Biology subjects

wang, s.

Publications and source records attributed to wang, s..

9 recordsLinked to original sources

bioRxiv2026

The U6 snRNA capping enzyme Amus/MePCE is required for the stability of U6atac and the processing of introns with features of both major and minor classes

Over 99% of the introns in humans are removed by the major spliceosome with the U6 snRNA as the catalytic component while the rest 1% by the minor spliceosome containing U6atac. Recent evidence supports mechanistic interactions between the two splicing pathways even though they have been considered as operating in parallel. Here we report a unique genetic condition in which both splicing pathways are partially impaired, and present evidence supporting inter-pathway interactions. We recently showed that the Drosophila Amus protein and its mammalian homolog MePCE serve as the methyltransferase of U6 and protect its stability. We show here that U6atac is greatly reduced under the same amus or mepce mutant backgrounds, suggesting Amus/MePCE serves the same protective function for U6atac. Interestingly, the processing of canonical minor introns is not significantly affected in the mutants, suggesting that a minimal amount of U6atac is sufficient for those reactions. In contrast, the splicing of a second class of minor spliceosome sensitive introns are inhibited. These introns harbor splice sites different from the minor consensus even though they interact favorably with the minor spliceosome as shown by others. Interestingly, most members of this class of minor-like introns are evolutionarily younger than their canonical counterparts, and our in-depth analysis of one such intron supports that some of them might be functionally neutral. Our study provides empirical evidence supporting the existence of a class of introns jointly regulated by the two splicing pathways, and revealing additional complexity in the functional evolution of minor splicing.

genetics↗

bioRxiv2026

A Phospho-Switch for Cell Fate Control

Cell fate control is thought to be complex, likely involving both cell intrinsic and extrinsic factors, and remains ill-defined at the molecular level. Here we show a phospho-switch controls cell fate both in vitro and in vivo. We first show that SALL4 is phosphorylated at multiple sites, but only T903 as the primary one as SALL4T903A lost BAF interaction and [~]90% activity in reprogramming. Mechanistically, we demonstrate that SALL4pT903 is sensitive to BMP4 signaling through DUSP9 axis. Finally, we show by tetraploid embryo complementation that mESCs harboring SALL4T903A can sustain embryo development but with severe defects including cranial hypoplasia and flattened skull vault after birth. A genome wide search identifies 608 transcription factors harboring the same HTG motif sandwiched by two zinc fingers, suggesting that this phospho-switch may be a conserved mechanism to control cell fate. In briefA phosphorylation switch, pT903, in SALL4 governs cell fate by modulating BAF complex interaction in response to BMP4-DUSP9 signaling, as its dephosphorylation abrogates reprogramming and causes cranial defects in mice, suggesting that such switches may broadly regulate cell fate machinery. HighlightsO_LIA T903 phospho-switch in SALL4 gates BAF interaction to orchestrate gene activation and reprogramming C_LIO_LIBMP4-DUSP9 signaling converges on SALL4pT903 dephosphorylation to couple extracellular cues with cell fate determination C_LIO_LIDisrupting the SALL4-T903 phospho-switch causes severe postnatal developmental defects in mice. C_LIO_LIThe HTGE motif may function as a central switch on key transcription factors for sensing upstream signals C_LI

developmental biology↗

bioRxiv2025

Ferritin Nanoparticle-Based Indirect ELISA for Immunodominant Region Screening and B-cell Epitope Validation of African Swine Fever Virus

