Biology subjects

Yamamoto, H.

Publications and source records attributed to Yamamoto, H..

At least 19 recordsLinked to original sources

bioRxiv2026

Information coding strategies in biological reservoir computing for multiclass time-series classification

Physical reservoir computing (RC) using biological neural networks (BNNs) offers a framework for probing how high-dimensional dynamics arising from multicellular systems support complex information processing in the brain. However, how electrical inputs and extracellular readouts should be designed to suit different task demands remains unclear. Here, using primary rat cortical neurons cultured on a high-density microelectrode array, we investigated how input encoding and spike detection methods influence neural representations and the resulting RC performance in BNNs. Comparison of pulse amplitude modulation and latency coding revealed that the latter exhibited superior performance in demanding multiclass conditions, including a 10% improvement in a 10-class spoken digit classification task. Spike sorting further improved classification performance by enhancing class separability, while largely preserving the low-dimensional structure of population activity. These findings highlight the importance of designing information coding strategies according to task difficulty.

neuroscience↗

bioRxiv2026

Handwritten Digit classification with neural cultures is influenced by neural architecture, network dynamics, and decoding methods

As silicon-based computing approaches fundamental physical limits, neurocomputing offers an energy-efficient alternative by leveraging the intrinsic non-linear dynamics of biological systems. To harness these dynamics, it is vital to understand the structure-function relationship governing how neural cultures process complex spatio-temporal information and how to appropriately decode the resulting neural electrophysiological activity. We investigated this utilizing a closed-loop electrophysiology platform, the CL1, to implement reservoir computing in human iPSC-derived neuronal networks. To systematically evaluate the variables driving neurocomputational capacity, we explored how cellular composition (cortical vs. hippocampal lineages), and the physical architecture (unstructured monolayers, 3D neural organoids, and modular networks confined by microfluidic devices) influenced electrophysiological properties and interacted with different decoding methodologies. Using a spatio-temporal version of a handwritten digit pattern recognition task (MNIST), we analyzed how these biological and analytical factors influenced classification accuracy. To ensure robust interpretation this required us to first demonstrated that reservoir computing decoding methods require strict artifact control and trial-based cross-validation to distinguish network computation from artifactual signal separability or temporal data leakage. Applying this validated frequency-domain pipeline, we suggest a clear functional hierarchy where structural modularity acts as a vital functional regularizer. Modular cortical cultures significantly outperformed unconstrained monolayers and organoids on MNIST. Furthermore, decoding frequency information from raw signals proved superior to typical time-bin decoding implementations. These findings establish that maximizing the computational potential of Synthetic Biological Intelligence, while avoiding false positives, requires a synergistic optimization of cellular identity, structural governance, and rigorous decoding logic. In doing so, this work provides a critical base establishing the criteria under which to evaluate neurocomputing implementations.

neuroscience↗

bioRxiv2026

OAC-PCA: orthogonal adjustment of confounding effects in principal component analysis for metabolomics data mining

Principal component analysis (PCA) is widely used in mass spectrometry-based metabolomics for exploratory data mining. Statistical testing of loading values can extract metabolite features associated with score patterns, but this approach requires principal components (PCs) to remain orthogonal while loadings are defined as correlation coefficients between PC scores and variables. Adjustment for Confounding PCA (AC-PCA) was previously developed to explore biologically meaningful components from data matrices affected by biological and technical confounders. However, AC-PCA does not simultaneously ensure PC orthogonality and a correlation-coefficient definition of loadings, limiting the statistical interpretation of its loadings. Here, we reformulated AC-PCA as Orthogonal Adjustment for Confounding effects in PCA (OAC-PCA). In OAC-PCA, PCs remain orthogonal, and loadings retain this correlation-coefficient interpretation. These properties enable statistical testing of metabolite associations while accounting for confounding effects.

bioinformatics↗

bioRxiv2026

Modeling interregional propagation of α-synuclein in human striatal-midbrain assembloids

