Biology subjects

Liu, J.

Publications and source records attributed to Liu, J..

At least 199 records · Page 11Linked to original sources

bioRxiv2026

Tripeptidyl peptidase II is essential for maintaining cerebrovascular homeostasis of female mice and represents a novel therapeutic target for vascular dementia

Although cerebrovascular impairment is a known common driver of both Vascular dementia (VaD) and Alzheimers disease (AD), the underlying mechanisms remain poorly defined. Substantial evidence has demonstrated that brain-derived estrogen is essential for cerebrovascular health and neuroprotection. Consequently, therapeutic strategies that can replicate the beneficial effects of estrogen replacement therapy (ERT) in the brain while avoiding its peripheral risks are being actively pursued. Here, we demonstrate that tripeptidyl peptidase II (TPP2) is essential for cerebrovascular homeostasis in adult female mice by orchestrating intracellular Ca{superscript 2} distribution and local estrogen biosynthesis. Specifically, TPP2 deficiency triggers a vicious cycle of Ca{superscript 2} imbalance and estrogen deficiency, thereby disrupting the anticipatory unfolded protein response (UPR). This disruption consequently drives aberrant pexophagic flux and ultimately depletes ether-linked phosphatidylcholine (PC-O)--the essential building block of endothelial cell (EC) membranes. In addition, we identified PC-O as a key facilitator of choline uptake through FLVCR2 and therefore its deficiency causes subsequent choline depletion as well as significantly decreased acetylcholine (ACh). Consistent with the fact that choline is a building block of PC and that ACh is a primary driver of vascular dilation, TPP2 depletion leads to significant cerebrovascular degeneration characterized by narrowed lumens, decreased EC number, and abnormal aggregation of ECs within the blood vessel. Additionally, AAV-mediated specific expression of Far1 in ECs to promote ether-PC biosynthesis significantly increases cerebrovascular volume and diameter in hippocampi of adult female TPP2 knockout (T2KO) mice. Importantly, ectopic expression of Far1 in ECs not only significantly ameliorates memory impairment in global and conditional TPP2-depleted female mice, but also significantly improves memory performance of naturally aged female mice. In total, our findings establish TPP2 as a key determinant of cerebrovascular homeostasis in adult female mice, positioning it as a novel therapeutic candidate for the treatment of VaD and AD.

neuroscience↗

bioRxiv2026

Cholesteryl Ester as a Prognostic Biomarker In IDH-wildtype Glioblastoma

Current treatment of IDH-wildtype glioblastoma (GBM) relies on the first-line chemotherapy-temozolomide. Although MGMT methylation is routinely conducted to predict chemosensitivity, its efficacy is often compromised. Thus, there is an urgent need to discover more accurate prognostic biomarkers. Cholesteryl ester (CE) has been recently recognized as a key feature of GBM, however, its role in GBM prognosis remains poorly understood. We first employed label-free stimulated Raman scattering (SRS) imaging to quantitatively analyze CE level in intact tumor tissues obtained from IDH-wildtype GBM patients. Our result revealed significantly prolonged 2-year overall survival (OS) in patients with CE level [≥] 40% compared to those with CE level < 40%. CE outperformed MGMT methylation for 2-year OS prognosis (AUC: 0.836 vs. 0.763). Importantly, CE also achieved superior prognostic performance over MGMT methylation on an independent cohort, with higher sensitivity (0.856 vs. 0.667), specificity (0.833 vs. 0.583), NPV (1.00 vs. 0.667), PPV (0.833 vs. 0.583). Given synergistic effects between CE and MGMT methylation, we developed a prognostic model combining these two biomarkers. Specially, machine learning (XGBoost) model exhibited optimal performance in the training cohort (AUC: 0.920), and maintained its superior performance on the independent cohort (sensitivity: 0.946, specificity: 0.873, NPV: 1.00; PPV: 0.917). Mechanistically, integrative analysis of TCGA database linked poor prognosis to the coordinated upregulation of genes involved in cholesterol efflux, hydrolysis, transport, and inhibition of de novo synthesis, unraveling a possible underlying mechanism between poor prognosis and cholesterol metabolism. This work identified CE as a prognostic biomarker for IDH-wildtype GBM.

