Proceedings of the National Academy of Sciences, Volume 123, Issue 19, May 2026.
Science Journals
Proceedings of the National Academy of Sciences, Volume 123, Issue 19, May 2026.
Proceedings of the National Academy of Sciences, Volume 123, Issue 19, May 2026.
The accuracy of spatial gene expression profiles generated by probe-based in situ spatially resolved transcriptomic technologies depends on the specificity with which probes bind to their intended target gene. Off-target binding, defined as a probe binding to something other than the target gene, can distort a gene’s true expression profile, making probe specificity essential for reliable transcriptomics. Here, we investigated off-target binding affecting the 10x Genomics Xenium technology. We developed a software tool, Off-target Probe Tracker (OPT), to identify putative off-target binding via alignment of probe target sequences and assessing whether mapped loci corresponded to the intended target gene across multiple reference annotations. Applying OPT to a Xenium human breast gene panel, we identified at least 14 out of the 313 genes in the panel potentially impacted by off-target binding to protein-coding genes. To substantiate our predictions, we leveraged a Xenium breast cancer dataset generated using this gene panel and compared results to orthogonal spatial and single-cell transcriptomic profiles from Visium CytAssist and 3′ single-cell RNA-seq derived from the same tumor block. Our findings indicate that for some genes, the expression patterns detected by Xenium demonstrably reflect the aggregate expression of the target and predicted off-target genes based on Visium and single-cell RNA-seq, rather than the target gene alone. We further applied OPT to identify potential off-target binding in custom gene panels and integrate tissue-specific RNA-seq data to assess effects. Overall, this work enhances the biological interpretability of spatial transcriptomics data and improves reproducibility in spatial transcriptomics research.
Circulating cell-free DNA (cfDNA) is valuable for molecular testing, but typically requires specialized collection tubes or immediate processing. We investigated whether residual plasma from heparin separators, routinely used in clinical chemistry, could serve as an accessible and underused source for cfDNA. We analyzed matched plasma samples from healthy volunteers in two experiments: an immediate-processing comparison across EDTA, Streck, and heparin separator tubes (n=5), and a clinical-handling simulation comparing EDTA and heparin separator tubes under delayed processing at room temperature or 4°C (n=6). We also analyzed matched plasma samples from viral PCR-positive patients in a hospital cohort (n=38). Whole-genome sequencing and enriched methylation sequencing were performed to assess concordance across metagenomics, copy number, methylation, and fragmentomic features. Under immediate processing, heparin separator plasma showed high concordance with EDTA and Streck plasma for methylation patterns (Spearman’s ρ=0.65–0.70) and fragmentation features. In the Hospital Cohort, heparin separator plasma showed strong concordance with matched EDTA plasma for viral detection (Spearman’s ρ=0.95), copy number alteration profiling (Spearman’s ρ=0.72–0.96), and methylation patterns (Spearman’s ρ=0.50–0.83). These findings support the feasibility of using refrigerated, promptly processed residual plasma from routine clinical chemistry as a supplementary source for cfDNA biobanking and molecular analyses.
<i>Fusobacterium nucleatum</i>, a gram-negative bacterium implicated in periodontal disease, contributes to tumor progression in various cancers. Whether the presence of <i>F. nucleatum</i> inhibits tumor progression of some cancers is largely unknown. Here, we identify an interaction between <i>F. nucleatum</i> and the natural killer (NK) cell receptor NKp46. Analysis of TCGA datasets revealed that the co-occurrence of <i>F. nucleatum</i> and high NKp46 expression correlates with improved survival in head and neck cancers but not in colorectal cancers. Using binding assays, we demonstrate that both human NKp46 and its murine ortholog, Ncr1, directly recognize the fusobacterial adhesin RadD. Genetic deletion of <i>radD</i> or blockade of NKp46 significantly impaired NK cell-mediated cytotoxicity in vitro and promoted tumor-cell growth. In vivo, infection with <i>F. nucleatum</i> accelerated tumor progression, with an exacerbated effect observed in the absence of RadD or NKp46. These findings highlight RadD as a critical ligand for NKp46 and establish the NKp46–RadD axis as a key interface in host–microbe–tumor interactions, offering a novel target for immunotherapeutic intervention in cancer influenced by microbial factors.