Parkinsons disease is characterized by the progressive degeneration of dopaminergic (DA) neurons in the substantia nigra that project to the striatum. Postmortem studies have found that -synuclein pathology spreads across brain regions in a prion-like manner. Human three-dimensional neural tissues provide experimentally tractable platforms for clarifying the molecular pathology of synucleinopathy that overcome the limitations of postmortem brain studies. In this study, we established a human striatal-midbrain assembloid model enabling spatially and temporally controlled -synuclein induction after formation of the nigrostriatal pathway within hStrMAs. Human induced pluripotent stem cells (hiPSCs) were differentiated into striatal spheroids and midbrain spheroids, which were fused to reconstruct the nigrostriatal pathway in vitro. Histological, electrophysiological, and single-cell RNA sequencing analyses confirmed the generation of region-specific neuronal populations, including striatal GABAergic neurons and midbrain DA neurons, and demonstrated that distinct forebrain and ventral midbrain identities were preserved within assembloids. To model synucleinopathy, we generated a doxycycline-inducible -synuclein hiPSC line, permitting selective induction of -synuclein expression in the striatal region after assembloid formation. Spatially restricted -synuclein induction resulted in the progressive accumulation of phosphorylated and fibrillar -synuclein in midbrain DA neurons, indicating interregional propagation through established neural circuits. Transcriptomic analysis of the midbrain region revealed significant downregulation of DA identity and synaptic function genes, suggesting functional impairment of DA neurons following -synuclein propagation. These findings demonstrate that the human neural circuit architecture contributes to the interregional propagation of -synuclein pathology. Our controllable human assembloid platform provides a powerful experimental framework for dissecting the mechanisms of pathological protein spread and developing therapeutic strategies targeting molecular pathology in neurodegenerative diseases. Significance statementOne pathological feature of Parkinsons disease is the progressive spread of -synuclein pathology across brain regions. However, the molecular mechanisms underlying this process remain poorly understood. In this study, we established a human striatal-midbrain assembloid model in which -synuclein expression can be spatially restricted and temporally induced after the nigrostriatal pathway is formed, thus enabling the experimental dissection of interregional -synuclein propagation across brain regions. We demonstrated that -synuclein pathology initiated in the striatal region propagated to midbrain dopaminergic neurons, leading to disease-relevant molecular and transcriptional alterations. These findings provide direct evidence that human neural circuits contribute to the spread of synucleinopathy and offer a controllable human model for studying disease progression and therapeutic strategies.

neuroscience↗

bioRxiv2026

The strictly conserved GGG-tracts in the 5' and 3' long terminal repeat of HIV-1 are critical to control multiple steps of HIV-1 replication to prevent acquisition of unwanted mutations in the region

Three consecutive deoxyguanosine residues (the GGG-tract) in the U3/R junction of the 5 long terminal repeat (LTR) are strictly conserved among all HIV-1 subtypes. Each deoxyguanosine within the tract has been reported to function as transcription initiation site for HIV-1 RNA. Furthermore, RNAs whose transcription initiates from the third deoxyguanosine in the tract (1G form RNAs) are predominant in virus particles and serve as the primary templates for reverse-transcription. In this study, we generated mutant HIV-1s by replacing the tracts in both the 5 and 3 LTR with other nucleotides to elucidate their functional significance. We identified several proviral sequences containing unexpected mutations near the 5 tract after infection with the mutant, but not the wild-type virus. Five-prime rapid amplification of cDNA end (5 RACE) analyses of RNAs purified from mutant virus particles revealed multiple RNA variants with 5 terminal sequences differing from the plasmid used for producing the particles. Some of the unexpected proviral sequences likely arise directly from these variants during reverse-transcription. We also found that replacing all three nucleotides in the 3 tract with deoxyadenosines decreased the proportion of the 1G form RNAs in particles to 32.6%. Nevertheless, up to 88% of provirus was likely generated with the 1G form RNAs, although they were no longer absolutely predominant in particles. Our results demonstrate that the GGG-tracts in the 5 and 3 LTR are conserved to maintain the integrity of reverse-transcription for LTR sequence generation by controlling multiple steps of HIV-1 replication, including transcription, RNA packaging and reverse-transcription. ImportanceHIV-1 is highly mutable, yet certain conserved sequences remain consistent across most strains. These sequences are thought to be maintained because mutations in these regions typically impair viral fitness, leading to the elimination of such variants through natural selections. This study suggests that HIV-1 has a unique mechanism to autonomously prevent acquisition of mutations in certain regions. Specifically, the GGG-tracts in the 5 and 3 LTR ensure accurate transcription of HIV-1 RNAs, selective packaging of the 1G form RNAs, and preferential use of these 1G forms as a template for reverse-transcription. GGG preservation minimizes unwanted mutations in this region, allowing progeny virus to inherit intact LTR sequences. Interestingly, the preferential usage of 1G form RNAs as template for reverse-transcription is not due to their predominance in virus particles. This suggests that genome packaging independently contributes to reverse-transcription as preparatory stage by concentrating the most suitable RNAs for reverse-transcription in particles.

