Biology subjects

Hsu, J.

Publications and source records attributed to Hsu, J..

At least 19 recordsLinked to original sources

bioRxiv2026

CA19-9 induces microenvironment remodeling in pancreatic ductal adenocarcinoma

Durable therapeutic efficacy remains a major barrier to improving outcomes for patients with pancreatic ductal adenocarcinoma (PDAC). An immunosuppressive tumor microenvironment (TME) is a hallmark of PDAC and has been demonstrated to be a dominant driver of therapeutic resistance. The aberrant glycan CA19-9 is prevalent in PDAC and drives tumor progression, but the paracrine mechanisms by which it contributes to TME remodeling are unknown. To address this, we mapped TME changes and performed functional analyses using a genetically engineered mouse model (GEMM) harboring KrasG12D mutation and inducible CA19-9 expression. Elevation of CA19-9 led to expansion of antigen-presenting cancer associated fibroblasts (apCAFs) and regulatory T cells (Tregs), which can drive immunosuppression. Antibody blockade of CA19 -9 resulted in significant restoration of normal histology and decreased apCAF and Treg populations. We dissected the paracrine signaling mechanisms that drive this TME remodeling in vitro using mouse and human organoid mono- and co-culture models as well as in vivo using GEMMs and syngeneic orthotopic transplantation models. CA19-9 induced IL1a and TGFb expression, reprogramming pancreatic mesothelial cells into apCAFs in vitro, which in turn directly ligated naive Cd4+ T cells resulting in Treg differentiation in co-cultures. Antibody blockade of IL1a and TGFb in mice led to reduced apCAF and Treg differentiation. We previously reported that CA19-9 modification of the secreted Fbln3 protein increased Egfr engagement and now find that the induction of IL1a and TGFb expression by CA19-9 is dependent on Fbln3 hyperactivation of EGFR signaling. Genetic depletion of Fbln3 led to reduced tumor progression and increased Cd8+ T cell infiltration in mice. Together these findings identify a previously unknown signaling axis driving immunosuppressive phenotypes in PDAC, uncovering multiple potential nodes to relieve the immunosuppressive pressures within the PDAC TME.

cancer biology↗

bioRxiv2026

CA19-9 promotes liver metastasis of pancreatic cancer through E-selectin mediated extravasation

Pancreatic ductal adenocarcinoma (PDAC) frequently metastasizes to the liver, which drives patient mortality. CA19-9 is elevated in most PDAC tumors and is widely used as a clinical biomarker. Elevated serum levels are associated with poor outcomes. However, whether CA19-9 functionally contributes to metastatic progression has not been fully defined, in part because mice lack endogenous CA19-9 expression. Here, using syngeneic murine PDAC cells engineered to express CA19-9, we investigated its functional role in liver metastasis. In splenic injection models, CA19-9 expression markedly increased liver metastatic burden by promoting both metastatic seeding and subsequent metastatic outgrowth. In vitro, CA19-9 enhanced tumor cell adhesion to endothelial cells through interaction with E-selectin. Metastatic seeding of CA19-9-expressing cells was reduced by genetic deletion of E-selectin or antibody neutralization of either CA19-9 or E-selectin in vivo. Therapeutic targeting of CA19-9 with a neutralizing antibody markedly reduced liver metastatic burden after metastatic seeding. CA19-9 expression increased AKT signaling in PDAC cells and liver metastases, and CA19-9 levels correlated with AKT activation in human PDAC tissues. These findings show that CA19-9 promotes PDAC liver metastasis through E-selectin-dependent metastatic seeding and AKT-associated metastatic outgrowth, highlighting CA19-9 as a functional mediator of PDAC metastasis and a potential therapeutic target.

cancer biology↗

bioRxiv2025

Structural and biochemical basis for cannabinoid cyclase activity in marine bacterial flavoenzymes