neuroscience↗

bioRxiv2026

Rewiring c-Myc Transcriptional Activity with an O-GlcNAcylation Targeting Chimera (OGTAC)

c-Myc is a transcription factor that drives tumorigenesis in many cancers. It is notoriously difficult to directly target c-Myc, mainly due to its lack of well-defined druggable pockets. O-linked {beta}-N-acetylglucosamine modification (O-GlcNAcylation) is a post-translational modification (PTM) playing an important role in regulating c-Myc functions in cancer. However, previous studies have primarily relied on global perturbations to investigate c-Myc O-GlcNAcylation, making it difficult to determine its direct functional consequences due to concurrent cellular effects. Here, we report a bifunctional O-GlcNAcylation TArgeting Chimera (OGTAC) molecule, which can induce the proximity of c-Myc and O-GlcNAc transferase (OGT) in living cells, thereby enhancing the O-GlcNAcylation of c-Myc. The c-Myc-targeting OGTAC exhibits anti-proliferation effect against cancer cells. Mapping of c-Myc occupancy on genome indicates that OGTAC rewires c-Myc transcriptional activity and reprograms expression of the downstream oncogene MALAT1, in an O-GlcNAcylation-dependent manner. Overall, OGTAC presents a novel chemically induced proximity (CIP)-based tool to target and rewire c-Myc activity in cancer. Graphic abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=135 SRC="FIGDIR/small/722559v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@d1c640org.highwire.dtl.DTLVardef@2eb70corg.highwire.dtl.DTLVardef@f38970org.highwire.dtl.DTLVardef@c421c8_HPS_FORMAT_FIGEXP M_FIG C_FIG

biochemistry↗

bioRxiv2026

A Continuum of Atrial Peristalsis Initiates the Bicuspid to Quadricuspid Valve Transition

How chamber-specific hemodynamic shear forces coordinate the spatial and temporal patterning of cardiac valves remains poorly understood. The zebrafish atrioventricular (AV) valve transitions from 2-cusp to 4-cusp at 14 days post-fertilization (dpf) through commissural remodeling, differential cell proliferation, and chondrogenic matrix secretion. Orthogonal valvular axes are present at 25 hpf (hours), preceding endocardial cushion formation. AV canal remodeling aligns the shear stress with leaflet formation at 48 hpf and 14 dpf. At 28 hpf, the ventricle develops functional syncytium, while the atrial peristalsis remains and is necessary for the Klf2a-Snai1b signaling in Notch-negative and ccm1-positive endocardial cells. Without atrial contraction, the 2-to 4-cusp transition (86.7%) fails to occur. In the absence of ventricular contraction and relaxation, the formation of the first pair of leaflets (100%) does not initiate. As a corollary, we demonstrate that KLF4/SNAI1 expression at the commissure is conserved in the mammalian AV 2- to 4-cushion transition.

developmental biology↗

bioRxiv2026

Local translation drives glioblastoma heterogeneity and tumor invasion

O_FIG O_LINKSMALLFIG WIDTH=174 HEIGHT=200 SRC="FIGDIR/small/722387v1_ufig1.gif" ALT="Figure 1"> View larger version (96K): org.highwire.dtl.DTLVardef@1da020corg.highwire.dtl.DTLVardef@1bc9b32org.highwire.dtl.DTLVardef@171198aorg.highwire.dtl.DTLVardef@43ce6f_HPS_FORMAT_FIGEXP M_FIG C_FIG Glioblastoma is characterized by diffuse brain invasion, yet the subcellular mechanisms enabling this aggressive behavior remain poorly understood. A subpopulation of glioblastoma cells forms invasive tumor microtubes (TMs), neurite-like extensions that drive whole-brain colonization. Here, we establish local protein translation as a fundamental driver of TM dynamics and invasive cell states. Developing a subcellular transcriptomics approach - integrating subcellular organelle organization with spatially resolved transcriptomics and functional readouts - we reveal that TM gene expression drives cell state identity. Invasive cells further exhibit significantly elevated local translation in protruding TMs, directly linking subcellular protein synthesis to functional invasive states associated with neurodevelopmental programs of axonal growth cones. Targeted disruption of TM-localized translation via photoswitchable puromycin, and specific knockdowns of the TM-enriched proteins GPM6A and GAP43, impaired TM dynamics, suppressed invasion, and reduced tumor growth. Together, these findings define local translation as a key determinant of tumor heterogeneity and glioblastoma invasion.

cancer biology↗

bioRxiv2026

Hippocampal Glucuronyl C5-epimerase promotes stress resilience by directly engaging PI3K through a non-enzymatic mechanism

