HK2 in microglia and macrophages contribute to the development of neuropathic pain.

Neuropathic pain is a complex pain condition accompanied by prominent neuroinflammation involving activation of both central and peripheral immune cells. Metabolic switch to glycolysis is an important feature of activated immune cells. Hexokinase 2 (HK2), a key glycolytic enzyme enriched in microglia, has recently been shown important in regulating microglial functions. Whether and how HK2 is involved in neuropathic pain-related neuroinflammation remains unknown. Using a HK2-tdTomato reporter line, we found that HK2 was prominently elevated in spinal microglia. Pharmacological inhibition of HK2 effectively alleviated nerve injury-induced acute mechanical pain. However, selective ablation of Hk2 in microglia reduced microgliosis in the spinal dorsal horn (SDH) with little analgesic effects. Further analyses showed that nerve injury also significantly induced HK2 expression in dorsal root ganglion (DRG) macrophages. Deletion of Hk2 in myeloid cells, including both DRG macrophages and spinal microglia, led to the alleviation of mechanical pain during the first week after injury, along with attenuated microgliosis in the ipsilateral SDH, macrophage proliferation in DRGs, and suppressed inflammatory responses in DRGs. These data suggest that HK2 plays an important role in regulating neuropathic pain-related immune cell responses at acute phase and that HK2 contributes to neuropathic pain onset primarily through peripheral monocytes and DRG macrophages rather than spinal microglia.

Crosstalk between Interleukin-1 Receptor-Like 1 and Transforming Growth Factor-ß Receptor Signaling Promotes Renal Fibrosis

IL-33, a member of the IL-1 family, acts as an alarmin in immune response. Epithelial-mesenchymal transition and transforming growth factor-ß (TGF-ß)­induced fibroblast activation are key events in the development of renal interstitial fibrosis. The current study found increased expression of IL-33 and interleukin-1 receptor-like 1 (IL1RL1, alias ST2), the receptor for IL-33, in human fibrotic renal tissues. In addition, IL-33­ or ST2-deficient mice showed significantly reduced levels of fibronectin, α-smooth muscle actin, and vimentin, and increased E-cadherin levels. In HK-2 cells, IL-33 promotes the phosphorylation of the TGF-ß receptor (TGF-ßR), Smad2, and Smad3, and the production of extracellular matrix (ECM), with reduced expression of E-cadherin. Blocking TGF-ßR signaling or suppressing ST2 expression impeded Smad2 and Smad3 phosphorylation, thereby reducing ECM production, suggesting that IL-33­induced ECM synthesis requires cooperation between the two pathways. Mechanistically, IL-33 treatment induced a proximate interaction between ST2 and TGF-ßRs, activating downstream Smad2 and Smad3 for ECM production in renal epithelial cells. Collectively, this study identified a novel and essential role for IL-33 in promoting TGF-ß signaling and ECM production in the development of renal fibrosis. Therefore, targeting IL-33/ST2 signaling may be an effective therapeutic strategy for renal fibrosis.

Microglia modulate general anesthesia through P2Y12 receptor.

General anesthesia (GA) is an unconscious state produced by anesthetic drugs, which act on neurons to cause overall suppression of neuronal activity in the brain. Recent studies have revealed that GA also substantially enhances the dynamics of microglia, the primary brain immune cells, with increased process motility and territory surveillance. However, whether microglia are actively involved in GA modulation remains unknown. Here, we report a previously unrecognized role for microglia engaging in multiple GA processes. We found that microglial ablation reduced the sensitivity of mice to anesthetics and substantially shortened duration of loss of righting reflex (LORR) or unconsciousness induced by multiple anesthetics, thereby promoting earlier emergence from GA. Microglial repopulation restored the regular anesthetic recovery, and chemogenetic activation of microglia prolonged the duration of LORR. In addition, anesthesia-accompanying analgesia and hypothermia were also attenuated after microglial depletion. Single-cell RNA sequencing analyses showed that anesthesia prominently affected the transcriptional levels of chemotaxis and migration-related genes in microglia. By pharmacologically targeting different microglial motility pathways, we found that blocking P2Y12 receptor (P2Y12R) reduced the duration of LORR of mice. Moreover, genetic ablation of P2Y12R in microglia also promoted quicker recovery in mice from anesthesia, verifying the importance of microglial P2Y12R in anesthetic regulation. Our work presents the first evidence that microglia actively participate in multiple processes of GA through P2Y12R-mediated signaling and expands the non-immune roles of microglia in the brain.

Microglia and macrophages contribute to the development and maintenance of sciatica in lumbar disc herniation.