The marine bacterial flavoenzymes Clz9 and Tcz9 can process cannabigerolic acid (CBGA) to the minor cannabinoid, cannabichromenic acid (CBCA), however, the mechanistic details of this extrinsic transformation are still obscure. Here, we report a thorough analysis of CBCA-formation by Clz9 and Tcz9 through high-resolution crystallographic characterization, biochemical analysis, and spectroscopic interrogation. Our work reveals that Clz9 and Tcz9 use different biochemical mechanisms from Cannabis cyclases and each other in their production of CBCA. Collection of a high-resolution substrate-bound structure, the first for any cannabinoid cyclase, provides key insights into how active site architecture affects substrate binding and stereoselectivity. Engineering approaches improve the stereoselectivity of CBCA formation by Clz9 and Tcz9, providing access to (R) and (S)-CBCA. Collectively, our work advances understanding of enzymatic cannabinoid formation and cements Clz9 and Tcz9 as two unique members of the BBE-like enzyme family with encouraging potential for biocatalytic cannabinoid production applications.

biochemistry↗

bioRxiv2025

Spatial isoform sequencing at sub-micrometer single-cell resolution reveals novel patterns of spatial isoform variability in brain cell types

Spatial long-read technologies are becoming increasingly common but lack nanometer and therefore often single-cell resolution. This leaves the question unanswered of whether spatially variable isoforms represent spatial variability within one cell type or differences in cell-type composition between different regions. Here, we developed Spl-ISO-Seq2 (220nm spot size and 500nm resolution), and the accompanying software packages Spl-IsoQuant-2 and Spl-IsoFind, enabling long-read sequencing using up to 500 million barcodes compared to 80,000 previously. Applying this to the adult mouse brain, we compared spatial variability by examining (a) differential isoform abundance between known brain regions and (b) spatial isoform patterns that do not align with predefined regions. While the former revealed more spatial isoform differences, both approaches identified overlapping hits, e.g., Rps24 in oligodendrocytes. For Snap25, previously known to exhibit spatial isoform variation, we now show that this variability occurs in excitatory neurons. The second approach also uncovered patterns not captured by predefined-region comparisons, e.g., Tnnc1 in excitatory neurons. Furthermore, we show that a surprising number of spatial isoform signals is not driven by cell-type composition alone. Finally, we applied our software to newly generated Visium HD 3 long-read data to demonstrate its applicability and strong reproducibility across protocols and biological replicates. Taken together, our experimental and analytical methods enrich spatial transcriptomics with a so-far elusive isoform view of spatial variation for individual cell types.

neuroscience↗

bioRxiv2025

A spatial long-read approach at near-single-cell resolution reveals developmental regulation of splicing and polyadenylation sites in distinct cortical layers and cell types.

Genome-wide single-cell and spatial long-read approaches have gained traction, but mostly lack single-cell resolution - and yield limited read lengths. Here, we introduce spatial ISOform sequencing (Spl-ISO-Seq), which reveals exons and polyadenylation sites from long reads with near-single-cell resolution. Spl-ISO-Seq selects long cDNAs and doubles to triples read lengths compared to standard preparations. Adding a highly specific software tool (Spl-ISOquant) and comparing human post-mortem pre-puberty samples of the visual cortex (8-11 years) to post-puberty samples (16-19 years), we find that cortical layers harbor stronger splicing and poly(A)-site regulation than the adjacent white matter, with enrichment of multiple protein-domain types. For oligodendrocytes however, developmental splicing changes are stronger in white matter. Among cortical layers, layer 4 has the most developmental changes in alternative-exon inclusion in excitatory neurons and in poly(A) sites. We also find many repeat elements, especially ERV1 long terminal repeats downstream of developmentally-regulated layer 4 exons. Overall, alternative splicing changes are linked to synapses - specifically at the post-synapse. Age-linked splicing changes in layers 1-3 and 4 are associated with autism spectrum disorder but not with schizophrenia, amyotrophic lateral sclerosis and Alzheimers disease. These results root developmental splicing changes during puberty and the resulting protein changes in specific layers and cell types. More generally, our new technologies enable new observations for any complex tissue.

genomics↗

bioRxiv2025

Macrophage ferroptosis inhibits Aspergillus conidial killing in lung transplantation

