Biology subjects

Sousa, A.

Publications and source records attributed to Sousa, A..

17 recordsLinked to original sources

bioRxiv2026

Elevated MPS1 converts the fibrous corona into a source of chromosomal instability in colon cancer

Chromosomal instability (CIN), characterised by recurrent chromosome mis-segregation, fuels intratumour heterogeneity and tumour evolution. Yet, its proximal molecular causes remain ill-defined. Tumour-scale genomic and transcriptomic analyses associate MPS1 overexpression with CIN in colon carcinomas. Consistently, CIN+ patient-derived colon cancer cells (PCCCs) display elevated MPS1, frequent merotelic kinetochore-microtubule attachments and lagging chromosomes that are suppressed by partial MPS1 inhibition. Conversely, ectopic MPS1 overexpression in otherwise stable near-diploid cells phenocopies these defects, inducing micronuclei formation and karyotypic divergence. Mechanistically, MPS1 overactivity maintains ROD phosphorylation at Thr13/Ser15, continuously sustaining fibrous corona assembly despite mature end-on attachments. 3D-STED microscopy reveals split-interface configurations in which microtubules from opposing poles engage the canonical outer-kinetochore surface and a spatially distinct corona domain, explaining how merotely forms and persists into anaphase. Disrupting corona assembly restores chromosome segregation fidelity in MPS1-overexpressing RPE-1 cells and in CIN+ PCCCs. Together, our findings establish MPS1 overexpression as a driver of CIN and uncover how oncogenic transcriptional rewiring of mitotic signalling converts a transient microtubule-capture structure into a pathological source of merotely and chromosome mis-segregation in colon cancer.

cell biology↗

bioRxiv2025

Does ecological restoration align prokaryotic community structure with natural references across European coastal wetlands?

AbstractCoastal wetlands are crucial for biodiversity and act as critical buffers for carbon sequestration and atmospheric greenhouse gases (GHG) concentrations, yet their degradation often turns them into GHG sources. Restoration is widely implemented to recover these services, but it remains unclear whether interventions successfully reestablish the microbial functional diversity underpinning biogeochemical cycles. We tested the hypothesis that restoration aligns prokaryotic community structure with natural references, analyzing, a European gradient of coastal wetlands, comparing well-preserved, altered, and restored sites in water and sediment. Using 16SrRNA-metabarcoding and IndVal-Analysis, we characterized community assembly identifying diagnostic functional consortia. Results revealed a marked difference in water and sediment recovery after restoration. Bacterioplankton communities rapidly converge to natural references, while sediment microbiome displayed significant "ecological memory". Restored wetlands show sediment communities structurally distinct from well-preserved, retaining alteration-associated guilds decades. Results support the initial hypothesis: restoration processes in coastal wetlands can re-establish communities and metabolisms resembling well-preserved conditions in the water in the short term, while sediments retain microbial communities and metabolisms inherited from altered conditions for a long time. Future strategies must integrate active sediment interventions using molecular bioindicators to validate not only the landscape appearance, but the effective reactivation of ecosystem processes and microbiota-related services.

ecology↗

bioRxiv2025

A scalable proteogenomic framework for dissecting phospho-signaling pathways in primary immune cells

Signaling networks modulated by post-translational modifications orchestrate cellular responses to external cues. Traditional approaches to study these pathways lack the throughput to systematically capture the causal architecture of these signaling pathways at scale. Here, we present an integrated proteogenomic framework that combines saturating genetic perturbations with high-throughput proteomics to systematically map cytokine-induced signaling in primary human T cells. Supporting this framework is simplePhos, a streamlined, low-input phosphoproteomics workflow that enables scalable, time-resolved analysis without the requirement for specialized equipment or robotics. We extensively validate the simplePhos pipeline by applying inflammatory stimuli, including type I and II interferons, lipopolysaccharide, and Sendai virus to primary T cells and myeloid cells, establishing foundational datasets in these treatment contexts. Ultimately, using type I interferon signaling in genetically modified T cells as a model, we demonstrate that combined application of genetic alterations and proteomic analyses can map key signaling nodes in primary immune cells. This represents a powerful strategy to mechanistically interrogate phospho-signaling networks in human immune cells, with broad applications in translational immunology and therapeutic development.