ABSTRACT: Lumbar disc herniation (LDH) is a major cause of sciatica. Emerging evidence indicated that inflammation induced by the herniated nucleus pulposus (NP) tissues plays a major role in the pathogenesis of sciatica. However, the underlying mechanisms are still elusive. Although microglia and macrophages have been implicated in nerve injury-induced neuropathic pain, their roles in LDH-induced sciatica largely remain unknown. This study successfully established and modified a mouse model of LDH. We found that nerve root compression using degenerated NP tissues can initiate remarkable and persistent sciatica, with increased and prolonged macrophage infiltration in dorsal root ganglia (DRG) and significant activation of microglia in the spinal dorsal horn. Instead, compression of the nerve root with nondegenerated NP tissues only led to transient sciatica, with transient infiltration and activation of macrophages and microglia. Moreover, continuous treatment of PLX5622, a specific colony-stimulating factor 1 receptor antagonist, ablated both macrophages and microglia, which effectively alleviated LDH-induced sciatica. However, mechanical allodynia reoccurred along with the repopulation of macrophages and microglia after the withdrawal of PLX5622. Using RNA sequencing analysis, the current study depicted transcriptional profile changes of DRG after LDH and identified several macrophage-related potential target candidates. Our results suggested that microglia and macrophages may play an essential role in the development and maintenance of LDH-induced sciatica. Targeting microglia and macrophages may be a promising treatment for chronic LDH-induced sciatica.

Dual roles of hexokinase 2 in shaping microglial function by gating glycolytic flux and mitochondrial activity.

Microglia continuously survey the brain parenchyma and actively shift status following stimulation. These processes demand a unique bioenergetic programme; however, little is known about the metabolic determinants in microglia. By mining large datasets and generating transgenic tools, here we show that hexokinase 2 (HK2), the most active isozyme associated with mitochondrial membrane, is selectively expressed in microglia in the brain. Genetic ablation of HK2 reduced microglial glycolytic flux and energy production, suppressed microglial repopulation, and attenuated microglial surveillance and damage-triggered migration in male mice. HK2 elevation is prominent in immune-challenged or disease-associated microglia. In ischaemic stroke models, however, HK2 deletion promoted neuroinflammation and potentiated cerebral damages. The enhanced inflammatory responses after HK2 ablation in microglia are associated with aberrant mitochondrial function and reactive oxygen species accumulation. Our study demonstrates that HK2 gates both glycolytic flux and mitochondrial activity to shape microglial functions, changes of which contribute to metabolic abnormalities and maladaptive inflammation in brain diseases.

mTOR-neuropeptide Y signaling sensitizes nociceptors to drive neuropathic pain.

Neuropathic pain is a refractory condition that involves de novo protein synthesis in the nociceptive pathway. The mTOR is a master regulator of protein translation; however, mechanisms underlying its role in neuropathic pain remain elusive. Using the spared nerve injury-induced neuropathic pain model, we found that mTOR was preferentially activated in large-diameter dorsal root ganglion (DRG) neurons and spinal microglia. However, selective ablation of mTOR in DRG neurons, rather than microglia, alleviated acute neuropathic pain in mice. We show that injury-induced mTOR activation promoted the transcriptional induction of neuropeptide Y (Npy), likely via signal transducer and activator of transcription 3 phosphorylation. NPY further acted primarily on Y2 receptors (Y2R) to enhance neuronal excitability. Peripheral replenishment of NPY reversed pain alleviation upon mTOR removal, whereas Y2R antagonists prevented pain restoration. Our findings reveal an unexpected link between mTOR and NPY/Y2R in promoting nociceptor sensitization and neuropathic pain.

Analyzing the Noise Robustness of Deep Neural Networks.

Adversarial examples, generated by adding small but intentionally imperceptible perturbations to normal examples, can mislead deep neural networks (DNNs) to make incorrect predictions. Although much work has been done on both adversarial attack and defense, a fine-grained understanding of adversarial examples is still lacking. To address this issue, we present a visual analysis method to explain why adversarial examples are misclassified. The key is to compare and analyze the datapaths of both the adversarial and normal examples. A datapath is a group of critical neurons along with their connections. We formulate the datapath extraction as a subset selection problem and solve it by constructing and training a neural network. A multi-level visualization consisting of a network-level visualization of data flows, a layer-level visualization of feature maps, and a neuron-level visualization of learned features, has been designed to help investigate how datapaths of adversarial and normal examples diverge and merge in the prediction process. A quantitative evaluation and a case study were conducted to demonstrate the promise of our method to explain the misclassification of adversarial examples.

mTOR-mediated metabolic reprogramming shapes distinct microglia functions in response to lipopolysaccharide and ATP.

