Intelligent Image-Activated Cell Sorting.

A fundamental challenge of biology is to understand the vast heterogeneity of cells, particularly how cellular composition, structure, and morphology are linked to cellular physiology. Unfortunately, conventional technologies are limited in uncovering these relations. We present a machine-intelligence technology based on a radically different architecture that realizes real-time image-based intelligent cell sorting at an unprecedented rate. This technology, which we refer to as intelligent image-activated cell sorting, integrates high-throughput cell microscopy, focusing, and sorting on a hybrid software-hardware data-management infrastructure, enabling real-time automated operation for data acquisition, data processing, decision-making, and actuation. We use it to demonstrate real-time sorting of microalgal and blood cells based on intracellular protein localization and cell-cell interaction from large heterogeneous populations for studying photosynthesis and atherothrombosis, respectively. The technology is highly versatile and expected to enable machine-based scientific discovery in biological, pharmaceutical, and medical sciences.

N-terminal region of Drosophila melanogaster Argonaute2 forms amyloid-like aggregates.

BACKGROUND: Argonaute proteins play a central role in RNA silencing by forming protein-small RNA complexes responsible for the silencing process. While most Argonaute proteins have a short N-terminal region, Argonaute2 in Drosophila melanogaster (DmAgo2) harbors a long and unique N-terminal region. Previous in vitro biochemical studies have shown that the loss of this region does not impair the RNA silencing activity of the complex. However, an N-terminal mutant of Drosophila melanogaster has demonstrated abnormal RNA silencing activity. To explore the causes of this discrepancy between in vitro and in vivo studies, we investigated the biophysical properties of the region. The N-terminal region is highly rich in glutamine and glycine residues, which is a well-known property for prion-like domains, a subclass of amyloid-forming peptides. Therefore, the possibility of the N-terminal region functioning as an amyloid was tested. RESULTS: Our in silico and biochemical assays demonstrated that the N-terminal region exhibits amyloid-specific properties. The region indeed formed aggregates that were not dissociated even in the presence of sodium dodecyl sulfate. Also, the aggregates enhanced the fluorescence intensity of thioflavin-T, an amyloid detection reagent. The kinetics of the aggregation followed that of typical amyloid formation exhibiting self-propagating activity. Furthermore, we directly visualized the aggregation process of the N-terminal region under fluorescence microscopy and found that the aggregations took fractal or fibril shapes. Together, the results indicate that the N-terminal region can form amyloid-like aggregates. CONCLUSIONS: Many other amyloid-forming peptides have been reported to modulate the function of proteins through their aggregation. Therefore, our findings raise the possibility that aggregation of the N-terminal region regulates the RNA silencing activity of DmAgo2.

Real-time observation of flexible domain movements in CRISPR-Cas9.

The CRISPR-associated protein Cas9 is widely used for genome editing because it cleaves target DNA through the assistance of a single-guide RNA (sgRNA). Structural studies have revealed the multi-domain architecture of Cas9 and suggested sequential domain movements of Cas9 upon binding to the sgRNA and the target DNA These studies also hinted at the flexibility between domains; however, it remains unclear whether these flexible movements occur in solution. Here, we directly observed dynamic fluctuations of multiple Cas9 domains, using single-molecule FRET We found that the flexible domain movements allow Cas9 to adopt transient conformations beyond those captured in the crystal structures. Importantly, the HNH nuclease domain only accessed the DNA cleavage position during such flexible movements, suggesting the importance of this flexibility in the DNA cleavage process. Our FRET data also revealed the conformational flexibility of apo-Cas9, which may play a role in the assembly with the sgRNA Collectively, our results highlight the potential role of domain fluctuations in driving Cas9-catalyzed DNA cleavage.

The extracellular loop of pendrin and prestin modulates their voltage-sensing property.

Pendrin and prestin belong to the solute carrier 26 (SLC26) family of anion transporters. Prestin is unique among the SLC26 family members in that it displays voltage-driven motor activity (electromotility) and concurrent gating currents that manifest as nonlinear cell membrane electrical capacitance (nonlinear capacitance (NLC)). Although the anion transport mechanism of the SLC26 proteins has begun to be elucidated, the molecular mechanism of electromotility, which is thought to have evolved from an ancestral ion transport mechanism, still remains largely elusive. Here, we demonstrate that pendrin also exhibits large NLC and that charged residues present in one of the extracellular loops of pendrin and prestin play significant roles in setting the voltage-operating points of NLC. Our results suggest that the molecular mechanism responsible for sensing voltage is not unique to prestin among the members of the SLC26 family and that this voltage-sensing mechanism works independently of the anion transport mechanism.