<p>by Yacine Ben Chehida, Eva S. M. van der Heijden, Edward Page, Patricio A. Salazar C, Neil Rosser, Kimberly Gabriela Gavilanes Córdova, Mónica Sánchez-Prado, María José Sánchez-Carvajal, Franz Chandi, Alex P. Arias-Cruz, Maya Radford, Gerardo Lamas, Chris D. Jiggins, James Mallet, Melanie McClure, Camilo Salazar, Marianne Elias, Caroline N. Bacquet, Nicola J. Nadeau, Kanchon K. Dasmahapatra, Joana I. Meier</p>
Convergent evolution, the repeated evolution of similar phenotypes, is widespread in nature, but there are few studies investigating the genetic mechanisms of convergence across wide evolutionary timescales. The extent to which the same genetic mechanisms contribute to convergent evolution could reveal whether the pathway towards these fitness optima is flexible or constrained to follow a particular route, informing us about the predictability of evolution. Wing color pattern mimicry in Lepidoptera is a well-known example of convergent evolution, but as studies are restricted to a few closely related species, it is difficult to make general inferences about the predictability of evolution in this system. Here we study convergent evolution in multiple mimetic neotropical lepidopteran lineages that diverged between ~1 and 120 Mya, including seven species of Ithomiini and <i>Heliconius</i> butterflies and a day-flying <i>Chetone</i> moth. Across butterfly lineages that diverged up to ~30 Mya, the genetic variants most strongly associated with convergent color pattern switches are located in similar noncoding regions near the genes <i>ivory</i> and <i>optix</i>. In the more distantly related moth species, color pattern variation is associated with a ~1 Mb inversion which also contains <i>ivory</i>, closely mirroring the supergene architecture of the co-mimetic butterfly <i>Heliconius numata</i>. In contrast to previous studies on <i>Heliconius</i> butterflies, we find limited evidence that convergence among closely related Ithomiini species results from all…
<p>by Julia Penndorf, Brendan J. Barrett, Sonja Wild, John M. Martin, Lucy M. Aplin</p>
There is extensive evidence that the spread of innovation via social learning can facilitate uptake of new foraging behaviours in populations. In comparison, social learning about novel food types has received comparatively little attention. Yet the adoption of novel food is vital to persistence in, or colonisation of, novel environments. Here, we present a novel food (almonds in the shell, coloured either blue or red) in a two-option and control cultural diffusion experiment to five neighbouring roosts of 705 individually-marked sulphur-crested cockatoos (<i>Cacatua galerita</i>) living in a highly urbanised environment. From 4 initially trained individuals, a total of 349 individuals across all roosts learned to feed on the novel food within 10 days of first exposure. Using network-based diffusion-analysis (<i>N</i> = 214 learners out of 322 individuals with available social information), we demonstrated that this spread occurred almost exclusively through social learning, with information spreading through social network ties. Second, using experience-weighted attraction models, we described age-differences in social learning strategies, with juveniles, but not adults, exhibiting a conformist bias to prefer the most frequently chosen food colour. Third, when analysing 539 opening techniques of the novel food by 147 individuals across the five roosts, we found that opening techniques were more similar between roost communities when the distance between sites was small, or the degree of movement between sites was high. In addition, when focusing on a subset for which social association data were available (273 openings by 78 individuals), techniques tended to be more similar between close associates. Taken together, our study suggests that the adoption of novel food in urban-living sulphur-crested cockatoos is facilitated by social transmission of knowledge through networks,…
mRNA vaccines emerged as a leading vaccine technology during the COVID-19 pandemic. However, their sustained protective efficacies were limited by relatively short-lived antibody responses and the emergence of SARS-CoV-2 variants, necessitating frequent and variant-updated boosters. We recently developed the ESCRT- and ALIX-binding region (EABR) mRNA vaccine platform, which encodes engineered immunogens that induce budding of enveloped virus-like particles (eVLPs) from the plasma membrane, thereby resulting in presentation of immunogens on cell surfaces and eVLPs. Prior studies showed that spike (S)-EABR mRNA-LNP immunizations elicited enhanced neutralizing antibody responses against ancestral and variant SARS-CoV-2 compared with conventional S mRNA-LNP in naïve mice, but the effectiveness of S-EABR mRNA-LNP boosters in the context of pre-existing immunity has not been investigated. Here, we evaluated monovalent Wuhan-Hu-1 (Wu1) and bivalent (Wu1/BA.5) S-EABR mRNA-LNP boosters in mice pre-vaccinated with conventional Wu1 S mRNA-LNP. Compared to conventional S mRNA-LNP boosters, the EABR approach enhanced monovalent and bivalent mRNA-LNP booster-induced neutralizing responses against Omicron subvariants BA.1, BA.5, BQ.1.1, and XBB.1, with bivalent S-EABR mRNA-LNP consistently eliciting the highest titers. Epitope mapping of polyclonal antisera by deep mutational scanning revealed that bivalent S-EABR mRNA-LNP boosted diverse ‘polyclass’ anti-receptor-binding domain (RBD) responses, suggesting balanced targeting of multiple RBD epitope classes. In contrast, monovalent S, bivalent S, and monovalent S-EABR mRNA-LNP boosters elicited less diverse polyclonal serum responses primarily targeting immunodominant RBD epitopes. Cryo-electron microscopy (cryo-EM) structures demonstrated that bivalent mRNA immunizations promote S heterotrimer formation, potentially enhancing bivalent S-EABR mRNA-LNP booster-induced antibody breadth and polyclass epitope targeting by activating…