Microglia constantly survey the brain microenvironment and rapidly adopt different phenotypes in response to environmental stimuli. Such dynamic functions require a unique metabolism and bioenergetics. However, little is known about the basic metabolism of microglia and how metabolic changes regulate microglia function. Here, we uncover that microglia activation is accompanied by extensive transcriptional changes in glucose and lipid metabolism-related genes. Using metabolic flux assays, we found that LPS, a prototype of the pathogen-associated molecular patterns (PAMPs), significantly enhanced glycolysis but suppressed oxidative phosphorylation (OXPHOS) in primary cultured microglia. By contrast, ATP, a known damage-associated molecular pattern (DAMPs) that triggers sterile activation of microglia, boosted both glycolysis and OXPHOS. Importantly, both LPS and ATP activated the mechanistic target of rapamycin (mTOR) pathway and enhanced the intracellular reactive oxygen species (ROS). Inhibition of mTOR activity suppressed glycolysis and ROS production in both conditions but exerted different effects on OXPHOS: it attenuated the ATP-induced elevation of OXPHOS, yet had no impact on the LPS-induced suppression of OXPHOS. Further, inhibition of mTOR or glycolysis decreased production of LPS-induced proinflammatory cytokines and ATP-induced tumor necrosis factor-α (TNF-α) and brain derived neurotrophic factor (BDNF) in microglia. Our study reveals a critical role for mTOR in the regulation of metabolic programming of microglia to shape their distinct functions under different states and shed light on the potential application of targeting metabolism to interfere with microglia-mediated neuroinflammation in multiple disorders.

Sphingosine kinase 2 cooperating with Fyn promotes kidney fibroblast activation and fibrosis via STAT3 and AKT.

Sphingosine kinases (Sphks) are the rate-limiting enzymes in the conversion of sphingosine to biologically active sphingosine-1-phosphate. The present study aimed to determine the role of Sphk2 and its downstream targets in renal fibroblast activation and interstitial fibrosis. In the kidney interstitium of patients with renal fibrosis, Sphk2high-expressing cells (mainly interstitial fibroblasts) were significantly elevated and highly correlated with disease progression in patients. In a murine model of renal interstitial fibrosis, Sphk2 was upregulated in the kidney of wild-type mice in response to disease progression. Importantly, Sphk2-knockout (KO) mice exhibited significantly lower levels of extracellular matrix (ECM) production and a suppressed inflammatory response in the kidney tissues, compared to those in their wild-type counterparts, whereas the expression of TGF-ß1 was unaffected. TGF-ß1 effectively upregulated Sphk2 expression in the renal interstitial fibroblast line, NRK-49F, independent of canonical Smad signaling activation. Furthermore, siRNA-mediated Sphk2 knockdown or suppression of Sphk2 activity by ABC294640 exposure effectively attenuated AKT and STAT3 activation and ECM production, but had no effects on Smad2 and Smad3 activation. Sphk2 phosphorylated Fyn to activate downstream STAT3 and AKT, thereby promoting ECM synthesis. Therefore, our findings indicate that targeting Sphk2-Fyn-STAT3/AKT signaling pathway may be a novel therapeutic approach for renal fibrosis.

IL-33/ST2 plays a critical role in endothelial cell activation and microglia-mediated neuroinflammation modulation.

BACKGROUND: Interleukin-33 (IL-33) is increasingly being recognized as a key immunomodulatory cytokine in many neurological diseases. METHODS: In the present study, wild-type (WT) and IL-33-/- mice received intracerebroventricular (i.c.v.) injection of lipopolysaccharide (LPS) to induce neuroinflammation. Intravital microscopy was employed to examine leukocyte-endothelial interactions in the brain vasculature. The degree of neutrophil infiltration was determined by myeloperoxidase (MPO) staining. Real-time PCR and western blotting were used to detect endothelial activation. Enzyme-linked immunosorbent assay and quantitative PCR were conducted to detect pro-inflammatory cytokine levels in the brain. RESULTS: In IL-33-/- mice, neutrophil infiltration in the brain cortex and leukocyte-endothelial cell interactions in the cerebral microvessels were significantly decreased as compared to WT mice after LPS injection. In addition, IL-33-/- mice showed reduced activation of microglia and cerebral endothelial cells. In vitro results indicated that IL-33 directly activated cerebral endothelial cells and promoted pro-inflammatory cytokine production in LPS-stimulated microglia. CONCLUSIONS: Our study indicated that IL-33/ST2 signaling plays an important role in the activation of microglia and cerebral endothelial cells and, therefore, is essential in leukocyte recruitment in brain inflammation. The role of IL-33/ST2 in LPS induced neuroinflammation.

FTY720 attenuates behavioral deficits in a murine model of systemic lupus erythematosus.

