Inferring Time-Lagged Causality Using the Derivative of Single-Cell Expression

Many computational methods have been developed to infer causality among genes using cross-sectional gene expression data, such as single-cell RNA sequencing (scRNA-seq) data. However, due to the limitations of scRNA-seq technologies, time-lagged causal relationships may be missed by existing methods. In this work, we propose a method, called causal inference with time-lagged information (CITL), to infer time-lagged causal relationships from scRNA-seq data by assessing the conditional independence between the changing and current expression levels of genes. CITL estimates the changing expression levels of genes by “RNA velocity”. We demonstrate the accuracy and stability of CITL for inferring time-lagged causality on simulation data against other leading approaches. We have applied CITL to real scRNA data and inferred 878 pairs of time-lagged causal relationships. Furthermore, we showed that the number of regulatory relationships identified by CITL was significantly more than that expected by chance. We provide an R package and a command-line tool of CITL for different usage scenarios.     

Amino Acid-Mediated Intracellular Ca2+ Rise Modulates mTORC1 by Regulating the TSC2-Rheb Axis through Ca2+/Calmodulin

Mechanistic target of rapamycin complex 1 (mTORC1) is a master growth regulator by controlling protein synthesis and autophagy in response to environmental cues. Amino acids, especially leucine and arginine, are known to be important activators of mTORC1 and to promote lysosomal translocation of mTORC1, where mTORC1 is thought to make contact with its activator Rheb GTPase. Although amino acids are believed to exclusively regulate lysosomal translocation of mTORC1 by Rag GTPases, how amino acids increase mTORC1 activity besides regulation of mTORC1 subcellular localization remains largely unclear. Here we report that amino acids also converge on regulation of the TSC2-Rheb GTPase axis via Ca²⁺/calmodulin (CaM). We showed that the amino acid-mediated increase of intracellular Ca²⁺ is important for mTORC1 activation and thereby contributes to the promotion of nascent protein synthesis. We found that Ca²⁺/CaM interacted with TSC2 at its GTPase activating protein (GAP) domain and that a CaM inhibitor reduced binding of CaM with TSC2. The inhibitory effect of a CaM inhibitor on mTORC1 activity was prevented by loss of TSC2 or by an active mutant of Rheb GTPase, suggesting that a CaM inhibitor acts through the TSC2-Rheb axis to inhibit mTORC1 activity. Taken together, in response to amino acids, Ca²⁺/CaM-mediated regulation of the TSC2-Rheb axis contributes to proper mTORC1 activation, in addition to the well-known lysosomal translocation of mTORC1 by Rag GTPases.     

IMGG: Integrating Multiple Single-Cell Datasets through Connected Graphs and Generative Adversarial Networks

There is a strong need to eliminate batch-specific differences when integrating single-cell RNA-sequencing (scRNA-seq) datasets generated under different experimental conditions for downstream task analysis. Existing batch correction methods usually transform different batches of cells into one preselected “anchor” batch or a low-dimensional embedding space, and cannot take full advantage of useful information from multiple sources. We present a novel framework, called IMGG, i.e., integrating multiple single-cell datasets through connected graphs and generative adversarial networks (GAN) to eliminate nonbiological differences between different batches. Compared with current methods, IMGG shows excellent performance on a variety of evaluation metrics, and the IMGG-corrected gene expression data incorporate features from multiple batches, allowing for downstream tasks such as differential gene expression analysis.     

The Characterization of a Subependymal Giant Astrocytoma-Like Cell Line from Murine Astrocyte with mTORC1 Hyperactivation

