee for the Purpose of Control and Supervision of Experiments on Animals, Ministry of Environment and Forestry, Government of India.

IL-1β was injected intraperitoneally (i.p.) at a dose of 10 ng/g of body weight of P10 BALB/c mice pups after every 24 h for different durations (1, 2, and 3 days) as described earlier [13]. Control mice group received intraperitoneal injection of equal volume of PBS.

Vivo-morpholino are morpholino oligos coupled with eight guanidinium head groups on dendrimer scaffold that enable delivery into cells [31]. Morpholino oligomers are proven antisense molecules used for the specific knockdown of the gene of interest both in vitro and in vivo. It either blocks the mRNA translation or interferes with RNA processing, including splicing and mRNA maturation [32]. HSP60 vivo-morpholino (HSP60-Mo) oligos were commercially procured from Gene Tools LLC (Philomath, OR, USA). HSP60-Mo was designed against sequences of mouse HSP60 (HSPD1) gene to specifically target it (5′ ACT GTG GGT AGT CGA TTT CT 3′). A 25-base scrambled morpholino of random sequence (SC-Mo) was used as a negative control (5′ TGG TTT CAG AAT AGT ATT CCA CTG C 3′).

For in vivo IL-1β experiments, animals were divided into six groups: (i) Control, (ii) IL-1β treatment, (iii) Sc-Mo, (iv) Sc-Mo + IL-1β treatment, (v) HSP60-Mo, and (vi) HSP60-Mo + IL-1β treatment group. Each group had a minimum of three animals. Among these, groups (v) and (vi) were intracranially injected with HSP60 vivo-morpholino at P8 (15 mg/kg of body weight of mice), while groups (iii) and (iv) received intracranial injection of scrambled vivo-morpholino at P8 (15 mg/kg of body weight of mice). As efficiency of vivo-morpholino in crossing the blood brain barrier is quite low, therefore, to achieve significant knockdown in the brain, our laboratory devised a slightly different strategy based on a previously published method [33, 34]. The intracranial injection was given manually in 8-day-old BALB/c mice pups (P8) at a single site as vivo-morpholino is believed to diffuse in the tissue [35, 36]. The amount of vivo-morpholino was calculated according to the body weight of each mice and calculated volume of vivo-morpholino was made up to 25 μl using 1× PBS. Then this 25 μl of vivo-morpholino solution was taken in the insulin syringe with needle size 31 G × 15/64 (0.25 × 6 mm), and it was slowly injected by gently pushing the syringe piston. Groups (i) and (ii) received an intracranial injection of PBS at P8 (same volume as vivo-morpholino). At P10, IL-1β was injected intraperitoneally (i.p.) in groups (ii), (iv), and (vi) at a dose of 10 ng/g of body weight of mice pups dissolved in 50 μl PBS, for three consecutive days. Groups (i), (iii), and (v) received same volume of PBS intraperitoneally. The mice were then sacrificed by repeated transcardial perfusion of chilled PBS and their brains were collected for protein and/or RNA analysis. The efficacy of the morpholino injection and its efficiency to knockdown HSP60 was checked through Western blot which was done by random sampling for morpholino-treated group. After we observed specific knockdown of HSP60 by vivo-morpholino (Additional file 1: Figure S1(A)), then only we proceeded for further experiments using vivo-morpholino with following four groups: (i) Control, (ii) IL-1β, (iii) HSP60-Mo, and (iv) HSP60-Mo + IL-1β groups.

All the in vitro experiments were performed in N9 murine microglial cells (N9 cells), which were a kind gift from Prof. Maria Pedroso de Lima (Center for Neuroscience and Cell Biology, University of Coimbra, Portugal) and were grown as described earlier [10]. N9 cells were chosen for the study as these microglial cells were derived from mouse brains and share many phenotypic characteristics with primary mouse microglia [37]. Transfection of HSP60 plasmid and endonuclease-prepared short interfering RNA (esiRNA) against mouse HSP60 gene was performed in N9 cells as described earlier for overexpression and knockdown experiments [10]. For the overexpression studies, 4 μg recombinant mouse HSP60 plasmid (MC206740, Origene) was used (Additional file 1: Figure S2), while 5 pM HSP60 eSiRNA (EMU151751, Sigma Aldrich) was used for knockdown experiments.