Animal models of epilepsy are critical in drug development and therapeutic testing. However, dominant methods for evaluating epilepsy treatments face a tradeoff between higher throughput and etiological relevance. Screening models are either based on acutely induced seizures in wild-type, naive animals or spontaneous seizures in chronically epileptic animals. Each has its disadvantages – acute convulsant or kindling-induced seizures do not account for the myriad neuropathological changes in the diseased, epileptic brains, and spontaneous behavioral seizures are sparse in chronically epileptic models, making it time-intensive to sufficiently power experiments. In this study, we developed the Opto-IHK (optogenetically induced seizures in intrahippocampal kainate mice) model, a mechanistic approach to precipitate seizures ‘on demand’ in chronically epileptic mice. We briefly synchronized principal cells in the CA1 region of the diseased hippocampus to reliably induce stereotyped on-demand behavioral seizures. These induced seizures resembled naturally occurring spontaneous seizures in the epileptic animals and could be stopped by commonly prescribed anti-seizure medications such as levetiracetam and diazepam. Furthermore, we showed that seizures induced in chronically epileptic animals differed from those in naive animals, highlighting the importance of evaluating therapeutics in the diseased circuit. Taken together, we envision the Opto-IHK model to accelerate the evaluation of both pharmacological and closed-loop interventions for epilepsy.
A new fluorescent sensor makes it possible to track the neurotransmitter GABA in freely moving animals.
A subset of nuclear receptors (NRs) function as permissive heterodimers with retinoid X receptor (RXR), defined by transcriptional activation in response to RXR agonist ligands. Permissive NR-RXR activation is generally understood to operate through a classical pharmacological mechanism in which RXR agonist binding enhances coactivator recruitment to the heterodimer. However, we previously demonstrated that transcriptional activation of permissive Nurr1-RXRα (NR4A2-NR2B1) heterodimers by an RXR ligand set, which included pharmacological RXR agonists and selective Nurr1-RXRα agonists that function as antagonists of RXRα homodimers, is explained by a non-classical activation mechanism involving ligand-binding domain (LBD) heterodimer dissociation (Yu et al., 2023). Here, we extend mechanistic ligand profiling of the same RXR ligand set to the evolutionarily related Nur77-RXRγ (NR4A1-NR2B3) heterodimer. Biochemical and NMR protein-protein interaction profiling, together with cellular transcription studies, indicate that activation of Nur77-RXRγ transcription by the RXR ligand set, which lacks selective Nur77-RXRγ agonists, is consistent with contributions from both classical pharmacological activation and LBD heterodimer dissociation. However, reanalysis of our previously published data for Nurr1-RXRα revealed that inclusion of selective Nurr1-RXRα agonists was essential for elucidating the LBD heterodimer dissociation mechanism. Together, our findings highlight the importance of using a more functionally diverse RXR ligand set to define the mechanism of Nur77-RXRγ activation and to further evaluate whether LBD heterodimer dissociation represents a shared activation mechanism among NR4A-RXR heterodimers relevant to neurodegenerative and inflammatory diseases.
A growing number of therapies are being developed to target the cell cycle machinery for the treatment of cancer and other human diseases. Consequently, a greater understanding of the factors regulating cell cycle progression becomes essential to help enhance the response to these new therapies. Here, using data from the Cancer Dependency Map, we identified FAM53C as a new regulator of cell cycle progression. We found that FAM53C is critical for this cell cycle transition and that it acts upstream of the Cyclin D-CDK4/6-RB axis and of p53 in the regulation of the G1/S transition. By mass spectrometry, biochemical, and cellular assays, we identified and validated DYRK1A as a cell cycle kinase that is inhibited by and directly interacts with FAM53C. Consistent with the role for FAM53C identified in cells in culture, <i>FAM53C</i> knockout human cortical organoids display increased cell cycle arrest and growth defects. <i>Fam53C</i> knockout mice show minor behavioral phenotypes. Because DYRK1A dysregulation contributes to developmental disorders such as Down syndrome as well as tumorigenesis, future strategies aiming at regulating FAM53C activity may benefit a broad range of patients.