African swine fever virus (ASFV) poses a severe threat to the global pig industry, creating an urgent need for effective vaccines and reliable diagnostic tools. While numerous ASFV B-cell epitopes have been reported based on murine-derived monoclonal antibodies (mAbs), such epitopes may not fully represent the antibody targets recognized in naturally infected pigs. Therefore, we need an approach to screen immunodominant regions by evaluating the reactivity of pre-defined epitopes with ASFV-infected pigs sera. In this study, we developed an indirect enzyme-linked immunosorbent assay (iELISA) based on recombinant epitope-ferritin nanoparticles (EFNs), as a platform for immunodominant region screening and epitope validation. This assay, hereafter referred to as EFNs-iELISA, shows an approximately twofold improvement in sensitivity compared with synthetic peptide-based ELISA for linear B-cell epitopes, and improved discrimination over GST-His-peptide and BSA-conjugated peptide formats. Using EFNs-iELISA, we evaluated 22 known murine-derived B-cell epitopes and 17 predicted epitopes from ASFV proteins CD2v, P30, P54, and P72 with ASFV-positive and negative porcine sera. According to the calculated cut-off threshold, tested epitopes were classified into different groups with positive or negative reactivity. Our analysis confirmed that among 22 known murine-derived ASFV B-cell epitopes, 16 (72%) also showed positive reactions with porcine-derived sera. Whats more, we also screened 7 new porcine-specific immunoreactive regions on these proteins. The EFNs-iELISA method provides a practical tool for screening immunodominant regions of ASFV, with potential applications in vaccine and diagnostic research.

immunology↗

bioRxiv2022

Genome-scale mRNA transcriptomic insights into the mature growth dynamic of Cigar tobacco leaves

The harvest time is a key factor for cigar leaves with high quality, which varies greatly depending on the environment. Here, we performed a genome-scale mRNA transcriptomic analysis on the cigar cultivar CX-26 (Nicotiana tabacum L.) to evaluate the relationship between gene expression and growth state. The leaves were harvested with 67 (T1), 71(T2) and 75 (T3)-day growth. A total of 80,502 genes were detected in the CX-26 leaves, of which 64,611 genes were annotated on the reference genome. Principal component analysis showed that T1 and T2 leaves had a high overlapping pattern, while T3 leaves were distinct. Indeed, T1 and T2 leaves had fewer differential expressing genes (DEGs), while T3 leaves had 26,456 DGEs from T2 leaves, supporting the distinct growth of T3 leaves. GO annotations mainly enriched the photosynthesis-related metabolic process, catalytic activity and binding biological processes. KEEG analysis identified the key pathways including photosynthesisantenna proteins, plant hormone/MAPK signaling pathway and plant-pathogen interaction. The maturity regulation and defense response-associated hormones abscisic acid and jasmonate acid were higher in the T3 leaves than that in T1 and T2 leaves confirming the KEEG analysis. Furthermore, several photosynthesis-related enzymes and a transcription factor were highlighted in the gene regulatory network, which might regulate the dynamics of carbohydrate metabolism, lipid metabolism and energy metabolism. In summary, our study provides insight into the growth state of CX-26 cigar leaves.

plant biology↗

bioRxiv2020

DrugOrchestra: Jointly predicting drug response, targets, and side effects via deep multi-task learning

Massively accumulated pharmacogenomics, chemogenomics, and side effect datasets offer an unprecedented opportunity for drug response prediction, drug target identification and drug side effect prediction. Existing computational approaches limit their scope to only one of these three tasks, inevitably overlooking the rich connection among them. Here, we propose DrugOrchestra, a deep multi-task learning framework that jointly predicts drug response, targets and side effects. DrugOrchestra leverages pre-trained molecular structure-based drug representation to bridge these three tasks. Instead of directly fine-tuning on an individual task, DrugOrchestra uses deep multi-task learning to obtain a phenotype-based drug representation by simultaneously fine-tuning on drug response, target and side effect prediction. By coupling these three tasks together, DrugOrchestra is able to make predictions for unseen drugs by only knowing their molecular structures. We constructed a heterogeneous drug discovery dataset of over 21k drugs by integrating 8 datasets across three tasks. Our method obtained significant improvement in comparison to methods that were trained on a single task or a single dataset. We further revealed the transferability across 8 datasets and 3 tasks, providing novel insights for understanding drug mechanisms. Availabilityhttps://github.com/jiangdada1221/DrugOrchestra

bioinformatics↗

bioRxiv2020

Constructing early diagnosis model of colorectal cancer based on expression profile

