Distinct Processing of lncRNAs Contributes to Non-conserved Functions in Stem Cells.

Long noncoding RNAs (lncRNAs) evolve more rapidly than mRNAs. Whether conserved lncRNAs undergo conserved processing, localization, and function remains unexplored. We report differing subcellular localization of lncRNAs in human and mouse embryonic stem cells (ESCs). A significantly higher fraction of lncRNAs is localized in the cytoplasm of hESCs than in mESCs. This turns out to be important for hESC pluripotency. FAST is a positionally conserved lncRNA but is not conserved in its processing and localization. In hESCs, cytoplasm-localized hFAST binds to the WD40 domain of the E3 ubiquitin ligase ß-TrCP and blocks its interaction with phosphorylated ß-catenin to prevent degradation, leading to activated WNT signaling, required for pluripotency. In contrast, mFast is nuclear retained in mESCs, and its processing is suppressed by the splicing factor PPIE, which is highly expressed in mESCs but not hESCs. These findings reveal that lncRNA processing and localization are previously under-appreciated contributors to the rapid evolution of function.

Hexokinase 1 cellular localization regulates the metabolic fate of glucose.

The product of hexokinase (HK) enzymes, glucose-6-phosphate, can be metabolized through glycolysis or directed to alternative metabolic routes, such as the pentose phosphate pathway (PPP) to generate anabolic intermediates. HK1 contains an N-terminal mitochondrial binding domain (MBD), but its physiologic significance remains unclear. To elucidate the effect of HK1 mitochondrial dissociation on cellular metabolism, we generated mice lacking the HK1 MBD (ΔE1HK1). These mice produced a hyper-inflammatory response when challenged with lipopolysaccharide. Additionally, there was decreased glucose flux below the level of GAPDH and increased upstream flux through the PPP. The glycolytic block below GAPDH is mediated by the binding of cytosolic HK1 with S100A8/A9, resulting in GAPDH nitrosylation through iNOS. Additionally, human and mouse macrophages from conditions of low-grade inflammation, such as aging and diabetes, displayed increased cytosolic HK1 and reduced GAPDH activity. Our data indicate that HK1 mitochondrial binding alters glucose metabolism through regulation of GAPDH.

Taking Me away: the function of phosphorylation on histone lysine demethylases.

Histone lysine demethylases (KDMs) regulate eukaryotic gene transcription by catalysing the removal of methyl groups from histone proteins. These enzymes are intricately regulated by the kinase signalling system in response to internal and external stimuli. Here, we review the mechanisms by which kinase-mediated phosphorylation influence human histone KDM function. These include the changing of histone KDM subcellular localisation or chromatin binding, the altering of protein half-life, changes to histone KDM complex formation that result in histone demethylation, non-histone demethylation or demethylase-independent effects, and effects on histone KDM complex dissociation. We also explore the structural context of phospho-sites on histone KDMs and evaluate how this relates to function.

Proximity RNA Labeling by APEX-Seq Reveals the Organization of Translation Initiation Complexes and Repressive RNA Granules.

Diverse ribonucleoprotein complexes control mRNA processing, translation, and decay. Transcripts in these complexes localize to specific regions of the cell and can condense into non-membrane-bound structures such as stress granules. It has proven challenging to map the RNA composition of these large and dynamic structures, however. We therefore developed an RNA proximity labeling technique, APEX-seq, which uses the ascorbate peroxidase APEX2 to probe the spatial organization of the transcriptome. We show that APEX-seq can resolve the localization of RNAs within the cell and determine their enrichment or depletion near key RNA-binding proteins. Matching the spatial transcriptome, as revealed by APEX-seq, with the spatial proteome determined by APEX-mass spectrometry (APEX-MS), obtained precisely in parallel, provides new insights into the organization of translation initiation complexes on active mRNAs and unanticipated complexity in stress granule composition. Our novel technique allows a powerful and general approach to explore the spatial environment of macromolecules.

Localization of RNAs to the mitochondria-mechanisms and functions.