systems biology↗

bioRxiv2025

Aging: an inevitable road toward gut microbiota pathoadaptation

Laboratory-raised mice live approximately seven times longer and healthier lives compared to their wild counterparts, due to a standardized healthy diet and limited exposure to environmental stressors1. Aging is associated with increased inflammation and microbial dysbiosis2-4. Collectively, these influence microbiota evolution and may contribute to the enrichment in pathobiont frequency observed in old age4. Alternatively, this increase could stem from a decline in colonization resistance5,6, creating favorable conditions for pathobiont invasion. Here, we sought to test whether aging in healthy, controlled conditions, could prevent the selection of age-associated pathobionts. We have followed the adaptive evolution of a commensal strain of Escherichia coli in the guts of mice of advanced age and found that it acquired several mutations common to bacteria colonizing young mice, which were absent in old animals. This, together with the increase in Akkermansia muciniphila in mice of advanced age, suggest healthy aging7,8. However, mutations acquired exclusively in the older were mainly pathoadaptive, tuning the metabolism to oxygen and iron availability, hypermotility, and biofilm formation. In summary, while the evolutionary signature in the guts of very old mice shows youth-like features that may be associated with longevity, the selection of pathoadaptive traits is magnified in very old age. While suggesting that a breach in colonization resistance is not needed to justify the abundance of age-associated pathobionts, our findings raise the question whether specialized bacteria, as opposing to generalists such as E. coli, will display the same ability to evolve pathoadaptive traits. HighlightsO_LIGut commensals face increasingly personalized selective pressures in the aging gut C_LIO_LIEven healthy aging selects for pathoadaptive traits in the gut microbiota C_LIO_LIE. coli s adaptive pattern better reflects the metabolome than microbiota composition C_LI In BriefPathobionts are often enriched in the microbiota of the elderly. Melo-Miranda et al. showed that irrespectively of limiting the opportunity for gut invasion, the strength of selection for pathoadaptation increases with aging. Yet, the gut environment of extreme ages seems to converge, highlighting the discontinuity of the aging process.

evolutionary biology↗

bioRxiv2024

Preemptive minocycline decreases allodynia and depressive-like behaviors in a peripheral neuropathy rat model: a preliminary study

Background and ObjectiveThe neuroimmune system plays a critical role at all phases of chronic pain including at its onset. We therefore hypothesized that preemptive immunomodulation could decrease susceptibility and/or offer protection to pain. Materials and MethodsSix days before the spared nerve injury (SNI), Wistar Han rats were treated with either the immunomodulator minocycline (MIN) or vehicle (VEH). After that, half of the animals from each group switch treatments for additional 7 days, resulting in 4 groups: continuous treatment (MIN/MIN), pretreatment (MIN/VEH), early treatment (VEH/MIN) and the control (VEH/VEH). Mechanical allodynia was recorded using Von Frey test until 4 weeks after SNI where depressive-like behaviors of the animals were also assessed using sucrose preference test. ResultsThe continuous treatment provided sustained protection against mechanical allodynia, with rats in this group showing a significantly higher threshold to pain sensitivity compared to those in VEH/VEH. In contrast, pain relief effects were not observed with MIN/VEH and VEH/MIN. Additionally, animals in MIN/MIN, and VEH/MIN exhibited decreased anhedonic-like behavior at 30 days after SNI, relative to the control. ConclusionsThe exposure to an anti-inflammatory drug circa the installation of a neuropathic lesion had a positive impact on allodynia and on anhedonic behavior for a relatively long period after treatment cessation. The results support the assertions that pain trajectories can be altered at pain early stages by targeting the neuroimmune system. This proof-of-concept has the potential to be broadened to other drugs and/or therapeutical schemes. HighlightsChronic pain susceptibility varies across individuals. (endo)phenotypes of susceptibility manifest prior/circa pain onset. Preemptive minocycline treatment transiently decreases allodynia. Preemptive minocycline treatment provides sustained reduction of depressive-like behavior. Immune system modulation prior to pain onset or at early stages have beneficial outcomes.