Convergent extension (CE) is a fundamental morphogenetic process where oriented cell behaviors lead to polarized extension of diverse tissues. In vertebrates, regulation of CE requires both non-canonical Wnt, its co-receptor Ror, and several ‘core members’ of the planar cell polarity (PCP) pathway. PCP was originally identified as a mechanism to coordinate the cellular polarity in the plane of static epithelium, where core proteins Frizzled (Fz)/Dishevelled (Dvl) and Van Gogh-like (Vangl)/Prickle (Pk) partition to opposing cell cortex. But how core PCP proteins interact with each other to mediate non-canonical Wnt/Ror signaling during CE is not clear. We found previously that during CE, Vangl cell-autonomously recruits Dvl to the plasma membrane and keeps Dvl inactive. In this study, we show that non-canonical Wnt induces Dvl to transition from Vangl to Fz in <i>Xenopus</i> embryos. Pk inhibits the transition and functionally synergizes with Vangl to suppress Dvl during CE. Conversely, Ror is required for the transition and functionally antagonizes Vangl. Biochemically, Vangl interacts directly with both Ror and Dvl. Ror and Dvl do not bind directly but can be co-fractionated with Vangl. Collectively, we propose that Pk assists Vangl to function as an unconventional adaptor that brings Dvl and Ror into a complex to serve two functions: (1) simultaneously preventing both Dvl and Ror from ectopically activating non-canonical Wnt signaling; and (2) relaying Dvl to Fz for signaling activation upon non-canonical Wnt-induced dimerization of Fz and Ror.
The exposure of frozen biological samples to the high-energy electron beam in a cryo-electron microscope commonly leads to beam-induced sample motion and distortions. Previously, we described <i>Unblur</i>, software to correct for beam-induced motion based on the alignment of full frames in a movie collected during the beam exposure (Grant and Grigorieff, 2015). Here, we present <i>Unbend</i>, extending <i>Unblur</i> by accommodating more localized sample bending and distortions using a 3D cubic B-spline model. <i>Unbend</i> is integrated into our <i>cis</i>TEM software with a new local motion visualization panel. We processed movie frames from various in situ sample types, including whole cells, lamellae, and cell lysates, to analyze motion behavior across different specimen types. To quantify the improvement in high-resolution signal, we utilized the 2D template matching method to search large ribosomal subunits from the motion-corrected micrographs. Overall, the signal-to-noise ratio of detected particles improved by 3–8% across different samples compared with full-frame aligned micrographs, while the number of detected target particles increased by up to ~300%. Furthermore, we processed micrograph montages to study motion patterns across an entire sample, revealing considerable variance in distortion scale within the same sample, suggesting a complex underlying mechanism.
The world needs WHO. This should be an unambiguous and uncontroversial statement. But it is not. There are signals that WHO's perceived value has eroded, and that its central position in the international system for health is under threat. Some critiques of the organisation have been politically natured, but that should not inhibit debate over legitimate concerns: WHO's lack of agility, insufficient transparency in key processes and decision making, and the absence of a clear narrative of the indispensable benefit the agency provides to the world.
New England Journal of Medicine, Ahead of Print.
New England Journal of Medicine, Ahead of Print.