Immune suppression heightens the risk for fungal infections, but the mechanisms that result in clinical disease are poorly understood. Here we demonstrate that macrophage ferroptosis, an iron-dependent form of regulated cell death, inhibits Aspergillus fumigatus (Af) killing. In a mouse tracheal transplant model of Af infection, we observed an increase in macrophage lipid peroxidation, a decreased expression of negative ferroptosis regulators Gpx4 and Slc7a11, and an increase in positive regulators Ptgs2 and Nox2, relative to syntransplants. Depletion of macrophages in transplant recipients decreased Af invasion. In vitro, iron overload reduced macrophage viability and decreased their capability to kill Af spores, through a decrease in lysosomal acidification and lysosomal loss. Treatment with ferrostatin-1, a ferroptosis inhibitor, and deferasirox (an iron chelator) restored Af killing. Ferroptotic alveolar macrophages isolated from lung transplant patients also showed a decreased ability to kill Af spores and the patients bronchoalveolar lavage was characterized by higher iron levels and markers of ferroptotic stress compared to non-lung transplants. These characteristics were strongly correlated with a clinical history of fungal infections, independent of immune suppressive medications. Our findings indicate that macrophage ferroptosis augments the risk of invasive aspergillosis, representing a novel mechanism for host immune dysfunction. Graphical AbstractSchematic of proposed mechanism underlying ferroptosis induced immune dysregulation and increased Af invasion in lung transplantation. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=150 SRC="FIGDIR/small/643092v1_ufig1.gif" ALT="Figure 1"> View larger version (104K): org.highwire.dtl.DTLVardef@1112102org.highwire.dtl.DTLVardef@18e2ec1org.highwire.dtl.DTLVardef@19f9c5forg.highwire.dtl.DTLVardef@1d325e8_HPS_FORMAT_FIGEXP M_FIG C_FIG

immunology↗

bioRxiv2024

Identification of molecular and functional subtypes using chronic pancreatitis patient-derived organoid models

Chronic pancreatitis (CP) affects [~]3 million people worldwide, yet altering the course of disease is challenging. We developed a patient-derived organoid (PDO) platform to investigate the molecular pathogenesis of this disease and identify therapeutic strategies. We generated 36 PDOs from patients with idiopathic, hereditary, and alcohol-related CP with high genetic concordance. PDOs retained inflammation-associated transcriptional and proteomic features. Transcriptomic profiling revealed three molecular subtypes of CP independent of etiology. We discovered widespread dysfunction of the cystic fibrosis transmembrane conductance regulator (CFTR) in half of the CP PDOs, including those with wildtype CFTR. Clinically available CFTR modulators stabilized mutant or wildtype CFTR, restored CFTR function, and decreased mitogenic and inflammatory signaling. This work provides the first comprehensive PDO platform for modeling CP. We demonstrate the utility of this platform for precision therapeutic investigations. Our findings reveal CFTR modulators as a broadly applicable and effective therapeutic strategy.

cell biology↗

bioRxiv2024

Dietary fiber content in clinical ketogenic diets modifies the gut microbiome and seizure resistance in mice

The gut microbiome is emerging as an important modulator of the anti-seizure effects of the classic ketogenic diet. However, many variations of the ketogenic diet are used clinically to treat refractory epilepsy, and how different dietary formulations differentially modify the gut microbiome in ways that impact seizure outcome is poorly understood. We find that clinically prescribed ketogenic infant formulas vary in macronutrient ratio, fat source, and fiber content and also in their ability to promote resistance to 6-Hz psychomotor seizures in mice. By screening specific dietary variables for their effects on a model human infant microbial community, we observe that dietary fiber, rather than fat ratio or source, drives substantial metagenomic shifts. Addition of dietary fiber to a fiber-deficient ketogenic formula restores seizure resistance, and supplementing protective ketogenic formulas with excess dietary fiber further potentiates seizure resistance. By screening 13 fiber sources and types, we identify distinct subsets of metagenomic responses in the model human infant microbial community that correspond with increased seizure resistance in mice. In particular, supplementation with seizure-protective fibers enriches microbial representation of genes related to queuosine biosynthesis and preQ0 biosynthesis and decreases representation of microbial genes related to sucrose degradation, which is also seen in seizure-protected mice that are fed fiber-containing ketogenic infant formulas. Overall, this study reveals that different formulations of clinical ketogenic diets, and dietary fiber content in particular, differentially impact seizure outcome in mice, likely through modification of the gut microbiome. Understanding interactions between dietary components of the ketogenic diet, the gut microbiome, and host susceptibility to seizures could inform novel microbiome-guided approaches to treat refractory epilepsy.