The mammalian mitochondrial proteome comprises over 1000 proteins, with the majority translated from nuclear-encoded messenger RNAs (mRNAs). Mounting evidence suggests many of these mRNAs are localized to the outer mitochondrial membrane (OMM) in a pre- or cotranslational state. Upon reaching the mitochondrial surface, these mRNAs are locally translated to produce proteins that are cotranslationally imported into mitochondria. Here, we summarize various mechanisms cells use to localize RNAs, including transfer RNAs (tRNAs), to the OMM and recent technological advancements in the field to study these processes. While most early studies in the field were carried out in yeast, recent studies reveal RNA localization to the OMM and their regulation in higher organisms. Various factors regulate this localization process, including RNA sequence elements, RNA-binding proteins (RBPs), cytoskeletal motors, and translation machinery. In this review, we also highlight the role of RNA structures and modifications in mitochondrial RNA localization and discuss how these features can alter the binding properties of RNAs. Finally, in addition to RNAs related to mitochondrial function, RNAs involved in other cellular processes can also localize to the OMM, including those implicated in the innate immune response and piRNA biogenesis. As impairment of messenger RNA (mRNA) localization and regulation compromise mitochondrial function, future studies will undoubtedly expand our understanding of how RNAs localize to the OMM and investigate the consequences of their mislocalization in disorders, particularly neurodegenerative diseases, muscular dystrophies, and cancers.

Interpretable feature extraction and dimensionality reduction in ESM2 for protein localization prediction.

As the application of large language models (LLMs) has broadened into the realm of biological predictions, leveraging their capacity for self-supervised learning to create feature representations of amino acid sequences, these models have set a new benchmark in tackling downstream challenges, such as subcellular localization. However, previous studies have primarily focused on either the structural design of models or differing strategies for fine-tuning, largely overlooking investigations into the nature of the features derived from LLMs. In this research, we propose different ESM2 representation extraction strategies, considering both the character type and position within the ESM2 input sequence. Using model dimensionality reduction, predictive analysis and interpretability techniques, we have illuminated potential associations between diverse feature types and specific subcellular localizations. Particularly, the prediction of Mitochondrion and Golgi apparatus prefer segments feature closer to the N-terminal, and phosphorylation site-based features could mirror phosphorylation properties. We also evaluate the prediction performance and interpretability robustness of Random Forest and Deep Neural Networks with varied feature inputs. This work offers novel insights into maximizing LLMs' utility, understanding their mechanisms, and extracting biological domain knowledge. Furthermore, we have made the code, feature extraction API, and all relevant materials available at https://github.com/yujuan-zhang/feature-representation-for-LLMs.

TransGCN: a semi-supervised graph convolution network-based framework to infer protein translocations in spatio-temporal proteomics.

Protein subcellular localization (PSL) is very important in order to understand its functions, and its movement between subcellular niches within cells plays fundamental roles in biological process regulation. Mass spectrometry-based spatio-temporal proteomics technologies can help provide new insights of protein translocation, but bring the challenge in identifying reliable protein translocation events due to the noise interference and insufficient data mining. We propose a semi-supervised graph convolution network (GCN)-based framework termed TransGCN that infers protein translocation events from spatio-temporal proteomics. Based on expanded multiple distance features and joint graph representations of proteins, TransGCN utilizes the semi-supervised GCN to enable effective knowledge transfer from proteins with known PSLs for predicting protein localization and translocation. Our results demonstrate that TransGCN outperforms current state-of-the-art methods in identifying protein translocations, especially in coping with batch effects. It also exhibited excellent predictive accuracy in PSL prediction. TransGCN is freely available on GitHub at https://github.com/XuejiangGuo/TransGCN.

NF-κB dynamics in the language of immune cells.

Biological discovery has been driven by advances in throughput and resolution of analysis technologies. They have also created an indelible bias for snapshot-based knowledge. Even though recent methods such as multi-omics single-cell assays have empowered immunological investigations, they still provide snapshots of cellular behaviors and thus, have inherent limitations in reconstructing unsynchronized dynamic events across individual cells. Here, we present a rationale for how NF-κB may convey specificity of contextual information through subtle quantitative features of its signaling dynamics. The next frontier of predictive understanding should involve functional characterization of NF-κB signaling dynamics and their immunological implications. This may help solve the apparent paradox that a ubiquitously activated transcription factor can shape accurate responses to different immune challenges.

BioID organelle mapping: you are the company you keep.

In a recent study, Go, Knight et al. combined a panel of protein markers with BioID proximity-dependent labeling to profile the composition of 20 distinct subcellular compartments. Comparison with similar global datasets acquired using imaging or fractionation-based approaches confirmed the consistency of the results while highlighting unique advantages.

Spatial organization of bacterial sphingolipid synthesis enzymes.

