Periodic Reporting for period 1 - CartoHostBug (Functional cartography of intestinal host-microbiome interactions)
Période du rapport: 2024-07-01 au 2025-12-31
Microbiome-host crosstalk is crucial for gut homeostasis and its dysregulation is a hallmark of diseases such as inflammatory bowel disease (IBD) and colorectal cancer (CRC). Despite its key relevance for a holistic understanding of the human superorganism and its (patho-)physiology, how local host-microbiome interactions form specific niches in the gut and how these niches function at the cellular and molecular level remains unexplored, mainly due to a lack of suitable technologies.
To fill this gap, we propose to jointly reconstruct the host transcriptional and microbiome compositional landscape of the human gut across a large number of healthy individuals as well as IBD and CRC patients. For this, we will leverage a combination of novel spatial profiling technologies for unbiased transcriptome sequencing and microbiome profiling at single-cell resolution in situ. First, we will spatially delineate local niches formed of specific microbes and host cells. To dissect this complex crosstalk into specific interactions, we will secondly use in vitro and in vivo models to introduce perturbations either on the host or the microbiome side. Finally, we will integrate the resulting data to deconvolute host-microbiome circuits computationally and to predict functional niches, in particular host responses to pathogens of relevance in IBD and CRC. Interesting predictions will be tested in organoid and animal models. Developing a spatially resolved computational model of the gut ecosystem will allow us to predict early local events in disease onset; from these we will identify and validate prognostic IBD and CRC biomarkers for future clinical translation.
This work will revolutionize our understanding of intestinal host-microbiome interactions by adding a first-ever functionally resolved spatial dimension with clinical relevance for the future diagnosis and treatment of intestinal disorders.
To fill this gap, we propose to jointly reconstruct the host transcriptional and microbiome compositional landscape of the human gut across a large number of healthy individuals as well as IBD and CRC patients. For this, we will leverage a combination of novel spatial profiling technologies for unbiased transcriptome sequencing and microbiome profiling at single-cell resolution in situ. First, we will spatially delineate local niches formed of specific microbes and host cells. To dissect this complex crosstalk into specific interactions, we will secondly use in vitro and in vivo models to introduce perturbations either on the host or the microbiome side. Finally, we will integrate the resulting data to deconvolute host-microbiome circuits computationally and to predict functional niches, in particular host responses to pathogens of relevance in IBD and CRC. Interesting predictions will be tested in organoid and animal models. Developing a spatially resolved computational model of the gut ecosystem will allow us to predict early local events in disease onset; from these we will identify and validate prognostic IBD and CRC biomarkers for future clinical translation.
This work will revolutionize our understanding of intestinal host-microbiome interactions by adding a first-ever functionally resolved spatial dimension with clinical relevance for the future diagnosis and treatment of intestinal disorders.
1.1 Cartography of human intestinal microbiome - host niches in health and disease
We have profiled twelve colorectal tumors (6 MSS and 6 MSI) using µFISH and Xenium (GZ). This initial analysis of the samples allowed the development of analysis methodology for image preprocessing, detection and taxonomic identification (decoding of fluorescence barcodes) of bacterial spots in tissue sections. In addition, we implemented image registration across adjacent tissue sections, each analysed with a different spatial modality (µFISH and Xenium), colocalization analysis (bacterial-bacterial and bacterial-host cell colocalization), and host gene expression analysis according to spatial niches and bacterial colocalization.
For inflammatory bowel disease (IBD) samples, we have performed all the ethical and bureaucratic work necessary to set up a biobank (EJV, SG). In addition, we have completed the set up of patient consent forms to collect IBD biopsies. In parallel, we have performed curation and initial analysis of publicly available single-cell datasets profiling large IBD patient cohorts to describe their conserved inflammatory multicellular processes (in collaboration with Carl Anderson)
1.2 Development of spatial profiling technology
We have been working experimentally on a new chemistry for Spatial metaTranscriptomics (SmT) to enable the capture of bacterial and fungal signals compatible with the Visium HD platform (SG). Our approach considers in situ priming of RNA molecules for both microbes and host genes. Preliminary results are encouraging, showing that our new chemistry is working. We are now focusing on further optimizations to increase the signal detected. This new approach opens the possibility of integrating SmT with different types of host transcriptomics assays and its applicability to formalin fixed paraffin embedded samples.
In parallel, we have also been working on further establishing and validating the µFISH methodology, in particular probe design workflow including computational and in vitro (against bacterial isolates) validation protocols. Our first design of FISH probes has been focused on CRC-associated bacteria.
2.1 Perturbational profiling of M→H
To perturb the microbiota in vivo, we colonized germ-free (GF) mice with distinct microbiomes, including control specific pathogen–free (SPF) and colitogenic microbiota. We validated that mice colonized with the colitogenic microbiome developed exacerbated colitis and harbored a dysbiotic microbiome composition.
Using spatial transcriptomics, we analyzed the impact of distinct microbiota on the colonic transcriptomic landscape and identified gene programs and pathways that are influenced—or not—by a colitogenic microbiota
To visualize host–microbiome niches and specific microbes co-localizing with immunological features of interest in colitogenic-colonized mice, we performed Spatial MetaTranscriptomics (SmT). The generated data were analyzed, and we identified specific bacterial taxa associated with high JAK/STAT activity.
Thus, we have conducted our first analysis of how microbiome perturbation shapes host transcriptomic profiles and we are currently working on writing the manuscript.
The Zeller lab has been collecting relevant bacterial strains and contributing to the experimental design to test the effects of specific strains on host transcriptomic responses.