microbiology↗

bioRxiv2024

Predicting cell-type-specific exon inclusion in the human brain reveals more complex splicing mechanisms in neurons than glia

Alternative splicing contributes to molecular diversity across brain cell types. RNA-binding proteins (RBPs) regulate splicing, but the genome-wide mechanisms remain poorly understood. Here, we used RBP binding sites and/or the genomic sequence to predict exon inclusion in neurons and glia as measured by long-read single-cell data in human hippocampus and frontal cortex. We found that alternative splicing is harder to predict in neurons compared to glia in both brain regions. Comparing neurons and glia, the position of RBP binding sites in alternatively spliced exons in neurons differ more from non-variable exons indicating distinct splicing mechanisms. Model interpretation pinpointed RBPs, including QKI, potentially regulating alternative splicing between neurons and glia. Finally, using our models, we accurately predict and prioritize the effect of splicing QTLs. Taken together, our models provide new insights into the mechanisms regulating cell-type-specific alternative splicing and can accurately predict the effect of genetic variants on splicing.

bioinformatics↗

bioRxiv2024

ScISOr-ATAC reveals convergent and divergent splicing and chromatin specificities between matched cell types across cortical regions, evolution, and in Alzheimer's disease.

Multimodal measurements have become widespread in genomics, however measuring open chromatin accessibility and splicing simultaneously in frozen brain tissues remains unconquered. Hence, we devised Single-Cell-ISOform-RNA sequencing coupled with the Assay-for-Transposase-Accessible-Chromatin (ScISOr-ATAC). We utilized ScISOr-ATAC to assess whether chromatin and splicing alterations in the brain convergently affect the same cell types or divergently different ones. We applied ScISOr-ATAC to three major conditions: comparing (i) the Rhesus macaque (Macaca mulatta) prefrontal cortex (PFC) and visual cortex (VIS), (ii) cross species divergence of Rhesus macaque versus human PFC, as well as (iii) dysregulation in Alzheimers disease in human PFC. We found that among cortical-layer biased excitatory neuron subtypes, splicing is highly brain-region specific for L3-5/L6 IT_RORB neurons, moderately specific in L2-3 IT_CUX2.RORB neurons and unspecific in L2-3 IT_CUX2 neurons. In contrast, at the chromatin level, L2-3 IT_CUX2.RORB neurons show the highest brain-region specificity compared to other subtypes. Likewise, when comparing human and macaque PFC, strong evolutionary divergence on one molecular modality does not necessarily imply strong such divergence on another molecular level in the same cell type. Finally, in Alzheimers disease, oligodendrocytes show convergently high dysregulation in both chromatin and splicing. However, chromatin and splicing dysregulation most strongly affect distinct oligodendrocyte subtypes. Overall, these results indicate that chromatin and splicing can show convergent or divergent results depending on the performed comparison, justifying the need for their concurrent measurement to investigate complex systems. Taken together, ScISOr-ATAC allows for the characterization of single-cell splicing and chromatin patterns and the comparison of sample groups in frozen brain samples.

genomics↗

bioRxiv2024

Changes in Gait Asymmetry May Be Caused by Adaptation of Spinal Reflexes

In a recent human study, we found that adaptive changes in step length asymmetry (SLA) are correlated with similar changes in the H-reflex gains of the leg muscles during split-belt treadmill locomotion. While this observation indicated a closer link between gait asymmetry and spinal reflex adaptation, it did not reveal their causal relationship. To better understand this relationship, here we use a neuromuscular model of human walking whose control relies primarily on spinal reflexes. Subjecting the model to split-belt treadmill locomotion with different combinations of belt speed and reflex gain adaptation patterns, we find that belt speed changes increase the variability in SLA but do not result in consistent SLA patterns as observed in human experiments, whereas reflex gain adaptations do. Furthermore, we find that the model produces SLA patterns similar to healthy adults when its reflex gains are adapted in a way similar to the H-reflex changes we observed in our previous human study. The model also predicts SLA patterns similar to the ones observed for cerebellar degeneration patients when the reflexes do not adapt beyond a sudden dip at the time the ipsilateral belt speed is lowered. Our results suggest that SLA does not arise from imposing belt speed changes but requires the adaptation of the reflex gains, and that the dynamic adaptation of these gains may be an essential part of human gait control when encountering unexpected environment changes such as the uneven speed changes in split-belt treadmill locomotion.