Heterogeneous pathways and timing of factor departure during translation initiation.

The initiation of translation establishes the reading frame for protein synthesis and is a key point of regulation. Initiation involves factor-driven assembly at a start codon of a messenger RNA of an elongation-competent 70S ribosomal particle (in bacteria) from separated 30S and 50S subunits and initiator transfer RNA. Here we establish in Escherichia coli, using direct single-molecule tracking, the timing of initiator tRNA, initiation factor 2 (IF2; encoded by infB) and 50S subunit joining during initiation. Our results show multiple pathways to initiation, with orders of arrival of tRNA and IF2 dependent on factor concentration and composition. IF2 accelerates 50S subunit joining and stabilizes the assembled 70S complex. Transition to elongation is gated by the departure of IF2 after GTP hydrolysis, allowing efficient arrival of elongator tRNAs to the second codon presented in the aminoacyl-tRNA binding site (A site). These experiments highlight the power of single-molecule approaches to delineate mechanisms in complex multicomponent systems.

Real-time tRNA transit on single translating ribosomes at codon resolution.

Translation by the ribosome occurs by a complex mechanism involving the coordinated interaction of multiple nucleic acid and protein ligands. Here we use zero-mode waveguides (ZMWs) and sophisticated detection instrumentation to allow real-time observation of translation at physiologically relevant micromolar ligand concentrations. Translation at each codon is monitored by stable binding of transfer RNAs (tRNAs)-labelled with distinct fluorophores-to translating ribosomes, which allows direct detection of the identity of tRNA molecules bound to the ribosome and therefore the underlying messenger RNA (mRNA) sequence. We observe the transit of tRNAs on single translating ribosomes and determine the number of tRNA molecules simultaneously bound to the ribosome, at each codon of an mRNA molecule. Our results show that ribosomes are only briefly occupied by two tRNA molecules and that release of deacylated tRNA from the exit (E) site is uncoupled from binding of aminoacyl-tRNA site (A-site) tRNA and occurs rapidly after translocation. The methods outlined here have broad application to the study of mRNA sequences, and the mechanism and regulation of translation.

Initiation factor 2, tRNA, and 50S subunits cooperatively stabilize mRNAs on the ribosome during initiation.

Initiation factor 2 (IF2) is a key factor in initiation of bacterial protein synthesis. It recruits initiator tRNA to the small ribosomal subunit and facilitates joining of the large ribosomal subunit. Using reconstituted translation system of Escherichia coli and optical tweezers, we directly measure the rupture force between single ribosomal complexes and mRNAs for initiation complexes in the presence and the absence of IF2. We demonstrate that IF2 together with codon recognition by initiator tRNA increases the force required to dislocate mRNA from the ribosome complexes; mRNA stabilization by IF2 required the presence of a joined 50S subunit, and was independent of bound guanine nucleotide. IF2 thus helps lock the 70S ribosome over the start codon during initiation, thus maintaining reading frame. Our results show how mRNA is progressively stabilized on the ribosome through distinct steps of initiation.

Isolation of novel fluorogenic RNA aptamers via in vitro compartmentalization using microbead-display libraries.

Fluorogenic RNA aptamers, which specifically bind to fluorogens and dramatically enhance their fluorescence, are valuable for imaging and detecting RNAs and metabolites in living cells. Most fluorogenic RNA aptamers have been identified and engineered through iterative rounds of in vitro selection based on their binding to target fluorogens. While such selection is an efficient approach for generating RNA aptamers, it is less efficient for isolating fluorogenic aptamers because it does not directly screen for fluorogenic properties. In this study, we combined a fluorescence-based in vitro selection technique using water-in-oil microdroplets with an affinity-based selection technique to obtain fluorogenic RNA aptamers. This approach allowed us to identify novel fluorogenic aptamers for a biotin-modified thiazole orange derivative. Our results demonstrate that our approach can expand the diversity of fluorogenic RNA aptamers, thus leading to new applications for the imaging and detection of biomolecules.

Quantitative live-cell imaging of secretion activity reveals dynamic immune responses.

Quantification of cytokine secretion has facilitated advances in the field of immunology, yet the dynamic and varied secretion profiles of individual cells, particularly those obtained from limited human samples, remain obscure. Herein, we introduce a technology for quantitative live-cell imaging of secretion activity (qLCI-S) that enables high-throughput and dual-color monitoring of secretion activity at the single-cell level over several days, followed by transcriptome analysis of individual cells based on their phenotype. The efficacy of qLCI-S was demonstrated by visualizing the characteristic temporal pattern of cytokine secretion of group 2 innate lymphoid cells, which constitute less than 0.01% of human peripheral blood mononuclear cells, and by revealing minor subpopulations with enhanced cytokine production. The underlying mechanism of this feature was linked to the gene expression of stimuli receptors. This technology paves the way for exploring gene expression signatures linked to the spatiotemporal dynamic nature of various secretory functions.