microbiology↗

bioRxiv2026

A Knock-In Igfn1iCre transgenic mouse line provides partial developmental access to type-7 bipolar cells

1.Functional neuronal circuits in the vertebrate retina emerge through coordinated developmental events, yet the timeline by which bipolar cells acquire visual feature selectivity remains unclear. A major barrier is the limited genetic access to bipolar subtypes during early postnatal stages. Recent comprehensive transcriptomic study points to Igfn1 as a molecular marker for type-7 bipolar cells (BC 7), a subtype that exhibits direction-selective glutamate releases in adult. Here, we generated an Igfn1iCre knock-in mouse line and characterized Igfn1-positive cell morphology from postnatal day 4(P4) to adult using Cre-dependent tdTomato reporter mice. We found Igfn1-positive cells in the inner retina by P12-P15, predominantly labelling bipolar cells and some amacrine populations. At P15, about 71% of labelled bipolar cells stratified their axons in the S4 sublamina of the inner plexiform layer, consistent with BC 7 morphology. In adult retina, the widespread Igfn1-labelling appears slightly dominated in amacrine cells. To validate these observations, we analysed Igfn1 expression in the Mouse Retina Cell Atlas and confirmed strong Igfn1 enrichment in BC 7 and expression in additional retinal cell types, mirroring experimental results. Overall, these results reveal Igfn1iCre as a potential developmental tool for BC 7 access in the retina.

neuroscience↗

bioRxiv2026

Predictive E-prop: A biologically inspired approach to train predictive coding-based recurrent spiking neural networks

The predictive coding framework offers a compelling model for temporal signal processing in the cortex. Recent studies explored its implementation in spiking architectures using Hebbian plasticity rules or offline learning; however, a biologically inspired model that enables gradient-based minimization of prediction errors remains an open challenge. In this work, we demonstrate that the predictive coding objective can be optimized using the online and local nature of the e-prop learning algorithm in recurrent spiking neural networks, creating the Predictive E-prop model. We demonstrate that the model is capable of learning complex time-series signals purely from self-supervised learning, using only its own prediction error as input, maintaining self-sustaining activity and reproducing the targets underlying dynamics even in the absence of external stimuli. Furthermore, Predictive E-prop shows robust signal reconstruction abilities, effectively filtering noise and successfully interpolating sparse data. A comparative study against a backpropagation-based approach reveals that the two achieve comparable performance after training, confirming the viability of our model for timeseries generation tasks. These findings are particularly relevant for future developments in neuromorphic hardware, offering a purely self-supervised, gradient-based model that could provide significant advantages in power efficiency and computational ability.

neuroscience↗

bioRxiv2026

Chemogenetic activation induces persistent transcriptomic remodeling and functional plasticity in cultured astrocytes