Neuropsychiatric (NP) involvement in systemic lupus erythematosus (SLE) severely impacts patients' quality of life and leads to a poor prognosis. The current therapeutic protocol, corticosteroid administration, can also induce neuropsychiatric disorders. FTY720 is an immunomodulator that selectively confines lymphocytes in lymph nodes and reduces autoreactive T cell recruitment to the central nervous system (CNS). This study aimed to identify a novel therapeutic strategy for NPSLE. B6.MRL-lpr mice were treated with oral administration of FTY720 (2 mg/kg) three times per week for 12 weeks, to evaluate its efficacy in a model of NPSLE. FTY720 significantly attenuated the impulsive and depression-like behavior of B6.MRL-lpr mice. Neuronal damage was reduced in the cortex, hippocampus, and amygdala of the FTY720-treated B6.MRL-lpr mice, as well as in TNF-α-treated HT22 cells. Additionally, FTY720 downregulated levels of inflammatory cytokines, and reduced the infiltration of T cells and neutrophils in the brain parenchyma. FTY720 also acted directly on cerebral endothelial cells and reduced the permeability of the blood-brain barrier (BBB) in B6.MRL-lpr mice, as evidenced by reduced central IgG and albumin levels. Finally, FTY720 significantly inhibited activation of PI3K/Akt/GSK3ß/p65 signaling, which further reduced the expression levels of adhesion molecules in bEND.3 cells treated with B6.MRL-lpr mouse serum. Collectively, our data indicate that oral administration of FTY720 at an early stage has beneficial effects in NPSLE-model B6.MRL-lpr mice, suggesting that it may represent an effective new therapeutic strategy for NPSLE.

Fingolimod targets cerebral endothelial activation to block leukocyte recruitment in the central nervous system.

Fingolimod (FTY720), an immunomodulator, is approved as an oral treatment for patients with relapsing forms of multiple sclerosis. Its effects are largely attributed to its mechanism of selectively retaining lymphocytes in the lymph nodes to reduce autoreactive T-cell recruitment in the CNS. In this study, we investigated the therapeutic effect of FTY720 on an animal model of CNS inflammation induced by intracerebral ventricle LPS injection. We found that FTY720 treatment significantly prevented LPS-induced neutrophil recruitment in the CNS by inhibiting leukocyte recruitment in cerebral microvessels. Furthermore, FTY720 also inhibited the expressions of adhesion molecules on the cerebral endothelium, but did not affect the expression levels of pro-inflammatory cytokines (TNF-α and IL-6) and chemokines (CXCL1 and CXCL2) in the CNS parenchyma. The inhibition of endothelial activation was accompanied by reduced phosphorylation of signaling molecules, including serine/threonine-specific protein kinase (Akt), STAT6, and nuclear factor-κB. This FTY720-attenuated inhibition of leukocyte recruitment and endothelial activation was reversed by blocking the functions of sphingosine kinase 2 or sphingosine-1-phosphate receptor 1. Our study demonstrated, for the first time, that FTY720 directly inhibits the phosphorylation of multiple signaling molecules in endothelial cells, thereby effectively blocking leukocyte recruitment in the CNS.

Analyzing the Training Processes of Deep Generative Models.

Among the many types of deep models, deep generative models (DGMs) provide a solution to the important problem of unsupervised and semi-supervised learning. However, training DGMs requires more skill, experience, and know-how because their training is more complex than other types of deep models such as convolutional neural networks (CNNs). We develop a visual analytics approach for better understanding and diagnosing the training process of a DGM. To help experts understand the overall training process, we first extract a large amount of time series data that represents training dynamics (e.g., activation changes over time). A blue-noise polyline sampling scheme is then introduced to select time series samples, which can both preserve outliers and reduce visual clutter. To further investigate the root cause of a failed training process, we propose a credit assignment algorithm that indicates how other neurons contribute to the output of the neuron causing the training failure. Two case studies are conducted with machine learning experts to demonstrate how our approach helps understand and diagnose the training processes of DGMs. We also show how our approach can be directly used to analyze other types of deep models, such as CNNs.

Online Visual Analytics of Text Streams.

We present an online visual analytics approach to helping users explore and understand hierarchical topic evolution in high-volume text streams. The key idea behind this approach is to identify representative topics in incoming documents and align them with the existing representative topics that they immediately follow (in time). To this end, we learn a set of streaming tree cuts from topic trees based on user-selected focus nodes. A dynamic Bayesian network model has been developed to derive the tree cuts in the incoming topic trees to balance the fitness of each tree cut and the smoothness between adjacent tree cuts. By connecting the corresponding topics at different times, we are able to provide an overview of the evolving hierarchical topics. A sedimentation-based visualization has been designed to enable the interactive analysis of streaming text data from global patterns to local details. We evaluated our method on real-world datasets and the results are generally favorable.