purposeIn order to break through the restrictive factors such as the fecal occult blood test (FOBT) in the routine detection of bowel cancer, which is susceptible to diet and drugs, and the high cost and inconvenience of microscopy, Seeking a possible FOBT alternative. MethodsAn error back propagation neural network (BPNN) algorithm was used to construct a CRC diagnosis model based on expression profiles. ResultsThe accuracy of the model on the training and test sets is 0.943 and 0.935, respectively. AUC all reached above 0.95. ConclusionThe CRC molecular detection model based on expression profiles provides a possible alternative to FOBT. It provides a new approach and method for the clinical diagnosis of bowel cancer.

genetics↗

bioRxiv2019

Time-ResNeXt for epilepsy recognition based on EEG signals

ObjectiveTo automatically detect dynamic EEG signals to reduce the time cost of epilepsy diagnosis. In the signal recognition of electroencephalogram (EEG) of epilepsy, traditional machine learning and statistical methods require manual feature labeling engineering in order to show excellent results on a single data set. And the artificially selected features may carry a bias, and cannot guarantee the validity and expansibility in real-world data. In practical applications, deep learning methods can release people from feature engineering to a certain extent. As long as the focus is on the expansion of data quality and quantity, the algorithm model can learn automatically to get better improvements. In addition, the deep learning method can also extract many features that are difficult for humans to perceive, thereby making the algorithm more robust. MethodBased on the design idea of ResNeXt deep neural network, this paper designs a Time-ResNeXt network structure suitable for time series EEG epilepsy detection to identify EEG signals. ResultsThe accuracy rate of Time-ResNeXt in the detection of EEG epilepsy can reach 90.50%. ConclusionThe Time-ResNeXt network structure produces extremely advanced performance on the benchmark dataset (Berne-Barcelona dataset), and has great potential for improving clinical practice.

bioinformatics↗

bioRxiv2018

MiR-505-3p is a Repressor of the Puberty Onset in Female Mice

Puberty onset is a complex trait regulated by multiple genetic and environmental factors. In this study, we narrowed a puberty related QTL down to a 1.7 Mb region on chromosome X in female mice and inferred miR-505-3p as the functional gene.\n\nWe conducted ectopic expression of miR-505-3p in the hypothalamus of prepubertal female mice through lentivirus-mediated orthotopic injection. The impact of miR-505-3p on female puberty was evaluated by the measurement of pubertal events and histological analysis. The results showed that female mice with overexpression of miR-505-3p in the hypothalamus manifested later puberty onset timing both in vaginal opening and ovary maturation, followed by weaker fertility lying in the longer interval time between mating and delivery, higher abortion rate and smaller litter size. We also constructed miR-505-3p knockout mice by CRISPR/Cas9 technology. MiR-505-3p knockout female mice showed earlier vaginal opening timing, higher serum gonadotrophin and higher expression of puberty-related gene, as well as its target gene Srsf1 in the hypothalamus than their wild type littermates.\n\nSrsf1 was proved to be the target gene of miR-505-3p that played the major role in this process. The results of RIP-seq (RNA Immunoprecipitation-sequencing) showed that SF2, the protein product of Srsf1 gene, mainly bound to ribosome protein (RP) mRNAs in GT1-7 cells. The collective evidence implied that miR-505-3p/SRSF1/RP could play a role in the sexual maturation regulation of mammals.\n\nAuthor summaryThe puberty onset in mammals is a vital biological process that signals the acquisition of reproductive capacity. The initiation of puberty is triggered by the activation of hypothalamic pulsatile GnRH surge. The dysregulation of pubertal development shows relevance to later health risks of type 2 diabetes, cardiovascular disease, breast cancer and other health disorders. Recent progress indicates that a lot of genes play a role in the excitatory or inhibitory regulation of GnRH release. However, the detailed pathway of pubertal timing remains unclear. Our previous studies isolated an X-linked QTL that was associated with the timing of puberty in mice. In this study, we proved that miR-505-3p was a female puberty onset regulator based on data from positional cloning, ectopic expression and knockout mouse models. We also assigned Srsf1 as the functional target gene of miR-505-3p underlying this process. The results of RIP-seq showed that SF2, the protein of Srsf1 gene, preferential bound to ribosome protein (RP) mRNAs in GT1-7 cells. We propose that miR-505-3p/SF2/RP could play a role in the sexual maturation regulation of mammals.