Major depressive disorder (MDD) seriously affects human physical and mental health and causes global socioeconomic burdens. Stress-induced depressive etiology is linked to dendritic atrophy in the hippocampus, however, the underlying mechanism was poorly understood. Here, we identify that glucuronyl C5-epimerase (Glce), a heparan sulfae-modifying enzyme, as a critical regulator of hippocampal dendritic integrity and stress resilience. Genetic ablation of Glce in the hippocampal excitatory neurons is sufficient to induce dendritic atrophy and depressive-like behaviors, whereas its restoration rescues these deficits. Notably, Glce levels are reduced in the plasma of depressed individuals. We further find that Glce functions independently of its enzymatic activity, instead binding directly to and activating PI3K (p85/p110), thereby triggering an Akt/CREB/BDNF signaling cascade. Together, our findings uncover a non-canonical role for Glce and establish that Glce-PI3K-BDNF axis is essential for maintaining hippocampal structure and behavioral resilience, thereby highlighting a critical role of Glce in the pathophysiology of depression.

neuroscience↗

bioRxiv2026

Rapid neural analysis of linguistic stress and meter in continuous speech

Continuous speech evolves around vowels, the centerpieces of individual syllables. Vowels vary in linguistic and acoustic salience: Linguistically, stressed syllables are more salient than unstressed syllables: Stress patterns convey critical lexico-semantic and prosodic information, and their regularity defines the speech meter. Acoustically, English vowel intensity cues lexical stress but also marks salient syllables irrespective of stress status. Recent evidence demonstrates rapid neural analysis of vowel intensity and identity during perception of continuous speech. Here, we probe how these processes integrate lexical stress and metrical regularity. We recorded EEG while participants (n=26) listened to childrens stories with either an irregular, speech-like meter, or a regular poetic meter. Stress and meter modulated cortical encoding of vowels throughout processing: Preparatory activity preceded vowel onsets in an irregular meter only, and early sensory responses were enhanced for unstressed vowels, suggesting additional resource allocation during processing of uncertain and less discriminable speech sounds. In contrast, later processing (300-500ms) was stronger for stressed syllables and in irregular meters, suggesting a combined effect of uncertainty and informational content. Finally, responses were stronger for small intensity rises within metrically predicted stressed vowels than in all other conditions. In the time-frequency domain, the spectral profile of neural phase-locking corresponded to spectral signatures of individual evoked responses, syllable and stress rates in the stimuli. Overall, our findings reveal rapid neural integration of stress and metrical expectations in neural processing of continuous speech. These dynamics may underlie the perceptual benefits of metrically regular speech, such as poetry and song lyrics.

neuroscience↗

bioRxiv2026

Uncertainty-aware localization microscopy by variational diffusion

Fast extraction of physically relevant information from images using deep neural networks has led to significant advances in fluorescence microscopy and its application to the study of biological systems. For example, the application of deep networks for kernel density (KD) estimation in single-molecule localization microscopy (SMLM) has accelerated super-resolution imaging of densely labeled structures in the cell. However, localization of fluorescent molecules in dense images is a difficult inverse problem with potentially multiple solutions. To model a probability distribution of solutions to this problem, we propose a generative modeling framework for KD estimation in SMLM based on a conditional variational diffusion model (CVDM). In this framework, CVDM is trained to perform localization tasks on low-resolution measurements by modeling a distribution of high-resolution KD estimates. This approach allows us to probe the structure of the distribution on KD estimates and express uncertainty, which is not currently offered by existing deep models for localization microscopy. We demonstrate that this model permits high-fidelity super-resolution, enables the uncertainty estimation of regressed KD estimates, and has important implications for image restoration in single-molecule and super resolution microscopy.

bioinformatics↗

bioRxiv2026

Counting fluorescent emitters with a single photon avalanche diode array

For decades, the photon counting histogram (PCH) was used as the sole method to quantify fluorophore numbers in a diffraction-limited focal volume. This technique combines spatial excitation profiles, and the distribution of photon counts to register the photon emission statistics of individual fluorophores. However, this approach has not yet been transferred to widefield fluorescent imaging due to the lack of fast and single photon sensitive camera sensors which can capture the photon emission statistics of a single fluorophore. Here, we explore avenues towards quantitative analysis of the active fluorophore number by leveraging recent advancements in single photon avalanche diode (SPAD) array technology. Binary exposures of a SPAD array can be synchronized with picosecond laser pulses to measure the PCH in a widefield setting. Then, by modeling the statistical relationship between the active fluorophore number and the PCH in a region of interest following a laser pulse, we can perform Bayesian inference of this number. The model is demonstrated experimentally by counting quantum dots and various numbers of fluorescent dye molecules bound to DNA origamis. We find that this method has several important applications in widefield microscopy, including enhanced localization microscopy and constrained fitting of multiple unresolvable fluorescent emitters.