Astrocytes are essential regulators of neuronal function and brain homeostasis. Recent methodological advances have increasingly revealed their active roles through precise manipulation of astrocyte function in the brain. While chemogenetic strategies, such as Designer Receptors Exclusively Activated by Designer Drugs (DREADDs), have been widely used to recapitulate endogenous G-protein-coupled receptors (GPCRs) signaling and selectively modulate astrocyte activity, the molecular mechanisms accompanying DREADD activation remain incompletely understood. Here, we characterized the transcriptional responses of cultured cortical astrocytes and found that DREADD activation induces profound changes in astrocytic gene expression profiles. Activation of Gi-, Gq-, and Gs-coupled DREADDs elicited distinct transcriptomic responses. Transcription factor activity profiling further revealed that each DREADD subtype selectively modulates a distinct set of transcription factors that collectively shape astrocytic transcriptomic responses, thereby indicating subtype-specific transcriptional mechanisms. Importantly, DREADD activation induced persistent changes in the astrocytic transcriptome that remained detectable even three days after stimulation. These transcriptomic alterations were associated with sustained changes in astrocytic calcium responses to extracellular ATP, indicating plastic changes in astrocytic physiological function. Together, these findings provide a molecular framework for interpreting astrocytic DREADD manipulations and reveal a mechanistic basis for functional plasticity and heterogeneity of astrocytes.

neuroscience↗

bioRxiv2026

Heterogeneous Nuclear Ribonucleoprotein A1 as a Key Regulator in Pulmonary Arterial Hypertension Development

BackgroundAlthough plexiform lesion (PL) formation in severe pulmonary arterial hypertension (PAH) is a therapeutic target, the mechanisms underlying their formation have not been fully elucidated. This study aimed to identify candidate proteins involved in PL formation by examining differentially expressed proteins (DEPs) in PAH lesions. MethodsProteomic remodeling was assessed before and after the formation of PLs in a SU5416 combined with hypoxia (SuHx) rat model of severe PAH using laser-capture microdissection coupled with mass spectrometry. Unobstructed pulmonary arteries with medial hypertrophy (UMHPAs) and PLs from SuHx rats were subjected to qualitative and quantitative proteomics, revealing DEPs between these structures. ResultsWe identified 718 proteins with 58 DEPs, of which 31 were upregulated in UMHPAs and 27 were upregulated in PLs. Immunostaining confirmed that DEPs detected in our proteomic analysis were differentially expressed between UMHPAs and PLs. Among them, we focused on heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) as a candidate protein that may be strongly associated with PL formation because of its strong association with cell proliferation. siRNA knockdown of hnRNPA1 in hypoxia-treated pulmonary artery smooth muscle cells reduced pyruvate kinase M2 expression and significantly decreased proliferative capacity. Treatment of SuHx rats with hnRNPA1 inhibitors also suppressed pulmonary artery remodeling in PAH and PL formation. ConclusionsSeveral DEPs associated with PL formation but with unclear relevance to pulmonary artery remodeling in PAH were discovered. Among these, hnRNPA1, which was not detected in transcriptome analysis, may be important in PL formation. What are the Clinical Implications?The prognosis for pulmonary arterial hypertension (PAH) with plexiform lesions (PLs) is poor. Although PLs are therapeutic targets, the mechanisms of PL formation have not been fully elucidated, and there have been no direct examinations of proteomic remodeling before and after their formation. We identified several differentially expressed proteins (DEPs) that may be strongly associated with PL formation. Of these, heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) was examined as a candidate protein involved in PA remodeling in PAH and PLs. Inhibition of hnRNPA1 in a severe PAH model not only suppressed pulmonary artery remodeling in PAH but also reduced PL formation. Several DEPs involved in pulmonary artery remodeling in PAH may be of significant value in the search for new therapeutic targets.

cell biology↗

bioRxiv2025

Catalytic regulation of CoA transferase by an NAD+-sensing accessory protein and protein acetylation

CoA transferases play essential roles in short-chain fatty acid metabolism by catalyzing the reversible transfer of CoA groups between acyl substrates. However, how their activities are regulated in response to cellular metabolic states remains unclear. Here we identify a dual regulatory mechanism of a CoA transferase from Thermus thermophilus, which is controlled by both interaction with an NAD+-sensing accessory protein and protein acetylation. The enzyme associates with an alanine dehydrogenase-like protein that lacks catalytic activity but binds NAD+ and NADH. Biochemical analyses revealed that inhibition occurs specifically upon NAD+ binding, but not NADH, indicating that the alanine dehydrogenase-like protein senses the intracellular NAD+/NADH ratio. Cryo-electron microscopy structures of the complex of CoA transferase and alanine dehydrogenase-like protein reveal the structural basis of this redox-dependent inhibition. Furthermore, acetylation of the CoA transferase alleviates the inhibition. Since the NAD+/NADH ratio and acetyl-CoA levels reflect the cellular energy and metabolic states, these findings uncover a previously unrecognized regulatory link through which redox sensing and acetylation cooperate to fine-tune the {beta}-oxidation flux.

