Background and objectives Immune-mediated inflammatory diseases remain clinically heterogeneous and often refractory to targeted therapies. A tissue-centric perspective places disease persistence not only in immune dysregulation but in dysfunctional tissue programs that fail to resolve, leading to chronic structural damage. Within this view, fibroblasts are central organizers of local immunity and repair. This dissertation, built on a computational systems biology framework that integrates multiomics with predictive modeling and multi-stage analyses, examines how fibroblast programs guide tissues toward destructive inflammation, resolution, or fibrosis in the synovium in chronic inflammatory arthritis and in the skin in systemic sclerosis. The objectives are to define fibroblast states linked to resolution versus persistent inflammation in arthritis, to map these states across species and spatial contexts, and to identify stiffness-driven fibroblast programs and trajectories that underpin dermal fibrosis in systemic sclerosis. Methods The work combines a longitudinal clinical study with preclinical models and in vivo imaging of fibroblast activation protein (FAP) to monitor mesenchymal activation over time in arthritis. A cross-model single-cell synovial fibroblast atlas is assembled from complementary mouse models, and cross-species mapping projects the resulting fibroblast states into human datasets, including spatially resolved synovium. Putative partner cells are identified by combining ligand–receptor inference from single-cell data with spatial neighborhood modeling. Downstream effects are inferred from co-expression network analysis. For systemic sclerosis, bulk RNA sequencing of fibroblasts cultured on stiff matrices yields an early stiffness-response gene signature that serves as the basis for subsequent computational analyses. The signature is projected into single-cell data to quantify mechanosensitive programs and to model trajectories. Mechanistic hypotheses are generated computationally including by in silico perturbation and are tested by targeted in vitro assays and in vivo interventions. Results and observations In arthritis, the synovial atlas resolves two broad fibroblast states. One is a FAP-high inflammatory and tissue-destructive state marked by IL6 or MMP3. The other is a low-FAP, CD200+/DDK3+ pro-resolving state that is associated with remission and that appears as conserved niches when mapped into human datasets, including spatial synovium. Computational analysis with in silico perturbation, followed by experimental validation in vitro and in vivo, indicates that the CD200⁺ DKK3⁺ niche engages CD200R1 on ILC2, reduces pro-inflammatory signaling, stabilizes resolution, and prevents a shift from ILC2 to ILC3. Pharmacologic agonism of CD200/CD200R1 mitigated arthritis in preclinical models, supporting the functional relevance of this stromal–immune checkpoint. In systemic sclerosis, the stiffness-response gene signature derived from stiff-matrix cultures quantifies mechanosensitive fibroblast programs in single-cell data and reveals a directed trajectory in which SFRP2⁺ PI16⁺ fibroblasts differentiate toward SFRP2⁺ COMP⁺ myofibroblast-like states with PU.1 activation and increased extracellular matrix production. This trajectory associates strongly with the stiffness-response signature and identifies matrix stiffness as a principal driver. Analysis of mechanical-strain datasets, with and without inhibition of mechanotransduction pathways, supports a direct link from matrix stiffness to myofibroblast reprogramming. Spatial mapping places these states in mechanically and clinically affected skin and suggests a positive feedback between stiffness and fibrosis. Conclusions and discussion Across arthritis and systemic sclerosis, fibroblasts act as architects of tissue fate. Through crosstalk with immune cells and sensitivity to physical cues, they encode destructive inflammation, resolution, or fibrosis. In the synovium, the data support CD200/CD200R1 agonism as a means to reinforce resolution and suggests tissue biomarkers such as CD200⁺ DKK3⁺ niches and CD200R1-responsive immune signatures for patient stratification and response monitoring. In the skin, the findings link matrix stiffness to fibroblast reprogramming and point to therapies that interrupt mechanotransduction or downstream programs within mechanosensitive fibroblast lineages. Methodologically, by combining integrative multiomics, predictive modeling, multi-stage computational analysis, and rigorous validation, the work moves beyond descriptive single-cell atlases to testable and clinically relevant hypotheses, outlining tissue-anchored biomarkers and targeted interventions.
Hashem Mohammadian (Thu,) studied this question.