To induce inflammation, the N9 cells were serum starved for 2 h at 70% confluency and treated with 5 ng/ml IL-1β for different time periods. The cells were then used for different assays. For Western blotting, caspase-1 assay, and enzyme-linked immunosorbent assay, 1.5 × 10⁶ cells were seeded in 90 mm × 20 mm plates, while for quantitative real-time PCR and flow cytometric analysis (reactive oxygen species analysis, cytokine bead array, and rhodamine 123 assays), 6 × 10⁵ cells were seeded in 60 mm × 15 mm plates.

Viral suspensions were prepared from the mice brain using the GP78 strain of JEV as described previously [38]. P10 BALB/c mouse pups were divided into six groups: (i) Control, (ii) JEV- infected, (iii) only Sc-Mo, (iv) Sc-Mo + JEV, (v) only HSP60-Mo, and (vi) HSP60-Mo + JEV group, and each group had a minimum of three pups. The HSP60-Mo group and HSP60-Mo + JEV infection group received an intracranial injection of HSP60-Mo at P8 (15 mg/kg of body weight of mice), while Sc-Mo and Sc-Mo + JEV groups were intracranially injected with scrambled vivo-morpholino (15 mg/kg of body weight of mice). Control and only JEV-infected groups received intracranial injection of PBS (same volume as vivo-morpholino) at P8. Mice from JEV group, Sc-Mo + JEV group, and HSP60-Mo + JEV group were injected with 1.5 × 10³ plaque-forming units (PFU) of virus in 50 μl PBS, whereas control group, Sc-Mo group, and HSP60-Mo group were given same volume of PBS, intraperitoneally. After the development of full symptoms (including tremors, ruffled fur, hunching, ataxia, gait abnormalities like hind limb paralysis and body stiffening), the animals were sacrificed and their brains were excised after repeated transcardial perfusion with ice-cold PBS. The animal brains were then used for either protein or RNA analysis. Knockdown of HSP60 by vivo-morpholino was confirmed at protein levels by Western blotting (Additional file 1: Figure S1(B)). After confirming specific knockdown of HSP60 in JEV-infected group by HSP60 Mo, we proceeded with following four groups for further experiments: (i) Control, (ii) JEV-infected, (iii) only HSP60-Mo, and (iv) HSP60-Mo + JEV group.

For the JEV infection of N9 cells, around 1.5 × 10⁶ cells were seeded in 90 mm × 20 mm plates in 5% RPMI and were allowed to grow for 12–15 h. After the cells reached 70% confluence, they were serum starved for 2 h and infected with JEV (strain GP78) at an MOI (multiplicity of infection) of 2 followed by incubation at 37 °C for 24 h to induce inflammation. The MOI of 2 was chosen for JEV infection as it significantly induces inflammation as compared to low MOI (Additional file 1: Figure S3). The cells were then harvested to isolate RNA for quantitative real-time PCR and protein for cytokine bead array and Western blotting.

Fresh frozen paraffin-embedded (FFPE) human brain tissue sections were obtained from the Human Brain Tissue Repository, National Institute of Mental Health and Neurosciences, Bangalore, India, in accordance with the institutional scientific ethics, protecting the confidentiality of the subjects. These sections were obtained from the frontal cortex/hippocampus postmortem from at least two confirmed patients of different brain disorders. For the control experimental sets, brain tissues from individuals who succumbed to traffic accidents and had no known prior neurological disease were used. The human glioma brain tissues were kindly provided by Dr. Ellora Sen (NBRC).