biochemistry↗

bioRxiv2025

Structural basis for the folding of PINK1 by the HSP90-CDC37 chaperone complex

PTEN-induced kinase 1 (PINK1) is a mitochondrial serine/threonine kinase that plays a central role in Parkin-dependent mitophagy. Mutations in PINK1 are associated with familial Parkinsons disease. PINK1 is a high-affinity client of the HSP90-CDC37 complex and is stabilized by this chaperone system. However, the molecular mechanism by which HSP90-CDC37 facilitates the folding of PINK1 remains unclear. Here, we present a cryogenic electron microscopy structure of the human PINK1-HSP90-CDC37 complex. The {beta}5 strand of the PINK1 N-lobe is accommodated in the central channel of the HSP90 dimer, which holds the PINK1 kinase domain in a partially unfolded state. The C-lobe and unique C-terminal extension (CTE) of PINK1 is folded. HSP90 covers the CTE of PINK1, which overlaps with interaction sites for TOM5, TOM20, and the PINK1 N-helix. The HPNI motif of CDC37 interacts with the C-lobe of PINK1, mimicking the HPNI motif in the N-lobe. The pathogenic mutation L347P is suggested to disrupt these interactions, while H271Q is located within the HPNI motif in the N-lobe of PINK1. These findings provide structural insights into the folding of PINK1 and its dysfunction in Parkinsons disease.

molecular biology↗

bioRxiv2025

The caninized anti-canine PD-1 monoclonal antibody in canine oral malignant melanoma: Efficacy and exploratory biomarker analysis

BackgroundCanine oral malignant melanoma (OMM) is a highly aggressive tumor with few treatment options. Because of its biological similarity to human mucosal melanoma, canine OMM represents a valuable spontaneous model for translational immunotherapy studies. Anti-PD-1 antibody therapy has shown promise in canine OMM; however, predictive biomarkers for treatment response and survival have not been identified. MethodsWe conducted a multicenter, prospective, investigator-initiated clinical trial to evaluate the safety and efficacy of a caninized anti-canine PD-1 monoclonal antibody (ca-4F12-E6) in 150 dogs with advanced OMM. The dogs were administered 3 mg/kg ca-4F12-E6 intravenously every two weeks. Treatment response was assessed using cRECIST v1.0. Biomarker analyses included peripheral blood parameters, cytokine/chemokines, peripheral lymphocyte subsets, microsatellite instability (MSI), immunohistochemistry for immune cell and mismatch repair protein markers, and RNA sequencing of the tumor tissue. Associations with clinical outcomes were determined by logistic regression and Cox proportional hazard models. ResultsThe best overall response rate was 16.7%. Treatment-related adverse events occurred in 40.0% of the dogs, which were primarily grade 1-3. Increased baseline white blood cell, neutrophil count, and C-reactive protein (CRP) levels were significantly associated with poor response, shorter progression-free survival (PFS), and reduced overall survival (OS). MSI-high tumors were associated with significantly prolonged OS compared with MSI-low/microsatellite stable tumors. Transcriptome analysis revealed differentially expressed genes and enriched immune-related pathways in responders, though limited by sample size. Conclusionca-4F12-E6 exhibited durable antitumor activity and an acceptable safety profile in dogs with OMM. Baseline systemic inflammatory markers and MSI status may serve as predictive biomarkers for clinical outcomes. The results support the use of canine OMM as a comparative model for human immuno-oncology and biomarker discovery.

cancer biology↗

bioRxiv2025

Online supervised learning of temporal patterns in biological neural networks under feedback control