neuroscience↗

bioRxiv2024

To FID or not to FID: Applying GANs for MRI Image Generation in HPC

With the rapid growth of Deep Learning models and neural networks, the medical data available for training - which is already significantly less than other types of data - is becoming scarce. For that purpose, Generative Adversarial Networks (GANs) have received increased attention due to their ability to synthesize new realistic images. Our preliminary work shows promising results for brain MRI images; however, there is a need to distribute the workload, which can be supported by High-Performance Computing (HPC) environments. In this paper, we generate 256x256 MRI images of the brain in a distributed setting. We obtained an FIDRadImageNet of 10.67 for the DCGAN and 23.54 for the WGAN-GP, which are consistent with results reported in several works published in this scope. This allows us to conclude that distributing the GAN generation process is a viable option to overcome the computational constraints imposed by these models and, therefore, facilitate the generation of new data for training purposes.

bioengineering↗

bioRxiv2024

MAC: An Artifact Correction Framework for Brain MRI based on Deep Neural Networks

The correction of artifacts in Magnetic Resonance Imaging (MRI) is increasingly relevant as voluntary and involuntary artifacts can hinder data acquisition. Reverting from corrupted to artifact-free images is a complex task. Deep Learning (DL) models have been employed to preserve data characteristics and to identify and correct those artifacts. We propose MOANA, a novel DL-based solution to correct artifacts in multi-contrast brain MRI scans. MOANA offers two models: the simulation and the correction models. The simulation model introduces perturbations similar to those occurring in an exam while preserving the original image as ground truth; this is required as publicly available datasets rarely have motion-corrupted images. It allows the addition of three types of artifacts with different degrees of severity. The DL-based correction model adds a fourth contrast to state-of-the-art solutions while im-proving the overall performance of the models. MOANA achieved the highest results in the FLAIR contrast, with a Structural Similarity Index Measure (SSIM) of 0.9803 and a Normalized Mutual Information (NMI) of 0.8030. With this, the MOANA model can correct large volumes of images in less time and adapt to different levels of artifact severity, allowing for better diagnosis.

bioinformatics↗

bioRxiv2024

A human-specific enhancer fine-tunes radial glia potency and corticogenesis

Humans evolved an extraordinarily expanded and complex cerebral cortex, associated with developmental and gene regulatory modifications1-3. Human accelerated regions (HARs) are highly conserved genomic sequences with human-specific nucleotide substitutions. Although there are thousands of annotated HARs, their functional contribution to human-specific cortical development is largely unknown4,5. HARE5 is a HAR transcriptional enhancer of the WNT signaling receptor Frizzled8 (FZD8) active during brain development6. Here, using genome-edited mouse and primate models, we demonstrate that human (Hs) HARE5 fine-tunes cortical development and connectivity by controlling the proliferative and neurogenic capacity of neural progenitor cells (NPCs). Hs-HARE5 knock-in mice have significantly enlarged neocortices containing more neurons. By measuring neural dynamics in vivo we show these anatomical features correlate with increased functional independence between cortical regions. To understand the underlying developmental mechanisms, we assess progenitor fate using live imaging, lineage analysis, and single-cell RNA sequencing. This reveals Hs-HARE5 modifies radial glial progenitor behavior, with increased self-renewal at early developmental stages followed by expanded neurogenic potential. We use genome-edited human and chimpanzee (Pt) NPCs and cortical organoids to assess the relative enhancer activity and function of Hs-HARE5 and Pt-HARE5. Using these orthogonal strategies we show four human-specific variants in HARE5 drive increased enhancer activity which promotes progenitor proliferation. These findings illustrate how small changes in regulatory DNA can directly impact critical signaling pathways and brain development. Our study uncovers new functions for HARs as key regulatory elements crucial for the expansion and complexity of the human cerebral cortex.

neuroscience↗

bioRxiv2021

Pan-Cancer landscape of protein activities identifies drivers of signalling dysregulation and patient survival

Genetic alterations in cancer cells trigger oncogenic transformation, a process largely mediated by the dysregulation of kinase and transcription factor (TF) activities. While the mutational profiles of thousands of tumours has been extensively characterized, the measurements of protein activities has been technically limited until recently. We compiled public data of matched genomics and (phospho)proteomics measurements for 1,110 tumours and 77 cell lines that we used to estimate activity changes in 218 kinases and 292 TFs. Kinase activities are, on average, not strongly determined by protein abundance but rather by their phosphorylation state while the reverse is more common for TFs. Co-regulation of kinase and TF activities reflects previously known regulatory relationships and allows us to dissect genetic drivers of signalling changes in cancer. Loss-of-function mutation is not often associated with dysregulation of downstream targets, suggesting frequent compensatory mechanisms. Finally, we identified the activities most differentially regulated in cancer subtypes and showed how these can be linked to differences in patient survival. Our results provide broad insights into dysregulation of protein activities in cancer and their contribution to disease severity.

