Project 10

Background: Patients with inflammatory bowel disease (IBD) have an increased risk for cardiovascular diseases (CVD). The underlying pathophysiological mechanisms are not fully understood particularly since the traditional CVD risk factors, such as obesity or hyperlipidemia, are typically absent in these patients. Recent evidence indicates a connection between chronic inflammation and the positive selection of somatic mutations in “driver” genes of hematopoietic stem cells (HSC) that lead to clonal events. These mutant HSC clones give rise to progeny immune cells that harbor the mutant allele, and this can lead to the functional corruption of the circulating immune cell, a condition called clonal hematopoiesis of indeterminate potential (CHIP) ¹. Of note, CHIP is also associated with a doubling of the risk of atherosclerotic CVD ², and somatic mutation in gene Tp53 was characterized as driver of atherosclerosis ³. Recently, it was reported that CHIP is elevated in patients with ulcerative colitis ⁴. It is speculated that the inflammatory environment of the disease promotes HSC clone growth, and this is of interest due to the increased risk of CVD in IBD patients. However, epidemiological studies do not address causality, directionality or mechanistic aspects of the relationships between IBD, changes in the heterocellular bone marrow (BM) niche, CHIP and CVD.

Preliminary work: In pilot studies we developed a model of inducible chronic colitis in Apolipoprotein E (Apoe)-deficient mice and showed that the induction of colitis by dextran sodium sulfate (DSS) led to increased atherogenic plaque progression. Further, structural changes in the BM were observed, which provides an essential microenvironment for hematopoiesis and proper hematopoietic stem cell (HSC) function and survival. These changes in the cellular and non-cellular composition of the BM niche might potentially influence CHIP. In addition, using an experimental model of CHIP, we showed that the Tp53^R270H/+-driver gene increased leukocyte progenitors and mature immune cells in the BM following the induction of colitis. 

Hypothesis: We hypothesize that IBD-driven changes in the heterocellular BM niche accelerate atherosclerosis via the development of CHIP.

Research program and work packages (WP): Specifically, we aim to (i) evaluate colitis-induced changes in BM mesenchymal stem cells (MSCs), HSCs and mature immune cells with a transcriptome analysis at single cell level, (ii) investigate how these colitis-mediated effects modulate CHIP driven by the Tp53 mutation, (iii) elucidate whether colitis favoring CHIP driven by the Tp53 mutation promotes smooth muscle cell (SMC) and endothelial cell (EC) phenotypic switching and increases atherosclerosis. 

WP1: Investigating colitis-induced cellular plasticity in the BM niche. In order to transfer the model of combined colitis and accelerated atherosclerosis to all genetic backgrounds, adeno-associated virus vector (AAV)8 – PCSK9 virus injections will be used to induce hyperlipidemia and subsequent atherosclerosis in C57BL6J wildtype mice. This model will be combined with DSS-induced colitis (both models established in the Grandoch lab) and changes in the BM microenvironment (BM adipose tissue, extracellular matrix, MSCs, HSCs, mature immune cells) will be investigated in detail by the doctoral researcher in the Grandoch lab. Specifically, flow cytometric analyses of the cellular composition of the BM niche will be performed to detect colitis-driven changes in MSCs, HSCs and mature immune cells. This will be completed by (immuno)histological methods for staining of different matrix components such as hyaluronan, collagen or proteoglycans. A doctoral researcher from the Walsh lab will functionally characterize HSCs and MSC from mice with and without colitis regarding their differentiation capacity and interaction ex vivo.

The colitis-induced transcriptomic signature will further be investigated in detail at single cell level together with the Doerr lab. To this end, single-cell RNA (scRNA) sequencing of isolated MSCs, HSCs and mature immune cells are performed both in mice with and without colitis. In addition, full-length transcriptome bulk-sequencing for each of the three cell types is carried out on our in-house PacBio Revio system with the KINNEX RNA kit, enabling the identification of isoforms and analysis of alternative splicing. The doctoral researcher in the Doerr lab will develop computational methods to integrate gene expression data from both scRNA sequencing and long-read transcriptome sequencing and subsequently study isoform-resolved expression profiles at clonal level. Then they will compare expression profiles between the three cell types and conditions and characterize changes in clonal composition and inflammatory responses. Further, they will derive intercellular crosstalk between MSCs and HSCs, and mature immune cells from scRNA sequencing data.