biophysics↗

bioRxiv2026

LNGCN: A Distance-Aware Dynamics Network for Protein-Protein Interaction Prediction

High-throughput PPI screening requires not only interaction discrimination but also stable, calibrated scores for candidate prioritization and efficient experimental allocation. Conventional graph neural networks may over-smooth deep residue representations, reducing score separation among high-confidence candidates. We therefore developed LNGCN, a distance-aware continuous-time graph framework for experimental PPI prioritization. LNGCN integrates residue-level structural graphs with liquid neural dynamics, uses radial distance to drive continuous graph evolution, and applies hierarchical calibration to convert raw outputs into prioritization scores. It performed robustly on balanced human, 1:10 imbalanced, and cross-species yeast bench-marks. On the imbalanced dataset, the top 5% of candidates achieved 85.1% precision. LNGCN also preserved residue-level representation diversity better than conventional GCN variants. In FGF23-FGFR1c--Klotho, SHP2 signaling, Tdk1 oligomer-dependent binding, and a TPR study, it enriched known partners near the ranking top, and experimentally supported novel interactions. Overall, LNGCN enables sustained enrichment of experimentally relevant PPI candidates.

bioinformatics↗

bioRxiv2026

The discovery of missing taxane C13α-O-deacetylases re-delineates the biosynthetic pathway of paclitaxel

The prevalence of naturally occurring C13-acetoxy taxanes, together with the presence of a native C13-acetyltransferase in yew trees, suggests that the natural biosynthetic pathway for paclitaxel may involve a cryptic C13-O-deacetylation step. However, whether a putative taxane C13-O-deacetylase (T13dA) acts in the pathway of paclitaxel biosynthesis remains elusive. Herein, we functionally characterized two novel taxane C13-O-deacetylases (T13dA1 and T13dA2) from Taxus x media cell cultures, providing experimental evidence for the molecular and biochemical plausibility of C13-O-deacetylation in paclitaxel biosynthesis in Taxus species. Furthermore, T7dA1, a novel taxane C7{beta}-O-deacetylase with higher activity than the reported T7dA was discovered and characterized here. Notably, we reconstituted a new 19-gene pathway enabled by the integration of a C13-O-acetylation-deacetylation module for the de novo biosynthesis of baccatin III in Nicotiana benthamiana leaves, reaching a yield of 71{square} g{square} g-1 DW, much higher than that obtained from the 17{square} gene pathway lacking this module. These findings not only demonstrate the functional importance of the C13-O-acetylation-deacetylation module as a branch within the paclitaxel biosynthetic network, but also expand the multi-branched taxane metabolic network in Taxus and provide novel enzymatic tools and engineering strategies to facilitate heterologous pathway reconstitution and efficient paclitaxel production.

biochemistry↗

bioRxiv2026

Defined Mars Media (DMM), a chemically defined simulant of the soluble macro- and micro- nutrients in Mars regolith for use in biological research

Mars harsh yet workable surface conditions, such as manageable temperatures, availability of solar energy, and in situ resources like water ice, carbon dioxide, and mineral-rich regolith, make it a compelling target for supporting life beyond Earth. However, existing experiments testing chemical habitability in Mars conditions generally rely on leachates of physical regolith simulants, which vary in composition across simulant types, leaching conditions, and production batches. We introduce a defined Mars media (DMM) that accurately simulates the biologically relevant nutrients (nitrogen, phosphorus, and sulfur) and stressors (perchlorates, heavy metals) in Martian regolith when it is leached in water at neutral pH. We formulated DMM by combining direct rover and lander measurements from Mars with laboratory measurements of regolith simulant leachates. We validate DMM from a 1x to 20x concentrate, equivalent to 40 g/L to 800 g/L of leached regolith. Using DMM with acetate as a Mars atmosphere-derived carbon source, we grew eight bacteria, demonstrating that organisms can source all essential nutrients from Martian resources. We also demonstrate that microbial growth in DMM is robust to uncertainties in Martian regolith composition: sensitivity experiments can identify limiting trace element nutrients and toxins in DMM, and show that bacterial growth is maintained across at least an order of magnitude variation in their concentrations. This is the first defined Mars regolith media recipe containing both macro- and micro-nutrients, and designed specifically for biological experimentation. By shifting from variable leachate-based approaches to a defined aqueous analog, we enable controlled hypothesis testing of microbial survival, growth, and function in a Martian chemical environment. DMM will enable further research on astrobiology, biological in situ resource utilization, large-scale soil remediation, and terraforming. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=121 SRC="FIGDIR/small/719001v2_ufig1.gif" ALT="Figure 1"> View larger version (27K): org.highwire.dtl.DTLVardef@37408forg.highwire.dtl.DTLVardef@16fabfforg.highwire.dtl.DTLVardef@84fa43org.highwire.dtl.DTLVardef@4a7c5e_HPS_FORMAT_FIGEXP M_FIG C_FIG