neuroscience↗

bioRxiv2023

Patient-Specific Vascularized Tumor Model: Blocking TAM Recruitment with Multispecific Antibodies Targeting CCR2 and CSF-1R

Tumor-associated inflammation drives cancer progression and therapy resistance, with the infiltration of monocyte-derived tumor-associated macrophages (TAMs) associated with poor prognosis in diverse cancers. Targeting TAMs holds potential against solid tumors, but effective immunotherapies require testing on immunocompetent human models prior to clinical trials. Here, we develop an in vitro model of microvascular networks that incorporates tumor spheroids or patient tissues. By perfusing the vasculature with human monocytes, we investigate monocyte trafficking into the tumor and evaluate immunotherapies targeting the human tumor microenvironment. Our findings demonstrate that macrophages in vascularized breast and lung tumor models can enhance monocyte recruitment via TAM-produced CCL7 and CCL2, mediated by CSF-1R. Additionally, we assess a novel multispecific antibody targeting CCR2, CSF-1R, and neutralizing TGF-{beta}, referred to as CSF1R/CCR2/TGF-{beta} Ab, on monocytes and macrophages using our 3D models. This antibody repolarizes TAMs towards an anti-tumoral M1-like phenotype, reduces monocyte chemoattractant protein secretion, and effectively blocks monocyte migration. Finally, we show that the CSF1R/CCR2/TGF-{beta} Ab inhibits monocyte recruitment in patient-specific vascularized tumor models. Overall, this vascularized tumor model offers valuable insights into monocyte recruitment and enables functional testing of innovative therapeutic antibodies targeting TAMs in the tumor microenvironment (TME).

bioengineering↗

bioRxiv2023

Characterization of Unique Pathological Features of COVID-Associated Coagulopathy: Studies with AC70 hACE2 Transgenic Mice Highly Permissive to SARS-CoV-2 Infection

COVID-associated coagulopathy seemly plays a key role in post-acute sequelae of SARS-CoV-2 infection. However, the underlying pathophysiological mechanisms are poorly understood, largely due to the lack of suitable animal models that recapitulate key clinical and pathological symptoms. Here, we fully characterized AC70 line of human ACE2 transgenic (AC70 hACE2 Tg) mice for SARS-CoV-2 infection. We noted that this model is highly permissive to SARS-CoV-2 with values of 50% lethal dose and infectious dose as [~] 3 and [~] 0.5 TCID50 of SARS-CoV-2, respectively. Mice infected with 105 TCID50 of SARS-CoV-2 rapidly succumbed to infection with 100% mortality within 5 days. Lung and brain were the prime tissues harboring high viral titers, accompanied by histopathology. However, viral RNA and many inflammatory mediators could be readily detectable in other organs, suggesting the nature of a systemic infection. Lethal challenge of AC70 hACE2 Tg mice caused acute onset of leukopenia, lymphopenia, along with an increased neutrophil-to-lymphocyte ratio. Importantly, infected animals recapitulated key features of COVID-19-associated coagulopathy, including significantly elevated levels of D-dimer, t-PA, PAI-1, and circulating NETs, along with activated platelet/endothelium marker. Immunohistochemical staining with anti-PF4 antibody revealed profound platelet aggregates especially within blocked veins of the lungs. ANXA2 is known to interact with S100A10 to form heterotetrametric complexes, serving as coreceptors for t-PA to regulate membrane fibrinolysis. Thus, our results revealing elevated IgG type anti-ANXA2 antibody production, downregulated de novo ANXA2/S100A10 synthesis, and reduced AnxA2/S100A10 association in infected mice support an important role of this protein in the pathogenesis of acute COVID-19. In summary, we showed that acute SARS-CoV-2 infection of AC70 hACE2 Tg mice triggered a hypercoagulable state coexisting with ill-regulated fibrinolysis, accompanied by dysregulation of ANXA2 system, which might serve as druggable targets for development of antithrombotic and/or anti-fibrinolytic agents to attenuate pathogenesis of COVID-19. Author SummaryAccumulating evidence strongly suggests that COVID-associated coagulopathy characterized by dysregulation of the coagulation cascade, fibrinolysis system and pulmonary microvascular immune-thrombosis during different stages of SARS-CoV-2 infection may have a "yet-to-be fully defined" impact on the development of post-acute sequela of COVID-19. Herein we initially reported a comprehensively characterized AC70 hACE2 Tg mouse model for SARS-CoV-2 infection and disease. We next demonstrated the subsequent onset of imbalanced coagulation and fibrinolysis pathways in infected Tg mice, focusing on dysregulated formation of ANXA2/S100A10 complexes, key coreceptors for t-PA that regulates membrane fibrinolysis, in which elevated production of autoantibodies against ANXA2 induced by SARS-CoV-2 might play an intriguing role. Taken together, we demonstrated that AC70 hACE2 Tg mice lethally challenged with SARS-CoV-2 recapitulated several features of COVID-associated coagulopathy observed in patients and highlighted the potential role of ANXA2 in this phenomenon. Thus, ANXA2 might serve as a potentially novel druggable target to attenuate COVID-19-associated thrombotic events.