The 5-μm-thick FFPE frontal cortex sections were deparaffinized by repeated incubation in xylene followed by washing in alcohol gradient. Age-matched control samples were obtained from accidental cases with least possible trauma to brain. The RNA isolation was performed from human FFPE sections, human glioma brain tissue, N9 cells, and mouse brains using Tri reagent (Sigma-Aldrich), and cDNA was synthesized using an Advantage RT-for-PCR kit (Clontech Laboratories) as per manufacturer’s protocol. qRT-PCR was carried out as described previously [10] from 500 ng RNA, using primers specific for mouse IL-1β, HSP60, and NLRP3 genes. The conditions used for the qRT-PCR were as follows: 95 °C for 3 min (1 cycle) and 94 °C for 20 s, 55 °C for 30 s, and 72 °C for 45 s (40 cycles). The relative mRNA abundance was determined by normalizing to GAPDH mRNA using the Pfaffl method [39]. To elucidate the changes in IL-1β and HSP60 transcript levels in different brain disorders, two different qRT-PCRs were carried out (for IL-1β and HSP60) for each neurological condition. The qRT-PCR was done in triplicates. The primer sequences used for qRT-PCR analysis are listed in Additional file 1: Table S1.

Western blotting was performed as previously described [10]. Around 3 × 10⁶ cells were pelleted and protein was isolated and quantified by the abovementioned protocol. For Western blotting of cytosolic as well as nuclear protein fractions, 30 μg protein was used. Primary antibodies against the following proteins were used: HSP60 (Abcam, #Ab46798), NLRP3 (Abcam, #Ab91525), inducible nitric oxide synthase (iNOS) (Abcam, #Ab3523), phospho-p65 nuclear factor-κB (NF-κB) p65 (S536) (Cell signaling Technology, #3033), Proliferating cell nuclear antigen (PCNA) (Cell Signaling Technology, #2586), Cycloxygenase-2 (COX2) (Millipore, #Ab5118), NF-κBp-65 (Santa Cruz Biotechnology, #SC372), and β-actin (Sigma Aldrich, #A3854). Secondary antibodies were labeled with horseradish peroxidase. The images were captured and analyzed using the Uvitec gel documentation system (Cambridge, UK) and ImageJ software respectively. Cytosolic protein levels were normalized to β-actin levels, whereas nuclear protein levels were normalized to PCNA. The fold change with respect to control cells was then calculated based on integrated density values (IDV).

For the quantitative analysis of various important cyto-chemokines in untreated and treated cells, the CBA kit (BD Biosciences, NJ, USA) was used. Around 1.5 × 10⁶ cells were pelleted down and protein was isolated and quantified. The beads coated with antibodies against interleukin 6 (IL-6), tumor necrosis factor alpha (TNF-α), and monocyte chemoattractant protein-1 (MCP-1) antibodies were mixed with 50 μg cell lysates and standards according to the manufacturer’s instructions. The experiments were performed in triplicates as described earlier [13]. A total of 10,000 events were acquired for each sample. The results were analyzed using FACS Calibur (Becton Dickinson) and CBA software that allows the calculation of cytokine concentrations in unknown lysates with the help of a standard curve.

The levels of ROS generated within N9 cells of each treatment groups were measured by the CM-H2DCFDA (5 (and 6)-chlromethyl-20,70-dichloro-dihydrofluoresceindiacetate) (Sigma Aldrich), which is a cell-permeable, non-polar, H₂O₂-sensitive probe. It diffuses into cells, where intracellular esterases cleave its acetate groups, releasing the corresponding dichlorodihydrofluorescein derivative which gives red fluorescence [30]. 6 × 10⁵ cells were seeded for the ROS analysis. After the completion of treatment, the untreated and treated N9 cells were incubated with 5 μM CM-H2DCFDA in dark at 37 °C room temperature for 20 min followed by washing and the relative mean fluorescence intensity was measured using FACS caliber (BD Biosciences, USA)). A total of 10,000 events were acquired in each treatment group.

