Granulomas increase the local concentration and activity of immune cells and, therefore, the likelihood of a self-directed immune response. As such, existing treatments are targeted at inhibiting the formation of granulomas. Granulomas consist of multiple immune cells, including T cells, B cells, antigen-presenting cells, and stromal cells such as fibroblasts (9). The spatial architecture and cellular composition of granulomas are highly reminiscent of tertiary lymphoid organs that form in peripheral tissues in response to chronic antigen stimulation such as upon infection and during cancer progression.

In this issue of the JCI, Sati and colleagues used single-cell RNA-seq (scRNA-seq) and spatial transcriptomics to determine that CXCR4 was highly and preferentially expressed in T cells, B cells, and type 1 innate lymphoid cells (ILCs) in sarcoidosis lesional skin relative to nonlesional skin (10). CXCL12, the cognate ligand for CXCR4, was highly expressed in fibroblasts and has previously been shown to be expressed in tertiary lymphoid structures found in the peripheral tissues of Sjögren’s syndrome and mucosa-associated lymphoid tissue in patients with lymphoma (11) and upon viral infection of mice (12). By employing spatial transcriptomics, Sati and authors found that CXCL12 expression was localized in the granulomas of sarcoidosis-affected tissues. Taken together, their results show that CXCL12-expressing fibroblasts in the granulomas may recruit CXCR4-expressing immune cells, resulting in the formation of granulomas. Administration of an approved CXCR4 inhibitor, plerixafor, inhibited the formation of pulmonary granulomas in mouse models, suggesting that such therapeutic intervention may be possible in patients with sarcoidosis (Figure 1) (10).

Diagnostic features of granulomas and peripheral blood indicate sarcoidosis and a target for intervention. Granulomas from sarcoidosis skin lesions possess higher expression of CXCR4 in T cells, B cells, and type 1 innate lymphoid cells (ILC1s) compared with granulomas from other skin diseases. Increased quantities of ILCs in peripheral blood of patients with sarcoidosis also indicate the potential of ILCs as a biomarker for sarcoidosis. Notably, plerixafor inhibits the CXCL12/CXCR4 axis and the formation of lung granulomas in a preclinical mouse model.

The mechanisms of protection against granuloma formation by CXCR4 inhibition remains to be further explored. The authors argued that CXCR4 inhibition leads to reduced ILC1 accumulation, and in turn, reduced granuloma formation. However, given that CXCR4 can induce chemotaxis in T cells and B cells (both express higher levels of CXCR4) (13, 14), it is plausible that these immune cell subsets have a greater contribution to sarcoidosis. Indeed, the authors show a drastic decrease in granuloma formation in T cell– and B cell–deficient Rag2-knockout (Rag2-KO) mice compared with wild-type mice, and a more modest decrease between ILC-deficient (Il2rg-KO) and Rag2-KO mice. As such, it may be argued that granuloma formation, and CXCR4 inhibition, is largely driven by T cells and B cells, and less so by ILCs. The authors also identified lymphotoxin A+ and lymphotoxin B+ T cell subsets, which may home to CXCL12-rich regions and exacerbate the expansion of granulomas by inducing further expression of chemokines (15, 16). Addressing these questions are technically challenging due to a lack of models that can selectively deplete ILCs.