Sphingolipids are produced by nearly all eukaryotes where they play significant roles in cellular processes such as cell growth, division, programmed cell death, angiogenesis, and inflammation. While it was previously believed that sphingolipids were quite rare among bacteria, bioinformatic analysis of the recently identified bacterial sphingolipid synthesis genes suggests that these lipids are likely to be produced by a wide range of microbial species. The sphingolipid synthesis pathway consists of three critical enzymes. Serine palmitoyltransferase catalyzes the condensation of serine with palmitoyl-CoA (or palmitoyl-acyl carrier protein), ceramide synthase adds the second acyl chain, and a reductase reduces the ketone present on the long-chain base. While there is general agreement regarding the identity of these bacterial enzymes, the precise mechanism and order of chemical reactions for microbial sphingolipid synthesis is more ambiguous. Two mechanisms have been proposed. First, the synthesis pathway may follow the well characterized eukaryotic pathway in which the long-chain base is reduced prior to the addition of the second acyl chain. Alternatively, our previous work suggests that addition of the second acyl chain precedes the reduction of the long-chain base. To distinguish between these two models, we investigated the subcellular localization of these three key enzymes. We found that serine palmitoyltransferase and ceramide synthase are localized to the cytoplasm, whereas the ceramide reductase is in the periplasmic space. This is consistent with our previously proposed model wherein the second acyl chain is added in the cytoplasm prior to export to the periplasm where the lipid molecule is reduced.

The N-terminal signature motif on the transporter MCT1 is critical for CD147-mediated trafficking.

The human Solute Carrier (SLC) family member, monocarboxylate transporter 1 (MCT1), transports lactic and pyruvic acid across biological membranes to regulate cellular pH and metabolism. Proper trafficking of MCT1 from the endoplasmic reticulum to the plasma membrane hinges on its interactions with the membrane-bound chaperone protein, CD147. Here, using AlphaFold2 modeling and copurification, we show how a conserved signature motif located in the flexible N-terminus of MCT1 is a crucial region of interaction between MCT1 and the C-terminus of CD147. Mutations to this motif-namely, the thymic cancer linked G19C and the highly conserved W20A-destabilize the MCT1-CD147 complex and lead to a loss of proper membrane localization and cellular substrate flux. Notably, the monomeric stability of MCT1 remains unaffected in mutants, thus supporting the role of CD147 in mediating the trafficking of the heterocomplex. Using the auxiliary chaperone, GP70, we demonstrated that W20A-MCT1 can be trafficked to the plasma membrane, while G19C-MCT1 remains internalized. Overall, our findings underscore the critical role of the MCT1 transmembrane one signature motif for engaging CD147 and identify altered chaperone binding mechanisms between the CD147 and GP70 glycoprotein chaperones.

Rice tetraspanins express in specific domains of diverse tissues and regulate plant architecture and root growth.

Tetraspanins (TETs) are small transmembrane scaffold proteins that distribute proteins into highly organized microdomains, consisting of adaptors and signaling proteins, which play important roles in various biological events. In plants, understanding of tetraspanin is limited to the Arabidopsis TET genes' expression pattern and their function in leaf and root growth. Here, we comprehensively analyzed all rice tetraspanin (OsTET) family members, including their gene expression pattern, protein topology, and subcellular localization. We found that the core domain of OsTETs is conserved and shares a similar topology of four membrane-spanning domains with animal and plant TETs. OsTET genes are partially overlapping expressed in diverse tissue domains in vegetative and reproductive organs. OsTET proteins preferentially targeted the endoplasmic reticulum. Mutation analysis showed that OsTET5, OsTET6, OsTET9, and OsTET10 regulated plant height and tillering, and that OsTET13 controlled root growth in association with the jasmonic acid pathway. In summary, our work provides systematic new insights into the function of OsTETs in rice growth and development, and the data provides valuable resources for future research.

Comparative analysis of BZR1/BES1 family transcription factors in Arabidopsis.

Brassinazole Resistant 1 (BZR1) and bri1 EMS Suppressor 1 (BES1) are key transcription factors that mediate brassinosteroid (BR)-responsive gene expression in Arabidopsis. The BZR1/BES1 family is composed of BZR1, BES1, and four BES1/BZR1 homologs (BEH1-BEH4). However, little is known about whether BEHs are regulated by BR signaling in the same way as BZR1 and BES1. We comparatively analyzed the functional characteristics of six BZR1/BES1 family members and their regulatory mechanisms in BR signaling using genetic and biochemical analyses. We also compared their subcellular localizations regulated by the phosphorylation status, interaction with GSK3-like kinases, and heterodimeric combination. We found that all BZR1/BES1 family members restored the phenotypic defects of bri1-5 by their overexpression. Unexpectedly, BEH2-overexpressing plants showed the most distinct phenotype with enhanced BR responses. RNA-Seq analysis indicated that overexpression of both BZR1 and BEH2 regulates BR-responsive gene expression, but BEH2 has a much greater proportion of BR-independent gene expression than BZR1. Unlike BZR1 and BES1, the BR-regulated subcellular translocation of the four BEHs was not tightly correlated with their phosphorylation status. Notably, BEH1 and BEH2 are predominantly localized in the nucleus, which induces the nuclear accumulation of other BZR1/BES1 family proteins through heterodimerization. Altogether, our comparative analyses suggest that BEH1 and BEH2 play an important role in the functional interaction between BZR1/BES1 family transcription factors.