Peptide bond formation destabilizes Shine-Dalgarno interaction on the ribosome.

The ribosome is a molecular machine that translates the genetic code contained in the messenger RNA into an amino acid sequence through repetitive cycles of transfer RNA selection, peptide bond formation and translocation. Here we demonstrate an optical tweezer assay to measure the rupture force between a single ribosome complex and mRNA. The rupture force was compared between ribosome complexes assembled on an mRNA with and without a strong Shine-Dalgarno (SD) sequence-a sequence found just upstream of the coding region of bacterial mRNAs, involved in translation initiation. The removal of the SD sequence significantly reduced the rupture force in complexes carrying an aminoacyl tRNA, Phe-tRNA(Phe), in the A site, indicating that the SD interactions contribute significantly to the stability of the ribosomal complex on the mRNA before peptide bond formation. In contrast, the presence of a peptidyl tRNA analogue, N-acetyl-Phe-tRNA(Phe), in the A site, which mimicked the post-peptidyl transfer state, weakened the rupture force as compared to the complex with Phe-tRNA(Phe), and the resultant force was the same for both the SD-containing and SD-deficient mRNAs. These results suggest that formation of the first peptide bond destabilizes the SD interaction, resulting in the weakening of the force with which the ribosome grips an mRNA. This might be an important requirement to facilitate movement of the ribosome along mRNA during the first translocation step.

Accumulation of TERT in mitochondria exerts two opposing effects on apoptosis.

Telomerase reverse transcriptase (TERT) is a protein that catalyzes the reverse transcription of telomere elongation. TERT is also expected to play a non-canonical role beyond telomere lengthening since it localizes not only in the nucleus but also in mitochondria, where telomeres do not exist. Several studies have reported that mitochondrial TERT regulates apoptosis induced by oxidative stress. However, there is still some controversy as to whether mitochondrial TERT promotes or inhibits apoptosis, mainly due to the lack of information on changes in TERT distribution in individual cells over time. Here, we simultaneously detected apoptosis and TERT localization after oxidative stress in individual HeLa cells by live-cell tracking. Single-cell tracking revealed that the stress-induced accumulation of TERT in mitochondria caused apoptosis, but that accumulation increased over time until cell death. The results suggest a new model in which mitochondrial TERT has two opposing effects at different stages of apoptosis: it predetermines apoptosis at the first stage of cell-fate determination, but also delays apoptosis at the second stage. As such, our data support a model that integrates the two opposing hypotheses on mitochondrial TERT's effect on apoptosis. Furthermore, detailed statistical analysis of TERT mutations, which have been predicted to inhibit TERT transport to mitochondria, revealed that these mutations suppress apoptosis independent of mitochondrial localization of TERT. Together, these results imply that the non-canonical functions of TERT affect a wide range of mitochondria-dependent and mitochondria-independent apoptosis pathways.

Time-dependent cell-state selection identifies transiently expressed genes regulating ILC2 activation.

The decision of whether cells are activated or not is controlled through dynamic intracellular molecular networks. However, the low population of cells during the transition state of activation renders the analysis of the transcriptome of this state technically challenging. To address this issue, we have developed the Time-Dependent Cell-State Selection (TDCSS) technique, which employs live-cell imaging of secretion activity to detect an index of the transition state, followed by the simultaneous recovery of indexed cells for subsequent transcriptome analysis. In this study, we used the TDCSS technique to investigate the transition state of group 2 innate lymphoid cells (ILC2s) activation, which is indexed by the onset of interleukin (IL)-13 secretion. The TDCSS approach allowed us to identify time-dependent genes, including transiently induced genes (TIGs). Our findings of IL4 and MIR155HG as TIGs have shown a regulatory function in ILC2s activation.

Annexin A1 is a cell-intrinsic metalloregulator of zinc in human ILC2s.