<p>by Aleksandra Otwinowska, Janusz Koszucki, Vyshakh R. Panicker, Jade Leconte, Sebastian Olejniczak, Kathryn E. Holt, Edward J. Feil, Eduardo P. C. Rocha, Bogna Smug, Barbara Maciejewska, Zuzanna Drulis-Kawa, Rafal J. Mostowy</p>
Virulent bacteriophages infecting <i>Klebsiella pneumoniae often show capsule-driven host tropism</i> due to the presence of capsule-specific depolymerases. Yet for temperate phages the genetic and functional basis of such capsular specificity remains less well understood. Depolymerases appear unexpectedly rare in prophage genomes, raising unresolved questions about which prophage genes mediate capsular specificity, whether this apparent scarcity reflects biological or ecological differences versus annotation limitation, and whether prophage-encoded receptor-binding proteins (RBPs) are functionally active. To address these questions, we analysed 3,900 <i>Klebsiella</i> genomes from diverse ecological niches to identify prophage-encoded proteins mediating capsular specificity. We conducted a genome-wide association study (GWAS) correlating prophage protein clusters (from 8,105 prophages) with confidently assigned bacterial K-loci. GWAS revealed statistically supported predictors of capsular specificity for 16 of the 35 most diverse K-loci analysed. These predictors were dominated by diverse RBPs, including classical β-helix depolymerases (6 predictors), SGNH-domain hydrolases predicted to deacetylate polysaccharides (6 predictors), and structurally novel RBPs lacking known depolymerase folds (2 predictors). Nearly one-third of K-loci yielded no statistically significant predictors. A targeted experimental screen of 50 candidate prophage depolymerases showed that 34 failed to yield detectable recombinant expression, and neither sequence similarity, structural prediction, nor prophage genomic context reliably predicted activity. Of the 14 active enzymes, 5 targeted a K-type different from that predicted of their bacterial host, and enzym…
<p>by Marc Talló-Parra, Mireia Puig-Torrents, Gemma Pérez-Vilaró, Sol Ribó Pons, Juana Díez</p>
Arboviruses induce acute lytic infection in human cells but establish persistent infection in their mosquito vectors, a viral strategy that is essential for sustained viral transmission. How mosquito cells maintain continuous production of viral progeny without compromising host cell viability remains a fundamental unresolved question. Because arbovirus replication in human cells relies on viral takeover of the host translational machinery, we investigated how translation is regulated during persistent infection in mosquito cells using chikungunya virus (CHIKV) as a model. A temporal analysis of viral RNA translation in RNAi-competent and RNAi-deficient <i>Aedes albopictus</i> cells revealed that persistence was associated with reduced viral protein production resulting from translation repression of viral RNAs. Subcellular localization analyses of the viral protein nsP2 and LC-MS/MS analyses of host tRNAs showed that, in contrast to human cells, CHIKV infection in mosquito cells neither induced nuclear relocalization of viral nsP2 to induce global host mRNA depletion, nor reshaped the tRNA modification landscape to compensate for the suboptimal codon usage of viral RNAs. Together, our results indicate that persistent infection in mosquito cells is characterized by a balanced host–virus translational state, in which limited viral translation is maintained while viral takeover of the host translational machinery is avoided. Notably, translation repression of viral RNAs was also observed during Zika virus (ZIKV) infection, suggesting that this mechanism may represent a general RNAi-independent feature of arbovirus persistence in mosquito cells.
B cells and the antibodies they produce are vital to health and survival, motivating research on the details of the mutational and evolutionary processes in the germinal centers (GCs) from which mature B cells arise. It is known that B cells with higher affinity for their cognate antigen (Ag) will, on average, tend to have more offspring. However, the exact form of this relationship between affinity and fecundity, which we call the ‘affinity–fitness response function’, is not known. Here we use deep learning and simulation-based inference to learn this function from a unique experiment that replays a particular combination of GC conditions many times in mice. All code is freely available at <a href="https://github.com/matsengrp/gcdyn">https://github.com/matsengrp/gcdyn</a>, while datasets and inference results can be found at <a href="https://doi.org/10.5281/zenodo.15022130">https://doi.org/10.5281/zenodo.15022130</a>.
The human voltage-gated proton channel (hH<sub>v</sub>1) is a dimer of voltage-sensor domains (VSDs) containing highly selective proton permeation pathways in each monomer. In addition to voltage, hH<sub>v</sub>1 is regulated by other stimuli, including pH gradients, mechanical forces, and ligands, such as Zn<sup>2+</sup>. Aside from the VSDs, this membrane protein contains an N-terminal domain and a C-terminal coiled-coil domain (CC) formed between the monomers. To address the need for direct measurements of conformational rearrangements in hH<sub>v</sub>1, we developed a Förster resonance energy transfer (FRET) approach to measuring the conformational rearrangements in full-length hH<sub>v</sub>1 purified from <i>E. coli</i>. We used genetic code expansion (GCE) to generate a library of 14 full-length hH<sub>v</sub>1 constructs, each incorporating the fluorescent noncanonical amino acid acridon-2-ylalanine (Acd) at a different site throughout the various structural domains. Following the expression and purification of these hH<sub>v</sub>1-Acd proteins, we found that 12 sites yielded stable and functional proton-permeable channels. The fluorescence properties of Acd at each site showed small site-specific differences. Furthermore, we measured site-specific FRET efficiencies from tryptophan (Trp) and tyrosine (Tyr) to Acd in the hH<sub>v</sub>1-Acd proteins and found results consistent with correct folding in detergent micelles. Finally, the addition of Zn<sup>2+</sup> produced reversible changes in FRET, with affected residues clustered on the intracellular side of the channel.