microbiology↗

bioRxiv2026

Improving AlphaFold3 by Engineering MSA and Template Inputs

AlphaFold3 introduces a unified framework for predicting the structures and interaction of several biomolecules including single-chain protein monomers, multi-chain protein multimers, and protein-ligand complexes. While it achieves the state-of-the-art performance in most predictions, its prediction accuracy depends on the quality of multiple sequence alignment (MSA) and structural template inputs. There are few works of using customized MSAs and templates to improve AlphaFold3. In this work, we systematically investigate how diverse and carefully engineered MSAs and templates can be leveraged to improve AlphaFold3 predictions. We evaluate our methods on protein monomers, multimers, and protein-ligand complexes, and observe consistent, sizable gains in structure prediction accuracy for monomers (TM-score 0.937 vs 0.882), for multimers (DockQ score 0.550 vs 0.525), and for protein-ligand complexes (ligand RMSD 3.258 [A] vs 4 [A]) compared to the default AlphaFold3. Moreover, for the first time, we demonstrate that AlphaFold3 performs significantly better than AlphaFold2 when both use the same customized MSA and template inputs. The results highlight the importance and effectiveness of using diverse MSAs and templates to improve AlphaFold3.

bioinformatics↗

bioRxiv2026

CHORD: a framework for cross-species single-cell integration across gene, cell and cell-type levels

Quantifying cross-species relationships among cell types from single-cell transcriptomic data can reveal both conserved and divergent patterns of cell-type hierarchies. However, existing cross-species integration methods can be limited in modeling genes beyond orthologs by leveraging cell-type-resolved transcriptional context, or in learning explicit type-level representations. Here we present CHORD, a cross-species integration framework that jointly learns representations of genes, cells and cell types. We demonstrate that CHORD can integrate cross-species single-cell atlases and support cell-type annotation with unknown cell-type detection. In the frog-zebrafish embryogenesis and mammalian motor cortex atlases, CHORD infers cell-type trees that place conserved cell types from different species in relative proximity and summarize hierarchical relationships among cell types. CHORD also supports cross-species comparison of continuous phenotypic variation by placing embryonic cells along an aligned developmental timeline. CHORD further yields gene embeddings that capture orthologous and functional relationships, and gene importance scores linking genes to cell types.

bioinformatics↗

bioRxiv2026

Application of a High-Biomimetic Tumor Organoid-CAF Co-Culture Model for the Efficacy Evaluation of CAR-T Drugs

The tumor microenvironment (TME) is a complex ecosystem composed of tumor cells, cancer-associated fibroblasts (CAFs), immune suppressive cells, and the extracellular matrix (ECM), playing a crucial role in tumor development and CAR-T cell therapy efficacy. CAR-T therapy has shown promise in hematological malignancies but faces challenges in solid tumors due to the TMEs ability to suppress CAR-T cell infiltration, proliferation, and cytotoxicity. Traditional drug evaluation models, such as 2D cell cultures and animal models, have significant limitations due to oversimplification of the in vivo environment or physiological differences between species. Organoid models offer a more biomimetic approach but often fail to fully recapitulate the TMEs complexity and heterogeneity. Our research developed a tumor organoid and CAF co-culture model using the IBAC co-culture chip, demonstrating that CAFs significantly impact CAR-T cell therapy efficacy by forming physical (e.g., fibronectin) and chemical (e.g., IL-10) barriers that prevent CAR-T cell infiltration and cytotoxicity. This model provides a high-biomimetic platform for investigating the TMEs effects on CAR-T therapy and highlights the importance of incorporating a comprehensive stromal component into in vitro models to enhance their predictive power for cancer treatment.

cell biology↗

bioRxiv2026

A global metagenomic atlas uncovers ubiquitous biosynthetic potential linked to adaptation in extreme environments

