Human herpesvirus (HHV)-8-negative, idiopathic multicentric Castleman disease (iMCD) is a rare hematologic illness involving progressive flares of inflammation, lymphoproliferation, and cytokine-induced organ failure. While there is an FDA-approved anti-interleukin-6 (IL-6) therapy, siltuximab, which is effective in 34-50% of iMCD patients, disease etiology and pathophysiology remains largely unknown. While mTOR inhibition with sirolimus is effective in a subset of patients, the cell types and immune signaling pathways dysregulated in iMCD are poorly understood. We previously demonstrated that circulating T cells from iMCD patients have elevated STAT1-dependent type I interferon (IFN-I) signaling. However, IFN-I signaling in lymph node (LN) tissue, the primary site of abnormal histopathology, is unknown.

To examine IFN-I signaling in lymph node cells, we performed 10X Genomics' FLEX single-cell RNA sequencing on dissociated formalin-fixed paraffin-embedded (FFPE) LN tissue from iMCD patients (n=4) and reactive controls (n=4). We defined major cell types using unbiased clustering followed by manual evaluation of top differentially expressed gene markers per cluster. IFN-I signaling (specifically IFN-alpha) was significantly enriched in iMCD LN tissue across most cell types and led to separation of sub-populations of B cells and CD4+ T cells, which were clustering distinctly based on an IFN-I signature. Interestingly, GSEA identified several cell types that had significant enrichment of both IFN-I and mTOR signaling, including M2 macrophages and T follicular helper cells.

Since regions of the lymph nodes have distinct functions, we asked whether increased IFN-I activation was confined to specific areas of the lymph node in iMCD. First, we performed regional transcriptomics on germinal centers within LN tissue from iMCD patients (n=4) and reactive controls (n=4) using GeoMx (Bruker) followed by gene set enrichment analysis and cell type deconvolution using the SpatialDecon package. We uncovered that naïve CD4+ and CD8+ T cell signatures were increased within germinal centers in iMCD. The IFN-I signaling pathway was also highly upregulated in iMCD germinal centers compared to controls. These data suggest that IFN-I signaling is increased in iMCD LN germinal centers, possibly in naïve T cells. To examine IFN-I dysregulation with greater resolution, we also performed spatial transcriptomic analysis using 10x Visium on LN tissue from iMCD patients (n=4) and reactive inflammatory controls (n=4), which allows for deconvolution of voxel-level data of approximately 50-100 cells each. To resolve the cellular composition within each voxel, we leveraged our LN single-cell RNA-seq FLEX dataset to inform Robust Cell Type Decomposition (RCTD), a Bayesian deconvolution method that estimates the proportion of each cell type present in spatial transcriptomic spots. We also manually identified whether voxels fell within germinal centers, mantle zones, or interfollicular space. Together these allowed us to determine the proportion of cell types in each region. We identified that within germinal center voxels, Interferon alpha was the second most highly upregulated gene pathway in iMCD compared to controls. Interestingly, IFN-alpha was not significantly different between iMCD and control tissue in non-germinal center regions (mantle zone, interfollicular space) of the lymph node. Together, our spatial transcriptomics data indicate that IFN-I is significantly upregulated specifically in germinal center regions of iMCD lymph nodes.These data demonstrate that IFN-I signaling is increased in the lymph nodes of iMCD patients across multiple cell types, likely in the germinal center. Since IFN-I can lead to mTOR activation through PI3K signaling, it is possible that elevated IFN-I contributes to the elevated mTOR signature observed in iMCD. Given that IFN-I is significantly elevated in both circulation and LNs of iMCD patients, targeting INF-alpha with Anifrolumab may be a potential therapy for treatment refractory patients.

This content is only available as a PDF.
Sign in via your Institution