Infiltration of Malignant B Cells Suppresses and Re-Organizes Lymph Node Fibroblasts in Chronic Lymphocytic Leukemia and Mantle Cell Lymphoma

Secondary lymphoid organs (SLOs) are organized by specialized stromal cells, including B-cell-interacting follicular dendritic cells (FDCs) and T-cell zone-enriched fibroblastic reticular cells (FRCs). These fibroblasts are crucial for T/B cell compartment formation and immune regulation. Inflammation, especially Type-2 inflammation, triggers B cells to produce lymphotoxin (LTαβ), which directs SLO fibroblasts to produce chemokines necessary for follicle development. Additionally, fibroblasts, including cancer-associated fibroblasts (CAFs), influence tumor growth by modulating tumor cell invasion. Altered FRC/FDC distributions have been described in mantle cell lymphoma (MCL), diffuse large B-cell lymphoma (DLBCL), and chronic lymphocytic leukemia (CLL). CLL and DLBCL cells produce LTαβ, which boosts fibroblast CXCL13 production. Blocking CXCR5 or LTαβ receptors restores FDCs and delays CLL progression in TCL-1 mice. However, the mechanisms by which healthy fibroblasts transform into CAF-like cells and their role in B-cell leukemia and lymphoma are not well understood.

Here, we focused on CLL and MCL, which differ in their fundamental pathogenic mechanisms but share similarities in epidemiological characteristics, cells of origin, molecular alterations, and clinical features. We first performed in vitro co-culture assays using healthy human lymph node fibroblasts (HLFs) and PBMCs from 4 CLL patients, 4 MCL patients, and 2 healthy individuals. In this 2-D culture system, CLL and MCL PBMCs adhered to HLFs and formed larger aggregates compared to healthy controls. The survival of CLL and MCL B cells was enhanced by HLF. However, unlike healthy PBMCs, CLL and MCL cells induced fibroblast stretching, pulling, and formation of gaps to allow leukemic B-cell infiltration. CLL and MCL infiltration also increased HLF size and elevated FAP, BAFF, and CXCL13 protein levels by 2-5-fold; the ratio of FDC to FRC however remained the same.

In SLOs, CLL cells receive stimulatory signaling that facilitate the homing and tissue-retention of tumor cells. Mimicking this, we stimulated CLL PBMCs with anti-IgM and IL-4, and found elevated LTαβ levels (2.2-fold, p=0.06) in CLL B cells. While resting CLL cells altered HLF subsets by blocking PDPN (MFI: 68,500 vs. 58,723, p=0.008), anti-IgM/IL-4 stimulated CLL cells further reduced PDPN+ FRCs (58.9% vs. 40.5%, p=0.008). Activated CLL cells caused FRCs to expand (2.8-fold, p=0.03), produce CXCL13 (5.46-fold, p=0.01) and display a CAF-like phenotype with upregulated FAP, αSMA, and PD-L1. Altogether, data here suggest BCR/IL-4 stimulated CLL cells reorganize fibroblasts into CAF-like cells.

To verify these results in vivo, we established xenografts by mixing unmanipulated CLL or MCL patient PBMCs with HLFs in Matrigel and subcutaneously injecting this mixture into a flank site of NSG mice. Mice receiving CLL-HLF Matrigel have CLL cell engraftment in spleen and lymph nodes (LNs). In contrast, without HLF help, mice received CLL-Matrigel alone showed CLL growth only in the spleen. When CLL-Matrigel was injected alongside either PBS or HLFs intraperitoneally, the HLF group had increased numbers of LNs and a higher number of CLL PBMCs in the spleen compared to the control PBS group. Similar phenomena were observed in MCL xenografts, suggesting the critical role of HLFs in facilitating the migration of leukemic B cells from the flank injection site to the spleen. Importantly, HLFs in both CLL and MCL xenografts showed a significantly decreased percentage of FRCs and an increased ratio of FDC to FRC, consistent with in vitro observations.

To determine how CLL cells infiltration impact HLFs over the course of disease progression, we then injected PBS or TCL-1 CLL cells into C57/B6 mice and studied LN fibroblasts weekly after tumor cell injection. We found a dramatic drop in CD45-PDPN+CD31- FRCs in C57/B6 mice 1 week after TCL-1 CLL cell injection (d0 to d7: 76% to 10.8%). As the disease progressed, the percentage of FRCs further decreased, and nearly all PDPN+ FRCs were lost by week 4 when mice succumbed to CLL.

Overall, we provide evidence of the dynamic interaction between malignant B cells and LN fibroblasts that transform fibroblasts into MCL- or CLL- specific CAFs. Further work exploring the mechanisms controlling stromal adaptations that lead to an immune-suppressive TME is required and may offer innovative therapeutic targets in B-cell leukemias and lymphomas.