Group 2 innate lymphoid cells (ILC2s) produce large amounts of type 2 cytokines including interleukin-5 (IL-5) and IL-13 in response to various stimuli, causing allergic and eosinophilic diseases. However, the cell-intrinsic regulatory mechanisms of human ILC2s remain unclear. Here, we analyze human ILC2s derived from different tissues and pathological conditions and identify ANXA1, encoding annexin A1, as a commonly highly expressed gene in non-activated ILC2s. The expression of ANXA1 decreases when ILC2s activate, but it increases autonomously as the activation subsides. Lentiviral vector-based gene transfer experiments show that ANXA1 suppresses the activation of human ILC2s. Mechanistically, ANXA1 regulates the expression of the metallothionein family genes, including MT2A, which modulate intracellular zinc homeostasis. Furthermore, increased intracellular zinc levels play an essential role in the activation of human ILC2s by promoting the mitogen-activated protein kinase (MAPK) and nuclear factor κB (NF-κB) pathways and GATA3 expression. Thus, the ANXA1/MT2A/zinc pathway is identified as a cell-intrinsic metalloregulatory mechanism for human ILC2s.

Zero-mode waveguides and nanopore-based sequencing technologies accelerate single-molecule studies.

Single-molecule technologies can provide detailed information regarding molecular mechanisms and interactions that cannot easily be studied on the bulk scale; generally, individual molecular behaviors cannot be distinguished, and only average characteristics can be measured. Nevertheless, the development of the single-molecule sequencer had a significant impact on conventional in vitro single-molecule research, featuring automated equipment, high-throughput chips, and automated analysis systems. However, the utilization of sequencing technology in in vitro single-molecule research is not yet globally prevalent, owing to the large gap between highly organized single-molecule sequencing and manual-based in vitro single-molecule research. Here, we describe the principles of zero-mode waveguides (ZMWs) and nanopore methods used as single-molecule DNA sequencing techniques, and provide examples of functional biological measurements beyond DNA sequencing that contribute to a global understanding of the current applications of these sequencing technologies. Furthermore, through a comparison of these two technologies, we discuss future applications of DNA sequencing technologies in in vitro single-molecule research.

Whole-embryonic identification of maternal microchimeric cell types in mouse using single-cell RNA sequencing.

Even though the mother and the fetus of placental mammals are immunologically non-self with respect to one other, mutual exchange of small numbers of cells between them is known to occur. Maternal cells entering the fetus, called maternal microchimeric cells (MMc cells), are thought to be involved in different physiological phenomena, such as establishing immune tolerance, tissue repair, and the pathogenesis or deterioration of some inflammatory diseases and congenital malformations. While specific MMc cell types have been reported as associated with these phenomena, the contribution of MMc cells to these different outcomes remains unknown. As one possibility, we hypothesized that different embryos have differing repertoires of MMc cell types, leading to or biasing embryos toward different fates. To date, no studies have succeeded in identifying the MMc cell type repertoire of a single embryo. Accordingly, here, we isolated MMc cells from whole mouse embryos, determined their types, and analyzed their MMc cell type variability. By combining our previously established, whole-embryonic MMc isolation method with single-cell RNA sequencing, we successfully estimated the cell type repertoires of MMc cells isolated from 26 mouse embryos. The majority of MMc cells were immune-related cells, such as myeloid cells and granulocytes. We also detected stem cell-like MMc cells expressing proliferation marker genes and terminally differentiated cells. As hypothesized, we noted statistically significant inter-individual variation in the proportion of immune-related cells in the different embryos. We here successfully estimated MMc cell types in individual whole mouse embryos. The proportion of immune-related cells significantly differed among the individual embryos, suggesting that the variations are one of the potential mechanisms underlying the differing MMc-related physiological phenomena in offspring. These findings provide insight into cell-level epigenetics by maternal cells.

Upregulation of IL-4 receptor signaling pathway in circulating ILC2s from asthma patients.

Background: Group 2 innate lymphoid cells (ILC2s) produce type 2 cytokines by stimulation with epithelial cell-derived cytokines and are implicated in the pathogenesis of various allergic diseases, including asthma. However, differences in the molecular characteristics of ILC2s between patients with asthma and healthy subjects remain unclear. Objective: We sought to evaluate differences in cytokine production capacity and gene expression profile of ILC2s in the peripheral blood of patients with asthma and healthy subjects. Methods: We evaluated ILC2s derived from 15 patients with asthma and 7 healthy subjects using flow cytometry, live-cell imaging of secretion activity analysis, and RNA-sequencing. Results: ILC2s were sorted as CD45+Lineage-CRTH2+CD127+CD161+ cells from the peripheral blood of patients with asthma and healthy subjects, and the number of ILC2s was decreased in patients with asthma (851 ± 1134 vs 2679 ± 3009 cells/20 mL blood; P = .0066). However, patient-derived ILC2s were activated to produce more IL-5 and IL-13 in response to stimulation with IL-2, IL-33, and thymic stromal lymphopoietin compared with healthy subject-derived ILC2s (P = .0032 and P = .0085, respectively). Furthermore, RNA-sequencing analysis revealed that patient-derived ILC2s had different gene expression profiles, such as increased expression in cell growth-related genes (CDKN1b, CCNG2, CCND2, CCN1), prostaglandin E receptor (PTGER2), and IL-4 receptor. In addition, a gene set of the IL-4 receptor signaling pathway was significantly upregulated in ILC2s in patients with asthma (P = .042). Conclusions: Our results suggest that circulating ILC2s in patients with asthma are preactivated via the IL-4 receptor signaling pathway and produce IL-5 and IL-13 vigorously by stimulation.