The integrity of the mitochondrial membrane was estimated by Rhodamine 123 (Rh 123) assay as described earlier [40]. Rh 123 is a cationic green fluorescent dye that can enter the mitochondrial matrix and the variation in the accumulation of Rh 123 in the cells is directly related to the change in the mitochondrial electrochemical potential (Δψ_M). A decrease in the fluorescence of Rh 123 indicates a loss in the mitochondrial transmembrane potential and thus is a good method to identify mitochondrial damage. 6 × 10⁵ cells were seeded for the Rh 123 assay. After the completion of treatment, control and treated N9 cells were incubated with Rh 123 (0.3 μg/ml) for 20 min at 37 °C, followed by washing and resuspension in FACS buffer. A total of 10,000 events were acquired in each treatment group on a flow cytometer (BD FACS Calibur, BD Biosciences, USA) and the relative mean fluorescence intensity of Rh 123 was assessed. Staurosporine (1 μM)-treated N9 cells were used as positive control (data not shown).

Levels of active caspase-1 were analyzed using caspase-1 activity assay kit (Millipore, USA, #21870) as per the manufacturer’s protocol. Briefly, around 3 × 10⁶ cells were pelleted and resuspended for 10 min in 50 μl chilled lysis buffer followed by centrifugation at 10,000g at 4 °C for 1 min. The supernatant containing cell lysate was quantified using BCA method. Two hundred micrograms of the cell lysates were incubated with 50 μl of 2× reaction buffer and the substrate (YVAD-p-Nitroaniline, at a final concentration of 200 μM) at 37 °C for 2 h followed by measurement of absorbance at 405 nm to quantify caspase-1 activity levels in different treatment groups. This assay is based upon the spectrophotometric detection of the chromophore p-nitroaniline (p-NA) after cleavage from the YVAD-pNA substrate because of the activation of caspase-1.

To quantify the levels of secreted IL-1β from different groups of N9 cells, ELISA was performed using mouse IL-1β ELISA kit (Biolegend, #432604) as per the manufacturer’s recommendations. Briefly, a rat monoclonal anti-mouse IL-1β capture antibody was coated in the 96-well plate overnight, followed by blocking for 1 h at room temperature (RT) and washing. For in vitro experiments, 1.5 × 10⁶ cells were seeded in 90 mm × 20 mm culture plates and media was collected after the completion of treatments. For in vivo experiments, BALB/c brain lysates were used. Control and treated samples (100 μl media supernatant for in vitro and 100 μg protein from the mice brain lysates for in vivo experiments) were incubated in these wells overnight at 4 °C. The samples were then incubated with biotin conjugated detection antibody for 1 h at RT, followed by addition of avidin-HRP substrate for 30 min. The absorbance was measured at 450 nm on a spectrophotometer (Biorad, Australia), and the concentrations were calculated using the IL-1β standard reference curves.

Data are represented as the mean ± standard deviation (SD) from at least three independent experiments performed in triplicates (n = 3). The data was analyzed statistically by Student’s t test or one-way analysis of variance (ANOVA) followed by Holm-Sidak post hoc test. P value < 0.05 was considered significant. For in vivo treatments, a minimum of three mice were used in each group and experiments were repeated at least three times.

As IL-1β is considered to be the master regulator of the inflammation, its levels have been reported to be increased in various neurodegenerative disorders and brain infections as a result of microglial activation [9]. To confirm this, we compared the mRNA levels of IL-1β from sections of various human brain disorders including Alzheimer’s disease, Parkinson’s disease, stroke, rabies, tuberculous meningitis, cerebral malaria, toxoplasma encephalitis, and cryptococcus meningitis with control brain sections. For this, we performed qRT-PCR analysis from the FFPE human brain sections from the abovementioned neurological diseases and we found more than threefold increase in the levels of IL-1β in comparison to control sections (Fig. 1). In our previous study, we discovered that HSP60 plays a very important role in microglial inflammation by regulating underlying mechanism of IL-1β action. Therefore, we next determined the transcript levels of HSP60 in these diseased brain sections and found significant increase in HSP60 levels in almost all of these diseases when compared to control brain sections (Fig. 1). Similarly, levels of IL-1β and HSP60 get significantly elevated in glioma tissue in comparison to control (Fig. 1). Graph in Fig. 1 represents pooled data of all the qRT-PCR runs. This result signifies critical involvement of HSP60 in the pathogenesis of these neuronal disorders and neuronal infections besides IL-1β, and it might play an important role as an immunomodulatory molecule during neuronal infection and neurodegeneration.