WP2: Investigating the expansion of CHIP in mice with chronic colitis. In order to verify the effects of the BM niche on CH, the adoptive transfer BM transfer model to nonconditioned mice established in the Walsh lab ⁵ will be used, thereby avoiding potential confounders such as irradiation. After three cycles of colitis and priming of the BM niche as described in WP1, mice undergo the adoptive bone marrow transfer using Tp53 mutated cells. During the stay at the partner lab at the UVA, doctoral researchers from the Grandoch and the Walsh lab will analyze changes in the cellular and non-cellular composition of the BM niche and expansion of the mutant donor cells during a time period of 9 weeks by flow cytometry and (immuno)histology as described in WP1. Additionally, spatial transcriptomics of the BM niche will be performed on bone sections thereby enabling the investigation of cellular plasticity within the context of their relative locations within the niche. The bioinformatic analysis is performed by the doctoral researcher of the Doerr lab during their visit at UVA.

WP3: Analyzing the impact of chronic colitis and CHIP on aortic cellular plasticity. Together with the Owens lab progression of atherosclerosis, SMC/EC plasticity and phenotypic switching will be analyzed comparing the effects of colitis alone or in combination with concomitant CHIP. For this approach, the model of combined colitis and CHIP will be used in SMC/EC dual lineage tracing mice from the Owens lab. These athero-prone mice with and without colitis and with and without CHIP will be analyzed for atherosclerotic plaque development, immune cell infiltration in the vascular wall and SMC/EC phenotypic switching by doctoral researchers of the Grandoch and Owens lab. En face staining of the aorta, (immune)histological staining of the aortic root and the brachiocephalic artery as well as flow cytometry of aortic inflammatory cell composition will be used to characterize atherosclerotic immune signatures in colitis-driven CHIP. As described in WP1, scRNA and full-length transcriptome sequencing will be performed in isolated cells from the BM and the circulation (MSCs, HSCs, mature immune cells), as well as aortic cells to gain deeper insights into intercellular communication networks.

The Grandoch lab has extensive experience with the analysis of matrix components ^6-8 and immune cells in inflammatory conditions ^9-11 using murine models of atherosclerosis as well as the experimental chronic colitis ¹². A broad panel of in vivo and ex vivo methods covering (immuno)histological methods, flow cytometry, transcriptome/proteome analyses (see full list of lab citations) is used to elucidate the underlying pathomechanisms in detail.

The Computational Diabetology group (lead: Dörr) specializes in the development of computational methods for the analysis of multiomics data sets. Further, Dörr is proficient in genomic data analysis ¹³, including phylogenetic reconstruction ^{14, 15} and pangenomics ¹⁶. 

The Walsh group has extensive expertise in clonal hematopoiesis and how it contributes to cardio-metabolic disease. Their 2017 study was the first to causally connect CHIP with CVD. In toto, they have employed various CHIP models to evaluate multiple “driver” genes (Tet2, Dnmt3a, JAK2^V617F, Ppm1d, Tp53, Asxl1 and mLOY) in experimental models of cardiometabolic disease including atherosclerosis, ischemic and non-ischemic heart failure ⁵, ischemic stroke, hypertension and diet-induced obesity ¹ Mechanistic findings from these experimental systems were the basis for the paradigm that CHIP promotes CVD via the overactivation of inflammatory signaling. These experimental findings relating CHIP to excessive inflammation have been subsequently validated in numerous clinical studies. 

The Owens` lab has a strong focus on smooth muscle cell phenotypic switching in the development and progression of atherosclerosis as well as mechanisms that control plaque stability ^{17, 18}. They also have extensive expertise in studies of the role of inflammation in late-stage lesion pathogenesis.¹⁹

Due to the high complexity of the project, collaboration of the four participating groups is necessary. Doctoral researchers from all labs will work together and combine their expertise and murine models. In the Grandoch lab, the model of chronic colitis and characterization of the BM niche are established and doctoral researchers from the Walsh´s and Owens´ labs will use their stay at HHU to transfer these expertise to the UVA. In turn, the doctoral researcher from the Grandoch lab will learn the model of adoptive transfer and analysis of CH in the Walsh lab as well as use and analyses of the lineage tracing mice in the Owens lab.

With this project, we aim to perform seminal studies that will elucidate the alterations in the BM microenvironment as well as in the vascular wall during chronic inflammatory processes that promote CHIP in IBD, thereby driving the increased CVD risk in patients with IBD. The laboratories involved in this project work in a perfectly complementary way to shed light on the issue from different angles. This work will be complemented by bioinformatic analyses carried out by the PhD student from the Dörr lab. In particular, the student will establish new methodology for integrating scRNA-seq data with full-length transcriptome data from bulk-sequencing. During the visit in the Walsh lab at UVA, the student will be trained in analyzing spatial transcriptomics data of WP2 and collaborate with the PhD student from the Grandoch lab by performing the bioinformatic analysis of their CHIP mouse model experiments. Conversely, a partnering doctoral researcher from the Walsh lab will visit the Doerr lab to conduct a comparative analysis of the spatial transcriptomics data from the Walsh lab and the single cell transcriptomics data from WP1/3.