genetics↗

bioRxiv2018

Single-cell level transcriptome of the maize pathogenic fungi cochliobolus heterostrophus race O in infection reveal the virulence related genes, and potential circRNA effector

Cochliobolus heterostrophus is a crucial pathogenic fungus that causes southern corn leaf blight (SCLB) in maize worldwide, however, the virulence mechanism of the dominant race O remains unclear. In this report, the single-cell level of pathogen tissue at three infection stages were collected from the host interaction-situ, and were performed next-generation sequencing from the perspectives of mRNA, circular RNA(circRNA) and long noncoding RNA(lncRNA). In the mRNA section, signal transduction, kinase, oxidoreductase, and hydrolase, et al. were significantly related in both differential expression and co-expression between virulence differential race O strains. The expression pattern of the traditional virulence factors nonribosomal peptide synthetases (NPSs), polyketide synthases (PKSs) and small secreted proteins (SSPs) were multifarious. In the noncoding RNA section, a total of 2279 circRNAs and 169 lncRNAs were acquired. Noncoding RNAs exhibited differential expression at three stages. The high virulence strain DY transcribed 450 more circRNAs than low virulence strain WF. Informatics analysis revealed numbers of circRNAs which positively correlate with race O virulence, and a cross-kingdom interaction between the pathogenic circRNA and host miRNA was predicted. An important exon-intron circRNA Che-cirC2410 combines informatics characteristics above, and highly expressed in the DY strain. Che-cirC2410 initiate from the pseudogene chhtt, which doesnt translate genetic code into protein. In-situ hybridization tells the sub-cellular localization of Che-cirC2410 include pathogen`s mycelium, periplasm, and the diseased host tissues. The target of Che-cirC2410 was predicted to be zma-miR399e-5P, and the interaction between noncoding RNAs was proved. More, the expression of zma-miR399e-5P exhibited a negative correlation to Che-cirC2410 in vivo. The deficiency of Che-circ2410 decreased the race O virulence. The host resistance to SCLB was weakened when zma-miR399e-5P was silenced. Thus, a novel circRNA-type effector and its resistance related miRNA target are proposed cautiously in this report. These findings enriched the pathogen-host dialogue by using noncoding RNAs as language, and revealed a new perspective for understanding the virulence of race O, which may provide valuable strategy of maize breeding for disease resistance.\n\nAuthor SummaryThe southern corn leaf blight (caused by Cochliobolus heterostrophus) is not optimistic in Asia, however we have limit knowledge about the infection mechanism of the dominant C.heterostrophus race O. We take full advantage of the ideal C.heterostrophus genome database, laser capture microdissection and single-cell level RNA sequencing. Hence, we could avert the artificial influence such as medium, and profile the real gene mobilization strategy in the infection. The results of coding RNA section were accessible, virulence related genes (such as the signal transduction, PKS, SSP) were detected in RNA-seq,which accord with previous reports. However, the results of noncoding RNA was astonished, 2279 circular RNAs (circRNA) and 169 long noncoding RNAs (lncRNA) were revealed in our results. Generally, the function of noncoding RNA was hypothesized in single species, but we boldly guess that the function of circRNA is rather complicated in the pathogen-host interaction. Finally, the circRNA in-situ hybridization (ISH) demonstrate the secretion of pathogen circRNA into the host tissue. By bioinformatic prediction, we found a sole microRNA target, and proved the interaction between circRNA and microRNA. These findings are likely to reveal a novel pathogen effector type: secreted circRNA.

microbiology↗