IL-1β after binding with its cognate receptor IL-1R1 can induce its own production by stimulating NLRP3 inflammasome complex [7]. It can also induce the phosphorylation of NF-κB and its nuclear localization in various cell types, which can signal the formation of inflammasome complex [41, 42]. Phosphorylation of NF-κB acts as a probing signal for the activation of NLRP3 inflammasome pathway which is responsible for endogenous IL-1β production by activated microglia. However, whether HSP60 plays a role in this endogenous IL-1β production via inflammasome pathway in microglial cells is not known. Hence, we set out to determine the effects of HSP60 on inflammasome pathway activation.

For this, we first assessed the effect of IL-1β on NF-κB phosphorylation both in vitro and in vivo in the cytosolic extract. We found that IL-1β was able to significantly induce phosphorylation of p65-NF-κB both in vitro and in vivo (Fig. 2a, b). Next, we overexpressed HSP60 in N9 microglial cells and found that HSP60 overexpression was also able to induce p65-NF-κB phosphorylation in vitro (Fig. 2c). We then knocked-down HSP60 in N9 cells and treated cells with IL-1β for 3 h. To our surprise, IL-1β was not able to induce NF-κB phosphorylation after HSP60 reduction (Fig  2d). For in vivo knockdown of HSP60, mice were intracranially injected with HSP60-Mo. After the confirmation of specific knockdown of HSP60 by HSP60-Mo, the animals were divided into four groups and were treated with HSP60-Mo and IL-1β as described in the “Methods” section. Supporting our in vitro results, after reduction of HSP60 by HSP60-Mo, IL-1β was not able to induce phosphorylation of p65-NF-κB in vivo (Fig. 2e) also. This result confirms the crucial involvement of HSP60 in IL-1β-induced NF-κB phosphorylation.

Phosphorylation of p65-NF-κB leads to its nuclear localization which is necessary for its function, i.e., regulation of expression of inflammatory genes. Hence, we checked the nuclear localization of phosphorylated p65-NF-κB (p-p65-NF-κB) upon IL-1β treatment in N9 microglial cells as well as BALB/c mice brains. We found that IL-1β treatment not only increases the phosphorylation of p65-NF-κB, but also leads to increase in the nuclear localization of phosphorylated p65-NF-κB, both in vitro and in vivo (Fig. 3a, d respectively). Simultaneously, we assessed the effect of HSP60 overexpression on the same and our results show that HSP60 overexpression in N9 microglial cells leads to increased nuclear localization of pNF-κB (Fig. 3b). To determine the role of HSP60 in IL-1β-induced nuclear localization of p-p65-NF-κB, we knocked-down HSP60 in N9 cells followed by IL-1β treatment and found that, after knockdown of HSP60, there was decrease in the nuclear localization of p-p65-NF-κB (Fig. 3c).We specifically knocked-down HSP60 in BALB/c mice brain using HSP60 vivo-morpholino and treated with IL-1β after 48 h of morpholino treatment. Our results show that in vivo knockdown of HSP60 led to decrease in the nuclear localization of NF-κB even after IL-1β treatment (Fig. 3e). These results suggest that HSP60 plays a critical role in the IL-1β-induced nuclear localization of pNF-κB.