Genome-Wide Analysis of Human Long Noncoding RNAs: A Provocative Review.

Do long noncoding RNAs (lncRNAs) contribute little or substantively to human biology? To address how lncRNA loci and their transcripts, structures, interactions, and functions contribute to human traits and disease, we adopt a genome-wide perspective. We intend to provoke alternative interpretation of questionable evidence and thorough inquiry into unsubstantiated claims. We discuss pitfalls of lncRNA experimental and computational methods as well as opposing interpretations of their results. The majority of evidence, we argue, indicates that most lncRNA transcript models reflect transcriptional noise or provide minor regulatory roles, leaving relatively few human lncRNAs that contribute centrally to human development, physiology, or behavior. These important few tend to be spliced and better conserved but lack a simple syntax relating sequence to structure and mechanism, and so resist simple categorization. This genome-wide view should help investigators prioritize individual lncRNAs based on their likely contribution to human biology.

The glycine arginine-rich domain of the RNA-binding protein nucleolin regulates its subcellular localization.

Nucleolin is a multifunctional RNA Binding Protein (RBP) with diverse subcellular localizations, including the nucleolus in all eukaryotic cells, the plasma membrane in tumor cells, and the axon in neurons. Here we show that the glycine arginine rich (GAR) domain of nucleolin drives subcellular localization via protein-protein interactions with a kinesin light chain. In addition, GAR sequences mediate plasma membrane interactions of nucleolin. Both these modalities are in addition to the already reported involvement of the GAR domain in liquid-liquid phase separation in the nucleolus. Nucleolin transport to axons requires the GAR domain, and heterozygous GAR deletion mice reveal reduced axonal localization of nucleolin cargo mRNAs and enhanced sensory neuron growth. Thus, the GAR domain governs axonal transport of a growth controlling RNA-RBP complex in neurons, and is a versatile localization determinant for different subcellular compartments. Localization determination by GAR domains may explain why GAR mutants in diverse RBPs are associated with neurodegenerative disease.

Automatic recognition of protein subcellular location patterns in single cells from immunofluorescence images based on deep learning.

With the improvement of single-cell measurement techniques, there is a growing awareness that individual differences exist among cells, and protein expression distribution can vary across cells in the same tissue or cell line. Pinpointing the protein subcellular locations in single cells is crucial for mapping functional specificity of proteins and studying related diseases. Currently, research about single-cell protein location is still in its infancy, and most studies and databases do not annotate proteins at the cell level. For example, in the human protein atlas database, an immunofluorescence image stained for a particular protein shows multiple cells, but the subcellular location annotation is for the whole image, ignoring intercellular difference. In this study, we used large-scale immunofluorescence images and image-level subcellular locations to develop a deep-learning-based pipeline that could accurately recognize protein localizations in single cells. The pipeline consisted of two deep learning models, i.e. an image-based model and a cell-based model. The former used a multi-instance learning framework to comprehensively model protein distribution in multiple cells in each image, and could give both image-level and cell-level predictions. The latter firstly used clustering and heuristics algorithms to assign pseudo-labels of subcellular locations to the segmented cell images, and then used the pseudo-labels to train a classification model. Finally, the image-based model was fused with the cell-based model at the decision level to obtain the final ensemble model for single-cell prediction. Our experimental results showed that the ensemble model could achieve higher accuracy and robustness on independent test sets than state-of-the-art methods.

DAmiRLocGNet: miRNA subcellular localization prediction by combining miRNA-disease associations and graph convolutional networks.