Extreme environments impose severe physicochemical stresses that drive microorganisms to evolve specialized survival strategies. Microbial secondary metabolites determined by biosynthetic gene clusters (BGCs) are recognized as important mediators of microbial adaptation to environmental stress. However, their ecological roles, particularly habitat-dependent preferences across different environments, remain poorly understood. Although extreme environments provide opportunities to mine microbiomes for unique adaptations, such research is hampered by a lack of systematic overview of its genomic diversity, BGC diversity, and the relationships between them. Here, we constructed a standardized extremophilic genomic catalogue (SEGC) from 1,462 metagenomic samples spanning seven representative extreme habitats. The catalogue comprised 54,661 metagenome-assembled genomes representing 21,805 species, 66.1% of which were previously uncharacterized. With this catalogue, we identified 162,855 BGCs distributed across 81.5% of MAGs. Gene cluster family analysis showed the strong habitat dependence largely explained by species-level habitat specificity. Terpene biosynthetic pathways illustrated habitat-linked adaptive strategies, with hopan-22-ol biosynthesis enriched in acid mine, deep sea and hydrothermal plume environments, while retinal-based phototrophy predominated in cryosphere and saline-alkaline habitats. Metatranscriptomic analyses supported in situ activity of these pathways. In conclusion, we presented a global atlas of biosynthetic potential across extreme-environment microbiota and revealed habitat-dependent patterns of secondary metabolism linked to microbial survival.

microbiology↗

bioRxiv2026

Pyruvate-driven hydrogen production promotes polyphenol bioconversion by gut bacteria

Gut microbial biotransformation of poorly absorbable polyphenols into bioactive, bioavailable metabolites is increasingly recognized as a key mechanism underlying their health benefits of polyphenols. Microbial ellagic acid (EA)-to-urolithin conversion represents a typical example, but the environmental factors that facilitate such metabolism remain underexplored. We discovered that urolithin production by a gut commensal bacterium, Gordonibacter urolithinfaciens (G. uro), is metabolically repressed by arginine. To overcome such limitations, we developed PhenolBoost Medium (PBM) that induces a metabolic shift by suppressing the arginine deiminase pathway while activating pyruvate metabolism and hydrogen production in G. uro, thereby driving urolithin dehydroxylation. Transcriptomic profiling and 13C-isotopic tracing analysis revealed that pyruvate metabolism in PBM upregulates hydrogenase expression, facilitating the dehydroxylation of EA. PBM also promoted the complete conversion of EA to urolithin A in G. uro-Enterocloster bolteae co-culture, and other polyphenol biotransformations. In addition, co-culturing G. uro with hydrogen-producing Bacteroides species significantly increased urolithin production. Furthermore, an arginine-limited, pyruvate-enriched dietary regimen proved effective in vivo, resulting in significantly higher urolithin production and bioavailability in a mouse model. Our findings reveal the critical role of hydrogen in facilitating polyphenol dehydroxylation, and offer a viable nutritional strategy for boosting microbial production of beneficial metabolites from polyphenols.

microbiology↗

bioRxiv2026

16S rRNA k-mer composition encodes microbial functional potential

16S rRNA amplicon sequencing is widely used for microbiome profiling, but most methods rely on reference databases of characterized organisms, limiting its accuracy in function prediction for underrepresented environments. We discovered that 16S rRNA k-mer composition carries substantial functional signal: (i) whole-genome k-mer profiles predict genome-encoded functions, and (ii) 16S rRNA k-mer profiles reflect their source genomes composition. Building on these relationships, we developed embeRNA, a neural network framework that predicts functions directly from 16S rRNA k-mer embeddings without requiring taxonomy assignment or phylogenetic placement. embeRNA outputs per-function probability scores, enabling users to tune decision thresholds to balance precision and recall or account for community novelty. In a stringent "novel microbes" benchmark - where all test sequences shared <97% identity with training data - embeRNA outperformed reference-based methods, particularly for hard-to-label functions. Applied to soil metagenomes with paired 16S and whole metagenome shotgun sequencing (WMS) data, embeRNA recovered most WMS-inferred functions and produced abundance profiles strongly correlated with WMS results, attaining better performance than a reference-based approach. Our findings demonstrate that 16S rRNA directly captures functional potential, and 16S amplicon sequencing data can complement WMS-based inference to broaden functional characterization of microbiomes, especially in understudied environments.

bioinformatics↗