Tuberous sclerosis complex (TSC) is a genetic disorder caused by inactivating mutations in TSC1 (hamartin) or TSC2 (tuberin), crucial negative regulators of the mechanistic target of rapamycin complex 1 (mTORC1) signaling pathway. TSC affects multiple organs including the brain. The neurologic manifestation is characterized by cortical tubers, subependymal nodules (SEN), and subependymal giant cell astrocytoma (SEGA) in brain. SEGAs may result in hydrocephalus in TSC patients and mTORC1 inhibitors are the current recommended therapy for SEGA. Nevertheless, a major limitation in the research for SEGA is the lack of cell lines or animal models for mechanistic investigations and development of novel therapy. In this study, we generated TSC1-deficient neural cells from spontaneously immortalized mouse astrocytes in an attempt to mimic human SEGA. The TSC1-deficient cells exhibit mTORC1 hyperactivation and characteristics of transition from astrocytes to neural stem/progenitor cell phenotypes. Rapamycin efficiently decreased mTORC1 activity of these TSC1-deficient cells in vitro. In vivo, TSC1-deficient cells could form SEGA-like tumors and Rapamycin treatment decreased tumor growth. Collectively, our study generates a novel SEGA-like cell line that is invaluable for studying mTORC1-driven molecular and pathological alterations in neurologic tissue. These SEGA-like cells also provide opportunities for the development of novel therapeutic strategy for TSC patients with SEGA.     

Identification of an AMPK Phosphorylation Site in Drosophila TSC2 (gigas) that Regulate Cell Growth

AMP-activated protein kinase (AMPK) is an important metabolic regulator that mediates cellular adaptation to diverse stresses. One of the AMPK substrates, tuberous sclerosis complex 2 (TSC2), was suggested to mediate AMPK-induced silencing of mTOR complex 1 (mTORC1) signaling that is critical for cell growth. However, it is not known whether the AMPK-dependent TSC2 phosphorylation, originally observed in mammalian cells, is conserved in invertebrates. Here we show that energy depletion inhibits mTORC1 signaling through the AMPK-TSC2 axis in Drosophila S2 cells. We have discovered an AMPK phosphorylation site in TSC2-like genes from many different invertebrate species including Drosophila. The site (Ser1338 in Drosophila TSC2) is specifically and efficiently phosphorylated by AMPK in vitro. To evaluate the functional role of this phosphorylation site in vivo, we generated transgenic flies that can express identical amount of either wild-type or phosphorylation-resistant mutant Drosophila TSC2 in a tissue-specific manner. In response to transgenic Sestrin induction, which causes ectopic AMPK activation and subsequent mTORC1 inhibition, wild-type Drosophila TSC2 synergistically reduced tissue growth in the dorsal epithelium of Drosophila wings. However, phosphorylation-resistant mutant Drosophila TSC2 was unable to show such a growth-inhibiting effect, suggesting that this phosphorylation is important for AMPK-dependent regulation of cell growth.     

Single Cell Gene Expression Analysis in a 3D Microtissue Liver Model Reveals Cell Type-Specific Responses to Pro-Fibrotic TGF-β1 Stimulation

3D cell culture systems are widely used to study disease mechanisms and therapeutic interventions. Multicellular liver microtissues (MTs) comprising HepaRG, hTERT-HSC and THP-1 maintain multicellular interactions and physiological properties required to mimic liver fibrosis. However, the inherent complexity of multicellular 3D-systems often hinders the discrimination of cell type specific responses. Here, we aimed at applying single cell sequencing (scRNA-seq) to discern the molecular responses of cells involved in the development of fibrosis elicited by TGF-β1. To obtain single cell suspensions from the MTs, an enzymatic dissociation method was optimized. Isolated cells showed good viability, could be re-plated and cultured in 2D, and expressed specific markers determined by scRNA-seq, qRT-PCR, ELISA and immunostaining. The three cell populations were successfully clustered using supervised and unsupervised methods based on scRNA-seq data. TGF-β1 led to a fibrotic phenotype in the MTs, detected as decreased albumin and increased αSMA expression. Cell-type specific responses to the treatment were identified for each of the three cell types. They included HepaRG damage characterized by a decrease in cellular metabolism, prototypical inflammatory responses in THP-1s and extracellular matrix remodeling in hTERT-HSCs. Furthermore, we identified novel cell-specific putative fibrosis markers in hTERT-HSC (COL15A1), and THP-1 (ALOX5AP and LAPTM5).     