Nuclear localization of pNF-κB facilitates the activation of NLRP3 inflammasome pathway by inducing the transcription of NLRP3 gene and pro-IL-1β [41, 43]. We also observed that IL-1β induces the phosphorylation and nuclear localization of NF-κB in a HSP60-dependent manner (Figs. 2 and 3); therefore, we next explored the role of HSP60 in IL-1β-induced NLRP3 expression through qRT-PCR and Western blot. For this, we first assessed the effect of IL-1β on NLRP3 expression and we found that IL-1β treatment significantly increases mRNA and protein levels of NLRP3 both in vitro (Fig. 4a, f) and in vivo (Fig. 4d, i). To investigate the role of HSP60, we overexpressed HSP60 gene in N9 cells as described in the “Methods” section. HSP60 induced the expression of NLRP3 both at transcript and protein levels (Fig. 4b, g), and its downregulation reduced the NLRP3 expression even after IL-1β treatment (Fig. 4c, h). Similarly, in BALB/c mice brain samples, IL-1β treatment increases expression of NLRP3 (Fig. 4d, i); however, the expression of NLRP3 did not increase after IL-1β treatment in the HSP60 vivo-morpholino-treated group and were comparable to control group (Fig. 4e, j). These results show that HSP60 expression is critical for IL-1β-induced NLRP3 expression.

NLRP3 inflammasome is activated in response to a variety of stimuli, supporting the fact that different signals induce similar downstream events that are sensed by NLRP3. The widely studied mechanisms of NLRP3 activation include the mitochondrial damage leading to the decrease in mitochondrial membrane potential and generation of mitochondrial reactive oxygen species (ROS) [44]. To assess the effect of IL-1β treatment and HSP60 modulation on mitochondrial membrane potential, we performed Rhodamine 123 (Rh 123) assay. We observed that IL-1β treatment (for 3 h) as well as HSP60 overexpression led to decrease in the mitochondrial membrane potential in microglial cells, indicating the mitochondrial damage (Fig. 5a (i-ii)). Cells with HSP60 knockdown do not display mitochondrial damage as mitochondrial membrane potential was comparable to control cells even after IL-1β treatment (Fig. 5a (iii)).

Literature suggests that IL-1β increases ROS generation in microglia [45]. We also confirmed increase in ROS generation in N9 cells after IL-1β treatment (Fig. 5b (i)). We found that ROS generation in N9 cells increased up to 3.5-fold after 3 h of IL-1β treatment in comparison to untreated control cells. Further, to determine the effects of HSP60 on ROS, we overexpressed and knocked-down HSP60 in N9 cells. Overexpression of HSP60 hugely induces ROS generation (6.2-fold in comparison to control) (Fig. 5b (ii)) whereas its knockdown drastically reduces the effect of IL-1β on ROS generation (Fig. 5b (iii)) and ROS levels become comparable to that of control cells.

NLRP3 inflammasome complex, when activated in response to different cell damage and/or, stress stimuli, leads to the cleavage of pro-caspase-1 to caspase-1 which is also known as interleukin-converting enzyme (ICE). Formation of caspase-1 from pro-caspase-1 is the executioner step of inflammasome pathway which is responsible for the maturation of IL-1β from pro-IL-1β. We analyzed the levels of active caspase-1, both in vitro and in vivo. Our in vitro data show that both IL-1β treatment and HSP60 overexpression increased the activity of caspase-1 in N9 cells by 5.8 folds and 8.1 folds respectively (Fig. 6a (i-ii)). However, HSP60 knockdown does not allow increase in caspase-1 activity even after IL-1β treatment (Fig. 6a (iii)). Further, our in vivo results recapitulate the in vitro results. In in vivo conditions, IL-1β increases the levels of active caspase-1 via HSP60 as the knockdown of HSP60 reduces IL-1β-induced active caspase-1 levels (Fig. 6b (i) and (ii)). This result suggests that HSP60 plays an important role in caspase-1 activation.