2.2 Perturbational profiling of H→M
To investigate host perturbation and its impact on host–microbiome niches, we will conduct experiments analogous to those described in Section 2.1 employing IL10-KO mice. A lead PhD student has been appointed to oversee this component of the project, and the acquisition of the required animals has been initiated. The experiments are scheduled to commence in April in the EJV laboratory.
In parallel, to facilitate controlled host perturbation studies, we have successfully established murine and human organoid cultures. These platforms are now fully operational and ready for pharmacological and genetic manipulation experiments.
2.3 Computational integration of perturbation signatures
We are performing a comparison of in vivo and in vitro transcriptional signatures of Fusobacterium infection on colorectal cancer, in collaboration with Jens Puschof. We established an RNA-Seq protocol for simultaneous evaluation of bacterial composition and host gene expression (via total cDNA synthesis and selective host rRNA depletion). We are currently exploring single-cell protocols that are based on similar total RNA sequencing to be able to assess bacterial invasion/adhesion to host cells in human and mouse gut tissues.
3.1 Computational reconstruction of a spatial ecosystem map of the intestine
We are developing an extension of the evaluations of the impact of sparse molecular information on the inference of transcriptional signatures.
We are also expanding the spatial colocalization metrics to be applicable for host-microbiome profiling.
We have profiled twelve colorectal tumors (6 MSS and 6 MSI) using µFISH and Xenium (GZ). This initial analysis of the samples allowed the development of analysis methodology for image preprocessing, detection and taxonomic identification (decoding of fluorescence barcodes) of bacterial spots in tissue sections. In addition, we implemented image registration across adjacent tissue sections, each analysed with a different spatial modality (µFISH and Xenium), colocalization analysis (bacterial-bacterial and bacterial-host cell colocalization), and host gene expression analysis according to spatial niches and bacterial colocalization.
For inflammatory bowel disease (IBD) samples, we have performed all the ethical and bureaucratic work necessary to set up a biobank (EJV, SG). In addition, we have completed the set up of patient consent forms to collect IBD biopsies. In parallel, we have performed curation and initial analysis of publicly available single-cell datasets profiling large IBD patient cohorts to describe their conserved inflammatory multicellular processes (in collaboration with Carl Anderson)
1.2 Development of spatial profiling technology
We have been working experimentally on a new chemistry for Spatial metaTranscriptomics (SmT) to enable the capture of bacterial and fungal signals compatible with the Visium HD platform (SG). Our approach considers in situ priming of RNA molecules for both microbes and host genes. Preliminary results are encouraging, showing that our new chemistry is working. We are now focusing on further optimizations to increase the signal detected. This new approach opens the possibility of integrating SmT with different types of host transcriptomics assays and its applicability to formalin fixed paraffin embedded samples.
In parallel, we have also been working on further establishing and validating the µFISH methodology, in particular probe design workflow including computational and in vitro (against bacterial isolates) validation protocols. Our first design of FISH probes has been focused on CRC-associated bacteria.
2.1 Perturbational profiling of M→H
To perturb the microbiota in vivo, we colonized germ-free (GF) mice with distinct microbiomes, including control specific pathogen–free (SPF) and colitogenic microbiota. We validated that mice colonized with the colitogenic microbiome developed exacerbated colitis and harbored a dysbiotic microbiome composition.
Using spatial transcriptomics, we analyzed the impact of distinct microbiota on the colonic transcriptomic landscape and identified gene programs and pathways that are influenced—or not—by a colitogenic microbiota
To visualize host–microbiome niches and specific microbes co-localizing with immunological features of interest in colitogenic-colonized mice, we performed Spatial MetaTranscriptomics (SmT). The generated data were analyzed, and we identified specific bacterial taxa associated with high JAK/STAT activity.
Thus, we have conducted our first analysis of how microbiome perturbation shapes host transcriptomic profiles and we are currently working on writing the manuscript.
The Zeller lab has been collecting relevant bacterial strains and contributing to the experimental design to test the effects of specific strains on host transcriptomic responses.
2.2 Perturbational profiling of H→M
To investigate host perturbation and its impact on host–microbiome niches, we will conduct experiments analogous to those described in Section 2.1 employing IL10-KO mice. A lead PhD student has been appointed to oversee this component of the project, and the acquisition of the required animals has been initiated. The experiments are scheduled to commence in April in the EJV laboratory.
In parallel, to facilitate controlled host perturbation studies, we have successfully established murine and human organoid cultures. These platforms are now fully operational and ready for pharmacological and genetic manipulation experiments.
2.3 Computational integration of perturbation signatures
We are performing a comparison of in vivo and in vitro transcriptional signatures of Fusobacterium infection on colorectal cancer, in collaboration with Jens Puschof. We established an RNA-Seq protocol for simultaneous evaluation of bacterial composition and host gene expression (via total cDNA synthesis and selective host rRNA depletion). We are currently exploring single-cell protocols that are based on similar total RNA sequencing to be able to assess bacterial invasion/adhesion to host cells in human and mouse gut tissues.
3.1 Computational reconstruction of a spatial ecosystem map of the intestine
We are developing an extension of the evaluations of the impact of sparse molecular information on the inference of transcriptional signatures.
We are also expanding the spatial colocalization metrics to be applicable for host-microbiome profiling.
Ethical permits have been reviewed and biobank approvals required for the collection of samples from patients with inflammatory bowel disease (IBD). A multidisciplinary working group has been established, comprising clinicians, nurses, and research personnel. In addition, a standardized workflow has been implemented to ensure high-quality sample collection and processing for spatial transcriptomics and spatial metatranscriptomics (SmT) analyses.