Cell Type-Specific Induction of Inflammation-Associated Genes in Crohn’s Disease and Colorectal Cancer

Based on the rapid increase in incidence of inflammatory bowel disease (IBD), the identification of susceptibility genes and cell populations contributing to this condition is essential. Previous studies suggested multiple genes associated with the susceptibility of IBD; however, due to the analysis of whole-tissue samples, the contribution of individual cell populations remains widely unresolved. Single-cell RNA sequencing (scRNA-seq) provides the opportunity to identify underlying cellular populations. We determined the enrichment of Crohn’s disease (CD)-induced genes in a publicly available Crohn’s disease scRNA-seq dataset and detected the strongest induction of these genes in innate lymphoid cells (ILC1), highly activated T cells and dendritic cells, pericytes and activated fibroblasts, as well as epithelial cells. Notably, these genes were highly enriched in IBD-associated neoplasia, as well as sporadic colorectal cancer (CRC). Indeed, the same six cell populations displayed an upregulation of CD-induced genes in a CRC scRNA-seq dataset. Finally, after integrating and harmonizing the CD and CRC scRNA-seq data, we demonstrated that these six cell types display a gradual increase in gene expression levels from a healthy state to an inflammatory and tumorous state. Together, we identified cell populations that specifically upregulate CD-induced genes in CD and CRC patients and could, therefore, contribute to inflammation-associated tumor development.     

CrossTORC and WNTegration in Disease: Focus on Lymphangioleiomyomatosis

The mechanistic target of rapamycin (mTOR) and wingless-related integration site (Wnt) signal transduction networks are evolutionarily conserved mammalian growth and cellular development networks. Most cells express many of the proteins in both pathways, and this review will briefly describe only the key proteins and their intra- and extracellular crosstalk. These complex interactions will be discussed in relation to cancer development, drug resistance, and stem cell exhaustion. This review will also highlight the tumor-suppressive tuberous sclerosis complex (TSC) mutated, mTOR-hyperactive lung disease of women, lymphangioleiomyomatosis (LAM). We will summarize recent advances in the targeting of these pathways by monotherapy or combination therapy, as well as future potential treatments.     

Quantitative Analysis of the Potency of Equimolar Two-Drug Combinations and Combi-Molecules Involving Kinase Inhibitors In Vitro: The Concept of Balanced Targeting

The median-effect principle proposed by Chou and Talalay is the most effective approach to parameterize interactions between several agents in combination. However, this method cannot be used to evaluate the effectiveness of equimolar drug combinations, which are comparative references for dual-targeting molecular design. Here, using data acquired through the development of “combi-molecules” blocking two kinases (e.g., EGFR-c-Src and EGFR-c-Met), we established potency indices for equimolar and dual-targeted inhibitors. If the fold difference (κ) between the IC50 of the two individual kinase inhibitors was >6, the IC50 of their equimolar combination resembled that of the more potent inhibitor. Hence, the “combi-targeting” of the two kinases was considered “imbalanced” and the combination ineffective. However, if κ ≤ 6, the IC50 of the combination fell below that of each individual drug and the combi-targeting was considered “balanced” and the combination effective. We also showed that combi-molecules should be compared with equimolar combinations only under balanced conditions and propose a new parameter Ω for validating their effectiveness. A multi-targeted drug is effective if Ω < 1, where Ω is defined as the IC50 of the drug divided by that of the corresponding equimolar combination. Our study provides a methodology to determine the in vitro potency of equimolar two-drug combinations as well as combi-/hybrid molecules inhibiting two different kinase targets.     

Angelman Syndrome and Angelman-like Syndromes Share the Same Calcium-Related Gene Signatures

Angelman-like syndromes are a group of neurodevelopmental disorders that entail clinical presentation similar to Angelman Syndrome (AS). In our previous study, we showed that calcium signaling is disrupted in AS, and we identified calcium-target and calcium-regulating gene signatures that are able to differentiate between AS and their controls in different models. In the herein study, we evaluated these sets of calcium-target and calcium-regulating genes as signatures of AS-like and non-AS-like syndromes. We collected a number of RNA-seq datasets of various AS-like and non-AS-like syndromes and performed Principle Component Analysis (PCA) separately on the two sets of signature genes to visualize the distribution of samples on the PC1–PC2 plane. In addition to the evaluation of calcium signature genes, we performed differential gene expression analyses to identify calcium-related genes dysregulated in each of the studied syndromes. These analyses showed that the calcium-target and calcium-regulating signatures differentiate well between AS-like syndromes and their controls. However, in spite of the fact that many of the non-AS-like syndromes have multiple differentially expressed calcium-related genes, the calcium signatures were not efficient classifiers for non-AS-like neurodevelopmental disorders. These results show that features based on clinical presentation are reflected in signatures derived from bioinformatics analyses and suggest the use of bioinformatics as a tool for classification.     