microbiology↗

bioRxiv2023

A Connectome of the Male Drosophila Ventral Nerve Cord

Animal behavior is principally expressed through neural control of muscles. Therefore understanding how the brain controls behavior requires mapping neuronal circuits all the way to motor neurons. We have previously established technology to collect large-volume electron microscopy data sets of neural tissue and fully reconstruct the morphology of the neurons and their chemical synaptic connections throughout the volume. Using these tools we generated a dense wiring diagram, or connectome, for a large portion of the Drosophila central brain. However, in most animals, including the fly, the majority of motor neurons are located outside the brain in a neural center closer to the body, i.e. the mammalian spinal cord or insect ventral nerve cord (VNC). In this paper, we extend our effort to map full neural circuits for behavior by generating a connectome of the VNC of a male fly.

neuroscience↗

bioRxiv2023

Measuring chondrocyte viability of articular cartilage based on label-free two-photon microscopy and deep learning image analysis

ObjectiveChondrocyte viability (CV) is an important indicator of articular cartilage health. Two-photon excitation autofluorescence (TPAF) and second harmonic generation (SHG) microscopy provide a label-free method for imaging chondrocytes. In this study, we propose an automated assessment of CV using deep learning cell segmentation and counting based on acquired TPAF/SHG images. DesignLabel-free TPAF/SHG images of cartilage samples from rats and porcine were acquired using both commercial and home-built two-photon microscopes, respectively. TPAF/SHG images were merged to form RGB color images with red, green, and blue channels assigned to TPAF (two channels) and SHG signals, respectively. To make the training datasets for the deep learning networks, individual chondrocyte areas on the RGB color images were manually circled and live or dead chondrocytes were validated by using Calcein-AM and Ethidium homodimer-1 dye labeling. We first built a chondrocyte viability network (MCV-Net) using the Mask R-CNN architecture, which could provide individual segmented cellular areas with live or dead status. Wiener deconvolution preprocessing was added before the input of MCV-Net to improve the accuracy of the CV analysis, forming the Wiener deconvolution CV network (wMCV-Net). ResultsTraining (300 images) and test (120 images) datasets were built for rats and porcine cartilage respectively. Wiener deconvolution could improve the Peak Signal-to-Noise Ratio (PSNR) for 30-40%. We demonstrated that both MCV-Net and wMCV-Net significantly improved the accuracy of the CV measurement. ConclusionA custom desktop TPAF/SHG microscope was used in collaboration with deep learning algorithm wMCV-Net based label-free method to assess the CV and get 95% accuracy with both rats and porcine samples.

bioengineering↗

bioRxiv2022

Comparing 10x Genomics single-cell 3' and 5' assay in short-and long-read sequencing