In vitro biological neural networks (BNNs) provide a well-defined model system to constructively investigate how living cells interact with their environment to shape high-dimensional dynamics that could be used to generate a coherent temporal output, such as those required for motor control. Here, we developed a real-time closed-loop BNN system capable of generating periodic and chaotic temporal signals by integrating cultured cortical neurons with microfluidic devices and high-density microelectrode arrays. We show that training a simple linear decoder with fixed feedback weights enables the system to learn and autonomously generate diverse temporal patterns. When feedback was switched on, irregular activity in BNNs is transformed into low-dimensional, structured dynamics, producing coherent trajectories characterized by stable transitions between neural states. BNNs trained on different target frequencies--ranging from 4 to 30 s--could be trained to sustain oscillations at distinct frequencies, demonstrating their adaptability. Importantly, a top-down control of self-organized network formation with microfluidic devices is the key to suppress excessive synchronization and increase dynamical complexity in BNNs, facilitating the training and robust output generation. This work offers a biologically inspired platform for understanding the physical basis of cortical computation and for advancing energy-efficient neuromorphic computation. Significance StatementReservoir computing is a machine learning paradigm that exploits the transient dynamics of high-dimensional nonlinear systems. Although it was originally inspired by the mammalian brain and widely explored in physical systems, its implementations in biological neural networks (BNNs) have been limited due to their excessive connectivity and global synchrony in vitro. Here, we use microfluidic devices to construct modular, nonrandomly connected BNNs and integrate them with microelectrode arrays in a closed-loop reservoir computing environment. We show that the system can be trained to autonomously output various temporal signals, with the modular connectivity that is essential for learning. In vitro BNNs provide unique alternatives for physical reservoirs with dynamic adaptability.

neuroscience↗

bioRxiv2025

MICRORNAS REGULATED BY PREGNANCY TARGET THE HIV INTERACTOME

Innate immunity predictors of HIV risk are influenced by reproductive hormone levels, pregnancy, and lactation status. However, the molecular mechanisms underlying these associations remain unclear. MicroRNAs (miRNAs), as post-transcriptional regulators, play key roles in immune regulation and host-virus interactions. We hypothesized that physiological adaptations to pregnancy include the modulation of systemic miRNA expression, which in turn regulates host genes involved in HIV interaction, potentially influencing susceptibility during pregnancy. To test this, we leveraged a large longitudinal cohort from Uganda and Zimbabwe and analyzed 174 serum samples from 88 participants in pre-pregnancy (PP), pregnancy (P), and breastfeeding (BF) states using the HTG EdgeSeq platform. Differentially expressed (DE) miRNAs in pregnancy were identified with false discovery rate (FDR) < 0.1 (29 upregulated and 131 downregulated) by intersecting pairwise comparisons (P vs. PP and P vs. BF). Validated gene targets (2,733) were identified via miRWalk and modular enrichment analysis performed using Cytoscape/ClueGO. Enriched pathways (FDR<0.05) included Adaptive Immune Response, Hippo Signaling, Cellular Senescence, HSV-1 Infection, and two cancer-related pathways. Genes from pregnancy-enriched pathways overlapped with the known HIV-host interactome at the range of 37 to 88%. From this overlap, the Maximal Clique Centrally (MCC) score (cytoHubba) identified 47 unique hub genes acting as essential regulatory nodes of protein-protein interaction (PPI) network. These hub genes were further examined by their expression patterns predicted by DE miRNAs, and explored for their interactions with the HIV interactome, revealing connectivity with 18 HIV proteins - highest with Tat and gp120 - and impact on the HIV replication process. HLA-A was the most connected host gene, followed by BCL2L1, EIF2AK2, PTEN, and CYCS. These findings support the hypothesis that pregnancy-driven systemic miRNAs may shape HIV susceptibility by regulating hub genes with central role in viral evasion of host immunity. GRAPHIC ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=138 HEIGHT=200 SRC="FIGDIR/small/665982v1_ufig1.gif" ALT="Figure 1"> View larger version (17K): org.highwire.dtl.DTLVardef@1cd2007org.highwire.dtl.DTLVardef@1f1b9d3org.highwire.dtl.DTLVardef@dee856org.highwire.dtl.DTLVardef@6796c2_HPS_FORMAT_FIGEXP M_FIG C_FIG