Lipid Cubic Mesophases Combined with Superparamagnetic Iron Oxide Nanoparticles: A Hybrid Multifunctional Platform with Tunable Magnetic Properties for Nanomedical Applications

Hybrid materials composed of superparamagnetic iron oxide nanoparticles (SPIONs) and lipid self-assemblies possess considerable applicative potential in the biomedical field, specifically, for drug/nutrient delivery. Recently, we showed that SPIONs-doped lipid cubic liquid crystals undergo a cubic-to-hexagonal phase transition under the action of temperature or of an alternating magnetic field (AMF). This transition triggers the release of drugs embedded in the lipid scaffold or in the water channels. In this contribution, we address this phenomenon in depth, to fully elucidate the structural details and optimize the design of hybrid multifunctional carriers for drug delivery. Combining small-angle X-ray scattering (SAXS) with a magnetic characterization, we find that, in bulk lipid cubic phases, the cubic-to-hexagonal transition determines the magnetic response of SPIONs. We then extend the investigation from bulk liquid-crystalline phases to colloidal dispersions, i.e., to lipid/SPIONs nanoparticles with cubic internal structure (“magnetocubosomes”). Through Synchrotron SAXS, we monitor the structural response of magnetocubosomes while exposed to an AMF: the magnetic energy, converted into heat by SPIONs, activates the cubic-to-hexagonal transition, and can thus be used as a remote stimulus to spike drug release “on-demand”. In addition, we show that the AMF-induced phase transition in magnetocubosomes steers the realignment of SPIONs into linear string assemblies and connect this effect with the change in their magnetic properties, observed at the bulk level. Finally, we assess the internalization ability and cytotoxicity of magnetocubosomes in vitro on HT29 adenocarcinoma cancer cells, in order to test the applicability of these smart carriers in drug delivery applications.     

Is Multiple System Atrophy a Prion-like Disorder?

Multiple system atrophy (MSA) is a rapidly progressive, fatal neurodegenerative disease of uncertain aetiology that belongs to the family of α-synucleinopathies. It clinically presents with parkinsonism, cerebellar, autonomic, and motor impairment in variable combinations. Pathological hallmarks are fibrillary α-synuclein (αSyn)-rich glial cytoplasmic inclusions (GCIs) mainly involving oligodendroglia and to a lesser extent neurons, inducing a multisystem neurodegeneration, glial activation, and widespread demyelinization. The neuronal αSyn pathology of MSA has molecular properties different from Lewy bodies in Parkinson’s disease (PD), both of which could serve as a pool of αSyn (prion) seeds that could initiate and drive the pathogenesis of synucleinopathies. The molecular cascade leading to the “prion-like” transfer of “strains” of aggregated αSyn contributing to the progression of the disease is poorly understood, while some presented evidence that MSA is a prion disease. However, this hypothesis is difficult to reconcile with postmortem analysis of human brains and the fact that MSA-like pathology was induced by intracerebral inoculation of human MSA brain homogenates only in homozygous mutant 53T mice, without production of disease-specific GCIs, or with replication of MSA prions in primary astrocyte cultures from transgenic mice expressing human αSyn. Whereas recent intrastriatal injection of Lewy body-derived or synthetic human αSyn fibrils induced PD-like pathology including neuronal αSyn aggregates in macaques, no such transmission of αSyn pathology in non-human primates by MSA brain lysate has been reported until now. Given the similarities between αSyn and prions, there is a considerable debate whether they should be referred to as “prions”, “prion-like”, “prionoids”, or something else. Here, the findings supporting the proposed nature of αSyn as a prion and its self-propagation through seeding as well as the transmissibility of neurodegenerative disorders are discussed. The proof of disease causation rests on the concordance of scientific evidence, none of which has provided convincing evidence for the classification of MSA as a prion disease or its human transmission until now.     