cancer biology↗

bioRxiv2021

MRI Radiomic Signature of White Matter Hyperintensities Is Associated with Clinical Phenotypes

IntroductionNeuroimaging measurements of brain structural integrity are thought to be surrogates for brain health, but precise assessments require dedicated advanced image acquisitions. By means of describing the texture of conventional images beyond what meets the naked eye, radiomic analyses hold potential for evaluating brain health. We sought to: 1) evaluate this novel approach to assess brain structural integrity by predicting white matter hyperintensities burdens (WMH) and 2) uncover associations between predictive radiomic features and patients clinical phenotypes. MethodsOur analyses were based on a multi-site cohort of 4,163 acute ischemic strokes (AIS) patients with T2-FLAIR MR images and corresponding deep-learning-generated total brain and WMH segmentation. Radiomic features were extracted from normal-appearing brain tissue (brain mask-WMH mask). Radiomics-based prediction of personalized WMH burden was done using ElasticNet linear regression. We built a radiomic signature of WMH with the most stable selected features predictive of WMH burden and then related this signature to clinical variables (age, sex, hypertension (HTN), atrial fibrillation (AF), diabetes mellitus (DM), coronary artery disease (CAD), and history of smoking) using canonical correlation analysis. ResultsRadiomic features were highly predictive of WMH burden (R2=0.855{+/-}0.011). Seven pairs of canonical variates (CV) significantly correlated the radiomics signature of WMH and clinical traits with respective canonical correlations of 0.81, 0.65, 0.42, 0.24, 0.20, 0.15, and 0.15 (FDR-corrected p-valuesCV1-6<.001, p-valueCV7=.012). The clinical CV1 was mainly influenced by age, CV2 by sex, CV3 by history of smoking and DM, CV4 by HTN, CV5 by AF and DM, CV6 by CAD, and CV7 by CAD and DM. ConclusionRadiomics extracted from T2-FLAIR images of AIS patients capture microstructural damage of the cerebral parenchyma and correlate with clinical phenotypes. Further research could evaluate radiomics to predict the progression of WMH. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSWe did a systematic review on PubMed until December 1, 2020, for original articles and reviews in which radiomics were used to characterize stroke or cerebrovascular diseases. Radiomic analyses cover a broad ensemble of high-throughput quantification methods applicable to digitalized medical images that extract high-dimensional data by describing a given region of interest by its size, shape, histogram, and relationship between voxels. We used the search terms "radiomics" or "texture analysis", and "stroke", "cerebrovascular disease", "small vessel disease", or "white matter hyperintensities". Our research identified 24 studies, 18 studying radiomics of stroke lesions and 6 studying cerebrovascular diseases. All the latter six studies were based on MRI (T1-FLAIR, dynamic contrast-enhanced imaging, T1 & T2-FLAIR, T2-FLAIR post-contrast, T2-FLAIR, and T2-TSE images). Four studies were describing small vessel disease, and two were predicting longitudinal progression of WMH. The average sample size was small with 96 patients included (maximum: 204). One study on 141 patients identified 7 T1-FLAIR radiomic features correlated with cardiovascular risk factors (age and hyperlipidemia) using univariate correlations. All studies were monocentric and performed on a single MRI scanner. Added value of this studyTo date and to the best of our knowledge, this is the largest radiomics study performed on cerebrovascular disease or any topic, and one of the very few to include a great diversity of participating sites with diverse clinical MRI scanners. This study is the first one to establish a radiomic signature of WMH and to interpret its relationship with common cardiovascular risk factors. Our findings add to the body of evidence that damage caused by small vessel disease extend beyond the visible white matter hyperintensities, but the added value resides in the detection of that subvisible damage on routinely acquired T2-FLAIR imaging. It also suggests that cardiovascular phenotypes might manifest in distinct textural patterns detectable on conventional clinical-grade T2-FLAIR images. Implications of all the available evidenceAssessing brain structural integrity has implications for treatment selection, follow-up, prognosis, and recovery prediction in stroke patients but also other neurological disease populations. Measuring cerebral parenchymal structural integrity usually requires advanced imaging such as diffusion tensor imaging or functional MRI. Translation of those neuroimaging biomarkers remains uncommon in clinical practice mainly because of their time-consuming and costly acquisition. Our study provides a potential novel solution to assess brains structural integrity applicable to standard, routinely acquired T2-FLAIR imaging. Future research could, for instance, benchmark this radiomics approach against diffusion or functional MRI metrics in the prediction of cognitive or functional outcomes after stroke.