molecular biology↗

bioRxiv2025

MICRORNAS AS REGULATORS OF DRUG METABOLISM AND TRANSPORT IN PREGNANT AND LACTATING WOMEN

BackgroundPhysiological changes during pregnancy result in altered maternal drug metabolism that impacts efficacy and safety of therapeutics in pregnant women. Pregnancy induced hormonal, immunologic, or metabolic changes may also influence and alter drug disposition. Despite research efforts focused on pharmacokinetics of medications used in pregnant women in the past decade, knowledge gaps exist in understanding how pregnancy influences drug disposition and placental drug transporters. Moreover, there is a scarcity of research in understanding the safety and effectiveness of therapeutics in lactating women. This study aimed to determine the effect of pregnancy on levels of miRNAs regulating drug metabolizing enzymes and transporters (DMET). MethodsWe utilized longitudinal serum specimens collected in 3-month intervals from 88 women who became pregnant during follow-up in a large prospective study of hormonal contraception and HIV acquisition in Uganda and Zimbabwe. We used the HTG EdgeSeq platform coupled with Illumina sequencing to obtain the global miRNA transcriptome in paired specimens collected before, during and after pregnancy. To identify differentially expressed (DE) miRNAs that distinguish pregnancy from pre-conception or breastfeeding we used mixed effect model accounting for multiple samples from the pregnancy event and controlling for fixed effects of batch, country, Nugent score category and sexually transmitted infections. To identify hormonally regulated miRNAs independently associated with Box-Cox-transformed levels of progesterone (P4), {beta}-estradiol (E2), and sex-hormone binding protein (SHBG) we controlled in addition for age, pregnancy and breastfeeding P-values were corrected using the Benjamini-Hochberg false discovery rate (FDR). DMET-targeting miRNAs were identified using miRTarBase focusing on interactions verified by 3-UTR luciferase reporter assay and overlapped with DE miRNAs with FDR < 0.05. ResultsOf 140 DMET-targeting miRNAs among the 2079 miRNAs in the global peripheral blood transcriptome, 41 unique DMET-targeting miRNAs were found to be DE during pregnancy - 38 differentiating pregnancy from preconception and 9 differentiating pregnancy from breastfeeding. The 56 DMETs confirmed as targets of the DE miRNAs included 8 members of the ABC (ATP-binding cassette) transporter family, all abundantly expressed in the placenta, and 4 members of the cytochrome P450 Phase 1 enzyme family with major role in xenobiotics detoxification. The study also revealed a strong (FDR<0.05), predominantly positive association between specific DMET-targeting miRNAs and sex hormone-binding globulin (SHBG) levels, suggesting a miRNA-mediated downregulation of DMETs as SHBG levels rise during pregnancy. ConclusionThis research provides crucial insights into the molecular mechanisms underlying altered drug disposition in pregnant and lactating women, paving the way for improved therapeutic management and personalized medicine in these populations.

molecular biology↗

bioRxiv2025

Abnormal Vaginal Microbiota Associated with miRNA Targeting the HIV-Host Interactome