To determine whether endogenous IL-1β production is mediated by HSP60, we finally checked the effect of HSP60 on endogenous IL-1β production in response to IL-1β treatment both in vitro (N9 cells) and in vivo (BALB/c mice brains). We assessed the levels of expression of IL-1β through qRT-PCR and its secretion by ELISA. We observed that IL-1β treatment and HSP60 overexpression increases IL-1β production and it gets secreted by microglial cells in vitro (Fig. 7a, b, f, g respectively). Knocking down HSP60 in N9 cells compromised the expression and secretion of IL-1β even after IL-1β treatment (Fig. 7c, h). Similarly, in BALB/c mice brains also, IL-1β induces its own production in vivo (Fig. 7d, i,). However, IL-1β treatment in mice brain preceded by HSP60 downregulation was unable to induce IL-1β production (Fig. 7e, j). These results show that HSP60 indeed plays a critical role in IL-1β inducing its own production by activated microglia via regulating NLRP3 inflammasome pathway.

JEV is a common cause of acute and epidemic viral encephalitis. JEV infection is associated with microglial activation resulting in the production of pro-inflammatory cytokines. As our data in the previous section show that HSP60 regulates IL-1β production (Fig. 7), hence, we were curious to explore whether it regulates IL-1β production during JEV infection also, which is a very good model to study neuroinflammation. We first determined levels of HSP60 in JEV-infected N9 cells, mice brains, and FFPE human brain sections through qRT-PCR and found that JEV infection was able to significantly increase the expression of HSP60 transcripts (Fig. 8a–c). Protein levels of HSP60 also significantly increased in JEV-infected N9 cells and mice brain as compared to control (Fig. 8d, e). Literature suggests that JEV infection induces IL-1β production by stimulating the NLRP3 inflammasome pathway [29, 30]. We tested this notion and confirmed induction of IL-1β in vitro and in vivo upon JEV infection through ELISA (Fig. 8f, g). Next, to explore the role of HSP60 in JEV-induced IL-1β production, we knocked-down HSP60 both in vitro (N9 cells) and in vivo (BALB/c mice brain) as described in the “Methods” section. To our surprise, knockdown of HSP60 was sufficient to reduce JEV infection-mediated IL-1β production (Fig. 8h, i). These results suggest that downregulation of HSP60 leads to alteration of inflammasome pathway which hampers JEV-induced IL-1β production by activated microglia.

HSP60 knockdown results in a decrease in IL-1β production after JEV infection both in vitro and in vivo (Fig. 8h, i) and as IL-1β is the main cytokine involved in microglial activation, we speculated that reducing HSP60 levels may also ameliorate JEV-induced inflammation. To test this, we assessed the levels of important pro-inflammatory enzymes (iNOS and COX2) by Western blotting (Fig. 9a, b) and performed cytometric bead array (CBA) for measuring the levels of pro-inflammatory cytokines (TNF-α, MCP-1, and IL-6) in N9 cells as well as BALB/c mice brains after JEV infection (Fig. 9c–h). We observed that downregulation of HSP60 both in vitro and in vivo leads to reduction of these pro-inflammatory markers after JEV infection.

As knocking down HSP60 reduced the inflammation in JEV-infected mice, so we questioned what would be the effect of HSP60 on survival of JEV-infected mice. We observed that in BALB/c mice knockdown of HSP60 not only reduced the level of inflammatory markers, but was able to significantly increase the survival of the infected animal also. Animals pretreated with HSP60 vivo-morpholino before JEV infection showed delayed onset of symptoms and the survival was significantly increased than those of JEV-infected group (more than 10 days after the death of JEV-infected mice) (Fig. 10a). In addition, the mice from JEV-infected groups showed behavioral deficits after the onset of symptoms (viz. tremor, hind limb paralysis, motor deficit) which were improved and delayed after knockdown of HSP60 (Fig. 10b). We compared the behavior of the HSP60-Mo + JEV-infected mice with only JEV-infected mice by giving scores based on the visible symptoms as shown in the graph. These results suggest HSP60 reduces the inflammation during JEV infection that leads to delayed infection and increased survival of the organism. Thus, our results illuminate HSP60 as a novel therapeutic target against JEV infection.