neuroscience↗

bioRxiv2020

Sex-specific lesion topographies explain outcomesafter acute ischemic stroke

Acute ischemic stroke affects men and women differently in many ways. In particular, women are oftentimes reported to experience a higher acute stroke severity than men. Here, we derived a low-dimensional representation of anatomical stroke lesions and designed a sex-aware Bayesian hierarchical modelling framework for a large-scale, well phenotyped stroke cohort. This framework was tailored to carefully estimate possible sex differences in lesion patterns explaining acute stroke severity (NIHSS) in 1,058 patients (39% female). Anatomical regions known to subserve motor and language functions emerged as relevant regions for both men and women. Female patients, however, presented a more widespread pattern of stroke severity-relevant lesions than male patients. Furthermore, particularly lesions in the posterior circulation of the left hemisphere underlay a higher stroke severity exclusively in women. These sex-sensitive lesion pattern effects could be discovered and subsequently robustly replicated in two large independent, multisite lesion datasets. The constellation of findings has several important conceptual and clinical implications: 1) suggesting sex-specific functional cerebral asymmetries, and 2) motivating a sex-stratified approach to management of acute ischemic stroke. To go beyond sex-averaged stroke research, future studies should explicitly test whether acute therapies administered on the basis of sex-specific cutoff volumes of salvageable tissue will lead to improved outcomes in women after acute ischemic stroke.

neuroscience↗

bioRxiv2019

Perceptual spatial suppression and sensitivity for motion are weakened in schizophrenia

Perceptual spatial suppression is a phenomenon in which the perceived strength of a stimulus in some region of the space is reduced when the stimulus is surrounded by other stimuli. For contrast perception, several studies suggest that spatial suppression is reduced in patients with schizophrenia. For motion perception, only one study has been conducted in a cohort of 16 patients, suggesting that spatial suppression is reduced. It is unknown, however, whether this reduction is related to the lower IQ that schizophrenic patients usually show; as there is evidence that spatial suppression for motion increases with IQ in healthy individuals. Here, we sought to determine the spatial suppression for motion in a larger cohort of 33 patients with schizophrenia controlling for IQ. We found a weakened spatial suppression in patients with schizophrenia, consistent with the previous study (g = 0.47, CI = [0.055, 0.88], combining the previous and our study). For comparison, we performed a meta-analysis on spatial suppression for contrast and found a similar effect size. We found that patients had a lower IQ than controls, but this difference did not explain their weaker spatial suppression. Further, we found that spatial suppression of patients, but not controls increased with their IQ and decreased with age in both groups. Finally, as we estimated lapses of attention, we could estimate motion sensitivity and found that it was decreased in patients. We speculate about possible alterations in neurotransmission that might explain the reduced spatial suppression and sensitivity that we found.

neuroscience↗

bioRxiv2019

Beyond the suprachiasmatic nucleus: circadian rhythmicity shapes astrocyte morphology and neuronal function in hippocampal area CA1

Most animal species operate according to a 24-hour period set by the suprachiasmatic nucleus (SCN) of the hypothalamus. The rhythmic activity of the SCN is known to modulate hippocampal-dependent memory processes, but the molecular and cellular mechanisms that account for this effect remain largely unknown. Here, we show that there are cell-type specific structural and functional changes that occur with circadian rhythmicity in neurons and astrocytes in hippocampal area CA1. Pyramidal neurons change the surface expression of NMDA receptors, whereas astrocytes change their proximity to synapses. Together, these phenomena alter glutamate clearance, receptor activation and integration of temporally clustered excitatory synaptic inputs, ultimately shaping hippocampal-dependent learning in vivo. We identify corticosterone as a key contributor to changes in synaptic strength. These findings identify important mechanisms through which neurons and astrocytes modify the molecular composition and structure of the synaptic environment, contribute to the local storage of information in the hippocampus and alter the temporal dynamics of cognitive processing.

neuroscience↗

bioRxiv2019

Whole-genome and RNA sequencing reveal variation and transcriptomic coordination in the developing human prefrontal cortex