Barcoding strategies are fundamental to droplet-based single-cell sequencing, and understanding the biases and caveats between approaches is essential. Here, we comprehensively evaluated both short and long reads of the cDNA obtained through the two marketed approaches from 10x Genomics, the "3 assay" and the "5 assay", which attach barcodes at different ends of the mRNA molecule. Although the barcode detection, cell-type identification, and gene expression profile are similar in both assays, the 5 assay captured more exonic molecules and fewer intronic molecules compared to the 3 assay. We found that 13.7% of genes sequenced have longer average read lengths and are more complete (spanning both polyA-site and TSS) in the long reads from the 5 assay compared to the 3 assay. These genes are characterized by long average transcript length, high intron number, and low expression overall. Despite these differences, cell-type-specific isoform profiles observed from the two assays remain highly correlated. This study provides a benchmark for choosing the single-cell assay for the intended research question, and insights regarding platform-specific biases to be mindful of when analyzing data, particularly across samples and technologies.

bioinformatics↗

bioRxiv2022

Predicting FDA approvability of small-molecule drugs

A high rate of compound attrition makes drug discovery via conventional methods time-consuming and expensive. Here, we showed that machine learning models can be trained to classify compounds into distinctive groups according to their status in the drug development process, which can significantly reduce the compound attrition rate. Using molecular structure fingerprints and physicochemical properties as input, our models accurately predicted which drug compounds would proceed to trial, with an area under the receiver operating curve (AUC) of 0.94 {+/-} 0.01 (mean {+/-} standard deviation). Our models also identified which drugs in clinical trials would be approved by the US Food and Drug Administration (FDA) to go on the market, with an AUC of 0.73 {+/-} 0.02. The predictive power of our models could reduce the attrition rate of preclinical compounds to enter clinical trials from 65%, as with conventional methods, to 12% (with 92% sensitivity) and the clinical trial failure rate from 80-90% to 29% (with 83% sensitivity). The results largely held in additional tests on new clinical trial compounds and new FDA-approved drugs, as well as on drugs uniquely approved for use in Europe and Japan. SIGNIFICANCE STATEMENTThe odds of developing a drug approved by the US Food and Drug Administration (FDA) are slim, meaning that the vast majority of drug candidates would fail tests for safety and efficacy in the drug discovery process, rendering it highly inefficient and costly. Here, we have developed machine learning models to predict drug compounds worthy of clinical trials with high accuracy, and clinical-trial compounds to receive FDA approval with a much higher success rate than that achieved by the traditional approach. Our computational prediction requires input of only the drug compounds chemical structure and physicochemical properties. It can help mitigate the long-standing problem of drug discovery.

bioinformatics↗

bioRxiv2022

A pan-variant mRNA-LNP T cell vaccine protects HLA transgenic mice from mortality after infection with SARS-CoV-2 Beta

Clinically licensed COVID-19 vaccines ameliorate viral infection by inducing vaccinee production of neutralizing antibodies that bind to the SARS-CoV-2 Spike protein to inhibit viral cellular entry (Walsh et al., 2020; Baden et al., 2021), however the clinical effectiveness of these vaccines is transitory as viral variants arise that escape antibody neutralization (Tregoning et al., 2021; Willett et al., 2022). Vaccines that solely rely upon a T cell response to combat viral infection could be transformational because they can be based on highly conserved short peptide epitopes that hold the potential for pan-variant immunity, but a mRNA-LNP T cell vaccine has not been shown to be sufficient for effective antiviral prophylaxis. Here we show that a mRNA-LNP vaccine based on highly conserved short peptide epitopes activates a CD8+ and CD4+ T cell response that prevents mortality in HLA-A*02:01 transgenic mice infected with the SARS-CoV-2 Beta variant of concern (B.1.351). In mice vaccinated with the T cell vaccine, 24% of the nucleated cells in lung were CD8+ T cells on day 7 post infection. This was 5.5 times more CD8+ T cell infiltration of the lungs in response to infection compared to the Pfizer-BioNTech Comirnaty(R) vaccine. Between days 2 and 7 post infection, the number of CD8+ T cells in the lung increased in mice vaccinated with the T cell vaccine and decreased in mice vaccinated with Comirnaty(R). The T cell vaccine did not produce neutralizing antibodies, and thus our results demonstrate that SARS-CoV-2 viral infection can be controlled by a T cell response alone. Our results suggest that further study is merited for pan-variant T cell vaccines, and that T cell vaccines may be relevant for individuals that cannot produce neutralizing antibodies or to help mitigate Long COVID.

immunology↗