MicroRNAs (miRNAs) are human post-transcriptional regulators in humans, which are involved in regulating various physiological processes by regulating the gene expression. The subcellular localization of miRNAs plays a crucial role in the discovery of their biological functions. Although several computational methods based on miRNA functional similarity networks have been presented to identify the subcellular localization of miRNAs, it remains difficult for these approaches to effectively extract well-referenced miRNA functional representations due to insufficient miRNA-disease association representation and disease semantic representation. Currently, there has been a significant amount of research on miRNA-disease associations, making it possible to address the issue of insufficient miRNA functional representation. In this work, a novel model is established, named DAmiRLocGNet, based on graph convolutional network (GCN) and autoencoder (AE) for identifying the subcellular localizations of miRNA. The DAmiRLocGNet constructs the features based on miRNA sequence information, miRNA-disease association information and disease semantic information. GCN is utilized to gather the information of neighboring nodes and capture the implicit information of network structures from miRNA-disease association information and disease semantic information. AE is employed to capture sequence semantics from sequence similarity networks. The evaluation demonstrates that the performance of DAmiRLocGNet is superior to other competing computational approaches, benefiting from implicit features captured by using GCNs. The DAmiRLocGNet has the potential to be applied to the identification of subcellular localization of other non-coding RNAs. Moreover, it can facilitate further investigation into the functional mechanisms underlying miRNA localization. The source code and datasets are accessed at http://bliulab.net/DAmiRLocGNet.

GraphLncLoc: long non-coding RNA subcellular localization prediction using graph convolutional networks based on sequence to graph transformation.

The subcellular localization of long non-coding RNAs (lncRNAs) is crucial for understanding lncRNA functions. Most of existing lncRNA subcellular localization prediction methods use k-mer frequency features to encode lncRNA sequences. However, k-mer frequency features lose sequence order information and fail to capture sequence patterns and motifs of different lengths. In this paper, we proposed GraphLncLoc, a graph convolutional network-based deep learning model, for predicting lncRNA subcellular localization. Unlike previous studies encoding lncRNA sequences by using k-mer frequency features, GraphLncLoc transforms lncRNA sequences into de Bruijn graphs, which transforms the sequence classification problem into a graph classification problem. To extract the high-level features from the de Bruijn graph, GraphLncLoc employs graph convolutional networks to learn latent representations. Then, the high-level feature vectors derived from de Bruijn graph are fed into a fully connected layer to perform the prediction task. Extensive experiments show that GraphLncLoc achieves better performance than traditional machine learning models and existing predictors. In addition, our analyses show that transforming sequences into graphs has more distinguishable features and is more robust than k-mer frequency features. The case study shows that GraphLncLoc can uncover important motifs for nucleus subcellular localization. GraphLncLoc web server is available at http://csuligroup.com:8000/GraphLncLoc/.

mRNA-CLA: An interpretable deep learning approach for predicting mRNA subcellular localization.

Messenger RNA (mRNA) is vital for post-transcriptional gene regulation, acting as the direct template for protein synthesis. However, the methods available for predicting mRNA subcellular localization need to be improved and enhanced. Notably, few existing algorithms can annotate mRNA sequences with multiple localizations. In this work, we propose the mRNA-CLA, an innovative multi-label subcellular localization prediction framework for mRNA, leveraging a deep learning approach with a multi-head self-attention mechanism. The framework employs a multi-scale convolutional layer to extract sequence features across different regions and uses a self-attention mechanism explicitly designed for each sequence. Paired with Position Weight Matrices (PWMs) derived from the convolutional neural network layers, our model offers interpretability in the analysis. In particular, we perform a base-level analysis of mRNA sequences from diverse subcellular localizations to determine the nucleotide specificity corresponding to each site. Our evaluations demonstrate that the mRNA-CLA model substantially outperforms existing methods and tools.

Recent Advancements in Subcellular Proteomics: Growing Impact of Organellar Protein Niches on the Understanding of Cell Biology.

The mammalian cell is a complex entity, with membrane-bound and membrane-less organelles playing vital roles in regulating cellular homeostasis. Organellar protein niches drive discrete biological processes and cell functions, thus maintaining cell equilibrium. Cellular processes such as signaling, growth, proliferation, motility, and programmed cell death require dynamic protein movements between cell compartments. Aberrant protein localization is associated with a wide range of diseases. Therefore, analyzing the subcellular proteome of the cell can provide a comprehensive overview of cellular biology. With recent advancements in mass spectrometry, imaging technology, computational tools, and deep machine learning algorithms, studies pertaining to subcellular protein localization and their dynamic distributions are gaining momentum. These studies reveal changing interaction networks because of "moonlighting proteins" and serve as a discovery tool for disease network mechanisms. Consequently, this review aims to provide a comprehensive repository for recent advancements in subcellular proteomics subcontexting methods, challenges, and future perspectives for method developers. In summary, subcellular proteomics is crucial to the understanding of the fundamental cellular mechanisms and the associated diseases.