<p>by Jian Guan, Rebecca Lee Hurto, Akash Rai, Janakraj Bhattrai, Christopher A. Azaldegui, Luis A. Ortiz-Rodríguez, Quancheng Liu, Julie S. Biteen, Lydia Freddolino, Ursula Jakob</p>
Uncovering what drives select biomolecules to form phase-separated condensates in vivo and identifying their physiological significance are topics of fundamental importance. Here, we show that nitrogen-starved <i>Escherichia coli</i> produces long-chain polyphosphates, which scaffold the RNA chaperone Hfq into high molecular weight complexes, which eventually phase separate together with components of the RNA translation and processing machinery. The presence of polyphosphate within these condensates controls Hfq function by selectively stabilizing polyadenylated RNAs involved in transcription and protein translation and by promoting interactions with translation- and RNA-metabolism-associated proteins involved in de novo protein synthesis. Lack of polyphosphate significantly impairs condensate formation, increases cell death, and hinders recovery from N-starvation. In functional analogy, we demonstrate that polyP contributes specifically to the formation of Processing (P)-bodies in human cell lines, revealing that a single, highly conserved and ancestral polyanion serves as a modulator for functional phase-separated condensate formation across the tree of life.
<p>by Thomas A. Püschel, Rachel M. Hurwitz, Chris Venditti</p>
Humans exhibit a striking and near-universal population-level right-hand preference, an evolutionary singularity unmatched among primates. Despite its pervasiveness, the origins of this lateralization remain poorly understood. Here, we combine phylogenetic comparative methods with meta-analysis to investigate manual lateralization across 41 anthropoid species (<i>n</i> = 2,025), testing longstanding eco-evolutionary hypotheses for handedness direction (mean handedness index, MHI) and strength (mean absolute handedness index, MABSHI). Our models reveal significant phylogenetic signal for both traits and identify <i>Homo sapiens</i> as an evolutionary outlier, exhibiting exceptional rightward bias and strength relative to phylogenetic expectations. However, this outlier status disappears when brain size (endocranial volume) and intermembral index are included, suggesting these factors are central to the emergence of human handedness. We also show that high MABSHI evolved early in hominin evolution, while MHI increased to unparalleled levels with the appearance of the genus <i>Homo</i>. Our findings identify bipedalism and neuroanatomical expansion as likely key drivers of uniquely human lateralization, while also revealing broader ecological patterns shaping handedness across primates. This work provides a framework for disentangling human-specific adaptations from general primate trends in the evolution of behavioral asymmetries.
<p>by Natan Horáček, Ondřej Lukšan, Zarley Rebholz, Karel Harant, Radek Pohl, Lana Mutabdžija-Nedelcheva, Simon Hellemans, Daniel Jungwirth, Jan Křivánek, Anna Amirianová, Pavlína Kyjaková, Thomas Bourguignon, Dorothea Tholl, Robert Hanus, Jitka Štáfková</p>
Termites produce the most diverse array of terpenoids among terrestrial metazoans, comprising over 200 structures. However, their biosynthesis has not yet been elucidated. Here, we identify a gene family which arose through a series of duplications of geranylgeranyl pyrophosphate synthase in the common ancestor of Neoisoptera, the terpene-producing termite lineage. We functionally characterized several proteins from this rich GGPPS-like family as terpene synthases generating biologically relevant sesqui- and diterpenes. These include the queen pheromone (3<i>R</i>,6<i>E</i>)-nerolidol in <i>Embiratermes neotenicus</i> and the presumed precursor of polycyclic defensive diterpenes (<i>E</i>,<i>E</i>,<i>E</i>)-neocembrene in <i>Nasutitermes takasagoensis</i>. We report significant enrichment of transposable elements in the GGPPS-like genomic loci, study the selection pressures acting in the evolution of the GGPPS-like family, and highlight an amino acid site crucial for cyclization capacity and enantiospecificity of the characterized enzymes. We conclude that we have identified an enzyme family that facilitated the emergence of the remarkable richness of termite terpenoids.