Aversive Learning Deficits and Depressive-Like Behaviors Are Accompanied by an Increase in Oxidative Stress in a Rat Model of Fetal Alcohol Spectrum Disorders: The Protective Effect of Rapamycin

Fetal alcohol spectrum disorders (FASDs) are one of the most common consequences of ethanol exposure during pregnancy. In adulthood, these disorders can be manifested by learning and memory deficits and depressive-like behavior. Ethanol-induced oxidative stress may be one of the factors that induces FASD development. The mammalian target of the Rapamycin (mTOR) signaling pathway that acts via two distinct multiprotein complexes, mTORC1 and mTORC2, can affect oxidative stress. We investigated whether mTOR-dependent or mTOR-independent mechanisms are engaged in this phenomenon. Thus, Rapamycin—a selective inhibitor of mTORC1, Torin-2—a non-selective mTORC1/mTORC2 inhibitor, and FK-506—a drug that impacts oxidative stress in an mTOR-independent manner were used. Behavioral tests were performed in adult (PND60-65) rats using a passive avoidance (PA) task (aversive learning and memory) and forced swimming test (FST) (depressive-like behaviors). In addition, the biochemical parameters of oxidative stress, such as lipid peroxidation (LPO), as well as apurinic/apyrimidinic (AP)-sites were determined in the hippocampus and prefrontal cortex in adult (PND65) rats. The rat FASD model was induced by intragastric ethanol (5 g/kg/day) administration at postnatal day (PND)4–9 (an equivalent to the third trimester of human pregnancy). All substances (3 mg/kg) were given 30 min before ethanol. Our results show that neonatal ethanol exposure leads to deficits in context-dependent fear learning and depressive-like behavior in adult rats that were associated with increased oxidative stress parameters in the hippocampus and prefrontal cortex. Because these effects were completely reversed by Rapamycin, an mTORC1 inhibitor, this outcome suggests its usefulness as a preventive therapy in disorders connected with prenatal ethanol exposure.     

Resilience in Long-Term Viral Infection: Genetic Determinants and Interactions

Virus-induced neurological sequelae resulting from infection by Theiler’s murine encephalomyelitis virus (TMEV) are used for studying human conditions ranging from epileptic seizures to demyelinating disease. Mouse strains are typically considered susceptible or resistant to TMEV infection based on viral persistence and extreme phenotypes, such as demyelination. We have identified a broader spectrum of phenotypic outcomes by infecting strains of the genetically diverse Collaborative Cross (CC) mouse resource. We evaluated the chronic-infection gene expression profiles of hippocampi and thoracic spinal cords for 19 CC strains in relation to phenotypic severity and TMEV persistence. Strains were clustered based on similar phenotypic profiles and TMEV levels at 90 days post-infection, and we categorized distinct TMEV response profiles. The three most common profiles included “resistant” and “susceptible,” as before, as well as a “resilient” TMEV response group which experienced both TMEV persistence and mild neurological phenotypes even at 90 days post-infection. Each profile had a distinct gene expression signature, allowing the identification of pathways and networks specific to each TMEV response group. CC founder haplotypes for genes involved in these pathways/networks revealed candidate response-specific alleles. These alleles demonstrated pleiotropy and epigenetic (miRNA) regulation in long-term TMEV infection, with particular relevance for resilient mouse strains.     

Making the Bend: DNA Tertiary Structure and Protein-DNA Interactions

DNA structure functions as an overlapping code to the DNA sequence. Rapid progress in understanding the role of DNA structure in gene regulation, DNA damage recognition and genome stability has been made. The three dimensional structure of both proteins and DNA plays a crucial role for their specific interaction, and proteins can recognise the chemical signature of DNA sequence (“base readout”) as well as the intrinsic DNA structure (“shape recognition”). These recognition mechanisms do not exist in isolation but, depending on the individual interaction partners, are combined to various extents. Driving force for the interaction between protein and DNA remain the unique thermodynamics of each individual DNA-protein pair. In this review we focus on the structures and conformations adopted by DNA, both influenced by and influencing the specific interaction with the corresponding protein binding partner, as well as their underlying thermodynamics.