Variation in gene expression underlies neurotypical development, while genomic variants contribute to neuropsychiatric disorders. BrainVar is a unique resource of paired whole-genome sequencing and bulk-tissue RNA-sequencing from the human dorsolateral prefrontal cortex of 176 neurotypical individuals across prenatal and postnatal development, providing the opportunity to assay genomic and transcriptomic variation in tandem. Leveraging this resource, we identified rare premature stop codons with commensurate reduced and allele-specific expression of corresponding genes, and common variants that alter gene expression (expression quantitative trait loci, eQTLs). Categorizing eQTLs by prenatal and postnatal effect, genes affected by temporally-specific eQTLs, compared to constitutive eQTLs, are enriched for haploinsufficiency, protein-protein interactions, and neuropsychiatric disorder risk loci. Expression levels of over 12,000 genes rise or fall in a concerted late-fetal transition, with the transitional genes enriched for cell type specific genes and neuropsychiatric disorder loci, underscoring the importance of cataloguing developmental trajectories in understanding cortical physiology and pathology.\n\nHighlightsO_LIWhole-genome and RNA-sequencing across human prefrontal cortex development in BrainVar\nC_LIO_LIGene-specific developmental trajectories characterize the late-fetal transition\nC_LIO_LIIdentification of constitutive, prenatal-specific, postnatal-specific, and rare eQTLs\nC_LIO_LIIntegrated analysis reveals genetic and developmental influences on CNS traits and disorders\nC_LI

neuroscience↗

bioRxiv2018

Multi-omics characterization of interaction-mediated control of human protein abundance levels

Proteogenomic studies of cancer samples have shown that copy number variation can be attenuated at the protein level, for a large fraction of the proteome, likely due to the degradation of unassembled protein complex subunits. Such interaction mediated control of protein abundance remains poorly characterized. To study this we compiled genomic, (phospho)proteomic and structural data for hundreds of cancer samples and find that up to 42% of 8,124 analyzed proteins show signs of post-transcriptional control. We find evidence of interaction dependent control of protein abundance, correlated with interface size, for 516 protein pairs, with some interactions further controlled by phosphorylation. Finally, these findings in cancer were reflected in variation in protein levels in normal tissues. Importantly, expression differences due to natural genetic variation were increasingly buffered from phenotype differences for highly attenuated proteins. Altogether, this study further highlights the importance of post-transcriptional control of protein abundance in cancer and healthy cells.

genomics↗

bioRxiv2018

Sex overrides mutation in Escherichia coli colonizing the gut

Bacteria evolve by mutation accumulation in laboratory experiments, but the tempo and mode of evolution in natural environments are largely unknown. Here we show, by experimental evolution of E. coli in the mouse gut, that the ecology of the gut controls bacterial evolution. If a resident E. coli strain is present in the gut, an invading strain evolves by rapid horizontal gene transfer; this mode precedes and outweighs evolution by point mutations. An epidemic infection by two phages drives gene uptake and produces multiple co-existing lineages of phage-carrying (lysogenic) bacteria. A minimal dynamical model explains the temporal pattern of phage epidemics and their complex evolutionary outcome as generic effects of phage-mediated selection. We conclude that phages are an important eco-evolutionary driving force - they accelerate evolution and promote genetic diversity of bacteria.\n\nOne Sentence SummaryBacteriophages drive rapid evolution in the gut.

evolutionary biology↗

bioRxiv2016

Rampant Reverse Evolution towards Specialization in Commensal Bacteria Colonizing the Gut

The maintenance of diversity in the gut microbiota is a signature of host health. Yet how strain variation emerges and changes over time in this ecosystem is poorly understood. Here we use a natural yet controlled system to track the effects of natural selection by the genetic signatures it leaves in evolving populations. By following the emergence of intra-species diversity in an Escherichia coli strain, we unravel a recurrent case of violation of Dollos law, which proposes that evolution is unidirectional and irreversible. We demonstrate de novo acquisition of a primordial lost phenotype via compensatory mutation and also genetic reversion, the latter leaving no trace of the past. We show that this reverse evolution generates two coexisting phenotypes, resource generalist and specialist, whose abundance can be controlled by diet supplementation. While specialists abundance is low, they avoid competition with the gut microbiota, whereas generalist abundance is dependent on microbiota composition. Our results highlight how a single genetic change can have large ecological consequences.

evolutionary biology↗