Understanding the molecular mechanisms underlying the ability of vaginal dysbiosis to alter the mucosal barrier to HIV acquisition is an essential step toward prevention. We hypothesized that micro(mi)-RNAs dysregulated by vaginal pathobiont bacteria epigenetically control host pathways exploited by the virus. The impact of these endogenous non-coding short RNAs on the anti-viral mucosal barrier function in the female reproductive tract is largely unknown. This study utilized cervicovaginal specimens collected during the luteal and follicular phase of the menstrual cycle along with data on age, race, ethnicity, education, and body mass index from 141 healthy reproductive-age women confirmed negative for sexually transmitted infections. Vaginal microbiota was classified by Nugent scoring. Shot-gun vaginal microbiome sequencing and metagenome taxonomic classification was performed on a subset of 21 women. Levels of miRNAs in exosomes isolated from cervicovaginal secretions were quantified using the EdgeSeq-NextGen global transcriptome platform. Differential expression (DE) was determined using R. Epigenetic target prediction was performed using MirTarBase. MiRNA profiles varied by both Nugent score categories (0-3 scores = normal, 4-6 = intermediate, and 7-10 = bacterial vaginosis, BV) and by metagenome classification. Higher microbiome diversity was associated with higher number of significantly dysregulated miRNAs (588 in BV compared to Nugent 0-3 versus 42 in Nugent 4-6 compared to Nugent 0-3, false discovery rate FDR<0.01) affecting over 400 experimentally validated genes targeted for post-transcriptional regulation. The miRNAs dysregulated by G. vaginalis-dominated compared to L. crispatus-dominated metagenomes included 24 DE miRNAs (92% overlap with BV by Nugent score) and 112 validated target genes. BV-dysregulated miRNA mediated the immunosuppressive effects of BV on cytokine levels previously associated with HIV acquisition risk. The gene ontology predictions based on BV-dysregulated miRNAs identified enrichment for 445 downregulated and 50 upregulated genes previously validated as part of the HIV-host interactome. miRNAs mediation revealed a mechanism of suppressed immunity by BV predictive of HIV risk. In conclusion, miRNAs dysregulated by vaginal dysbiosis may facilitate immune imbalance and cellular pathways associated with HIV risk.

molecular biology↗

bioRxiv2025

The generation of viable, structurally integrated human-mouse chimaeras through enhanced hPSCs proliferation

The generation of human organs in animals through blastocyst complementation offers a promising solution to the shortage of transplantable organs. While human pluripotent stem cells (hPSCs) can contribute to interspecies chimaeric embryos when injected into preimplantation embryos of mice, pigs, or monkeys, their integration is often limited due to low chimaerism and segregation from host tissues, significantly impeding progress toward exogenic organ generation. Here, we demonstrate that co-overexpression of the anti-apoptotic gene BCL2 and the proto-oncogene MYCL, along with various cell cycle regulators significantly enhances human cell chimaerism by promoting cell proliferation. This strategy facilitates the generation of viable mouse pups containing hPSC-derived tissues without tumorigenesis. scRNA-seq analysis revealed that hPSCs already exit pluripotent by early post-implantation stage yet hPSCs with enhanced proliferation were able to integrate effectively into the cardiomyocytes and vasculature of both embryonic and extraembryonic tissues with gene expression profiles reflecting their structural integration. These findings highlight the critical role of cell cycle regulation in overcoming xenogeneic barriers. Our findings offer new insights into strategies for enhancing interspecies organogenesis and advancing the field of regenerative medicine. HighlightO_LIEnhancing hPSC proliferation increases human cell contribution in interspecies chimaeras C_LIO_LIAchieving the generation of viable human-mouse chimaeras without tumour formation C_LIO_LIhPSCs in post-implantation mouse epiblast exit pluripotency but retain the capacity to integrate into mouse embryogenesis C_LIO_LIhPSCs derivatives integrate into vasculature and cardiac tissues with lineage-matched transcriptional profiles C_LI

developmental biology↗

bioRxiv2025

Efficient learning and intrinsic noise filtering in recurrent spiking neural networks trained with e-prop

ObjectiveBiologically plausible learning rules for neural networks, such as e-prop (eligibility propagation), are essential both for advancing neuromorphic computing and for understanding fundamental mechanisms of learning in animal brains. However, their behavior under different network conditions remains unclear. ApproachHere, we investigate the performance of the e-prop learning algorithm in recurrent spiking neural networks (RSNNs) across different levels of recurrent connectivity and input noise using a complex temporal credit assignment task, a supervised classification problem known to be solvable by rodents. Main resultsWe show that increased sparsity in the recurrent layer significantly enhances learning performance by promoting the generation of more diverse activation patterns. Analysis of the networks evolution further reveals that the e-prop-trained input layer evolves to route distinct inputs to different regions of the recurrent layer while suppressing the contribution of noise. This partially resembles signal routing functions attributed to the thalamus in mammalian sensory systems, providing additional support for the biological plausibility of e-prop. SignificanceThese findings offer promising insights for efficiency and advantages of biologically inspired training in RSNNs.

neuroscience↗