Heterogeneous dissociation process of truncated RNAs by oligomerized Vasa helicase.

RNA helicases are enzymes that generally unwind double-stranded RNA using ATP hydrolysis energy, mainly involved in RNA metabolism, transcription, translation, and mRNA splicing. While the helicase core is crucial for RNA unwinding activity, N- and C-terminal extensions of specific helicases may contain an intrinsically disordered region for electrostatic interaction, resulting in the formation of droplets in the cytoplasm. However, how the disordered region of the RNA helicase contributes to RNA unwinding and dissociation remains unclear. Here, we focused on Bombyx mori Vasa, which unwinds truncated target transposon RNAs from the piRNA-induced silencing complex piRISC. In this study, we used single-molecule techniques to visualise how Vasa dynamically interacts with piRISC and investigate how Vasa oligomerization is involved in the process of piRNA amplification, named the ping-pong pathway. We found that Vasa's oligomerization is required during these processes in vitro and in vivo, and that Vasa triggers the dissociation of truncated RNA in heterogeneous pathways. Our single-molecule results suggest that oligomerized Vasa guides the timing of the process regulating overall dissociation efficiency.

Single-molecule imaging of full protein synthesis by immobilized ribosomes.

How folding of proteins is coupled to their synthesis remains poorly understood. Here, we apply single-molecule fluorescence imaging to full protein synthesis in vitro. Ribosomes were specifically immobilized onto glass surfaces and synthesis of green fluorescent protein (GFP) was achieved using modified commercial Protein Synthesis using Recombinant Elements that lacked ribosomes but contained purified factors and enzyme that are required for translation in Escherichia coli. Translation was monitored using a GFP mutant (F64L/S65T/F99S/M153T/V163A) that has a high fluorophore maturation rate and that contained the Secretion Monitor arrest sequence to prevent dissociation from the ribosome. Immobilized ribosomal subunits were labeled with Cy3 and GFP synthesis was measured by colocalization of GFP fluorescence with the ribosome position. The rate of appearance of colocalized ribosome GFP was equivalent to the rates of fluorescence appearance coupled with translation measured in bulk, and the ribosome-polypeptide complexes were stable for hours. The methods presented here are applicable to single-molecule investigation of translational initiation, elongation and cotranslational folding.

Development of a sequencing system for spatial decoding of DNA barcode molecules at single-molecule resolution.

Single-cell transcriptome analysis has been revolutionized by DNA barcodes that index cDNA libraries, allowing highly multiplexed analyses to be performed. Furthermore, DNA barcodes are being leveraged for spatial transcriptomes. Although spatial resolution relies on methods used to decode DNA barcodes, achieving single-molecule decoding remains a challenge. Here, we developed an in-house sequencing system inspired by a single-molecule sequencing system, HeliScope, to spatially decode DNA barcode molecules at single-molecule resolution. We benchmarked our system with 30 types of DNA barcode molecules and obtained an average read length of ~20 nt with an error rate of less than 5% per nucleotide, which was sufficient to spatially identify them. Additionally, we spatially identified DNA barcode molecules bound to antibodies at single-molecule resolution. Leveraging this, we devised a method, termed "molecular foot printing", showing potential for applying our system not only to spatial transcriptomics, but also to spatial proteomics.

Virtual-freezing fluorescence imaging flow cytometry.

By virtue of the combined merits of flow cytometry and fluorescence microscopy, imaging flow cytometry (IFC) has become an established tool for cell analysis in diverse biomedical fields such as cancer biology, microbiology, immunology, hematology, and stem cell biology. However, the performance and utility of IFC are severely limited by the fundamental trade-off between throughput, sensitivity, and spatial resolution. Here we present an optomechanical imaging method that overcomes the trade-off by virtually freezing the motion of flowing cells on the image sensor to effectively achieve 1000 times longer exposure time for microscopy-grade fluorescence image acquisition. Consequently, it enables high-throughput IFC of single cells at >10,000 cells s-1 without sacrificing sensitivity and spatial resolution. The availability of numerous information-rich fluorescence cell images allows high-dimensional statistical analysis and accurate classification with deep learning, as evidenced by our demonstration of unique applications in hematology and microbiology.