Late inflammatory monocytes define circulatory immune dysregulation observed in skin microbiome-stratified atopic dermatitis

Late inflammatory monocytes define circulatory immune dysregulation observed in skin microbiome-stratified atopic dermatitisTo the Editor,Atopic dermatitis (AD) is a skin inflammatory disorder well described for significant disease heterogeneity. (1) Previously, we defined steady-state microbial configurations dermotypes A and B that robustly reflected heterogeneity in AD clinical severity, cutaneous barrier properties and skin microbiome composition. (2) Here, we explored circulating immune dysregulation underlying dermotype-stratified AD.We performed single-cell RNA sequencing (scRNA-seq) of PBMCs from healthy subjects with dermotype A (n = 4) and AD patients with dermotypes A (n = 4) and B (n = 6). UMAP reduction distinguished 13 major immune clusters using transcriptional profile differences, identified using DISCO CELLiD (3) and top differentially expressed genes (DEGs; Supplementary Figure 1A ). Major clusters corroborated well across all groups as a testimony to limited sample distortion by inter-individual or batch effects (Supplementary Figure 1B ). Critically, AD patients harboured a marked enrichment in monocyte population cluster 1, whereby dermotype stratification saw a further enrichment in dermotype B (Supplementary Figure 1B ).To reveal finer expression details, we sub-clustered myeloid cells, obtaining 11 sub-populations whose identities were similarly assigned following DISCO CELLiD and top DEGs (Supplementary Figure 1C ). Clusters with low cell counts and not of myeloid origin were removed prior to down sampling for normalisation, yielding 7 myeloid clusters of interest. Majority of clusters represent CD14+ monocytes except for clusters 5 and 7 corresponding to CD16+ monocytes and conventional dendritic cells (cDC) respectively (Figure 1A ). CD14+ monocytes segregated into 4 transcriptionally distinct states (Figure 1A and B ): early activation monocytes expressing proinflammatory alarmins (S100A8 and S100A12; clusters 0 and 1), transitional monocytes concomitantly expressing alarmins and chemokines (S100A12 and CCL3; cluster 2), intermediate monocytes expressing antigen presentation genes (HLA-DPB1 andHLA-DPA1 ; cluster 3), and late inflammatory monocytes expressing proinflammatory cytokines (IL1B, CCL3 and TNF ; cluster 4,Supplementary Figure 1D ) (4, 5). Late inflammatory monocytes (cluster 4) were notably enriched in AD dermotype B (Figure 1A and C ) and exhibited upregulation in MHC class II (HLA-DQA2, HLA-DQB1 ), monocyte-related (ITGAM, FGD2 ) and proinflammatory genes (CCL4L2, CCL3L1 and CCL20 ), suggestive of inflammatory dysregulation (Figure 1D and Supplementary Table 1 ). Taken together, dermotype stratification addressed part of the cellular heterogeneity observed in circulating immune profiles, characterised particularly by late inflammatory monocytes enrichment in dermotype B.Complimenting scRNA-seq findings with CytoFLEX (Supplementary Figure 2A ), unstimulated PBMCs were stained and secretion of inflammatory cytokines CCL3, CCL4, TNF-α, IL-1β, and IL-8 was monitored (Supplementary Table 2 ). Cytokine secretion by classical (CD14+ CD16-) monocytes were significantly elevated in AD compared to healthy controls (Figures 2A, B and Supplementary Figure 2B ). Atopy-relevant cell types were examined, revealing similar levels between AD dermotypes A and B but elevated presence of Th2 and Th2a in AD dermotype B as opposed to healthy controls (Figures 2C and D ). The reverse was observed for NK and ILC3s, where AD subjects with dermotype B registered cell deficiencies (Figures 2E and F ). Comparing between AD dermotypes, IL3Cs were significantly diminished in dermotype B (Figure 2F ). Healthy and AD subjects harbouring dermotype A shared similar expression of aforementioned cells, suggesting that disease alone could not differentiate immune dysregulation observed. Next, we quantified 92 inflammatory biomarkers in plasma using Olink Proteomics Proximity Extension Assay (PEA) Target 96 Inflammation Panel. AD subjects recorded significantly higher expression of TNF-associated cytokines TRANCE and TRAIL, monocyte chemoattractant MCP-4, IL-18, and IL-2 than healthy controls (Figures 2G-J and Supplementary Figure 2C ). Noticeably, higher basal abundance of TRANCE and MCP-4 was observed in AD dermotype B than dermotype A (Figures 2G and I ), lending additional support to circulatory immune dysregulation previously observed in inflammatory monocytes.Our findings provide unique insights into immune heterogeneity underlying AD patients stratified by defined skin microbiome configurations. Definingly, dermotype B is characterised by microbial dysbiosis with reduced species richness, fewer commensal species and increased presence of S. aureus virulence genes. (2) Rather than atopy-relevant Type 2 cells, late inflammatory monocytes demonstrated a striking enrichment in AD dermotype B, particularly in the upregulation of proinflammatory cytokine transcripts. Coupled with skin microbial dysbiosis that increases cutaneous exposure to potent allergic inflammation inducers such as δ-toxin (2, 6), immune dysregulation directed by monocytes could predispose AD skin towards flare predisposition and clinical severity as observed in dermotype B. Basal secretion levels of proinflammatory cytokines by classical monocytes are elevated in AD regardless of dermotype, suggesting higher baseline inflammation driven by disease. While no stimulation was conducted here, earlier reports have demonstrated that LPS activation of monocytes upregulates transcriptional expression of proinflammatory cytokines IL-1β, CCL3 and IL-8. (4) Similarly, heightened proinflammatory cytokines and monocyte-related proteins were detected in plasma of AD dermotype B compared to dermotype A subjects, validating scRNA-seq findings of monocyte-driven immune dysregulation and inflammation.Collectively, circulatory immune profiling of dermotypes, defined by alterations in skin microbiota compositions, suggest a pathophysiological role for inflammatory monocytes. Future studies are needed to ascertain the functional and mechanistic roles of late inflammatory monocytes in dermotype-stratified AD. Importantly, our study highlights the potential clinical utility of dermotype stratification to indicate the cellular basis for increased inflammation and exacerbated immune dysregulation, setting precedence for better understanding of disease prognosis.