Abstract This study performed an in-depth investigation into the myeloid cellular landscape in the synovium of patients with rheumatoid arthritis (RA), “individuals at risk” of RA, and healthy controls (HC). Flow cytometric analysis demonstrated the presence of a CD40-expressing CD206^+CD163^+ macrophage population dominating the inflamed RA synovium, associated with disease activity and treatment response. In-depth RNA sequencing and metabolic analysis demonstrated that this macrophage population is transcriptionally distinct, displaying unique inflammatory and tissue-resident gene signatures, has a stable bioenergetic profile, and regulates stromal cell responses. Single-cell RNA sequencing profiling of 67,908 RA and HC synovial tissue cells identified nine transcriptionally distinct macrophage clusters. IL-1B^+CCL20^+ and SPP1^+MT2A^+ are the principal macrophage clusters in RA synovium, displaying heightened CD40 gene expression, capable of shaping stromal cell responses, and are importantly enriched before disease onset. Combined, these findings identify the presence of an early pathogenic myeloid signature that shapes the RA joint microenvironment and represents a unique opportunity for early diagnosis and therapeutic intervention. __________________________________________________________________ In-depth analysis of myeloid cells residing within the RA joint reveals a unique early signature associated with disease activity. INTRODUCTION Rheumatoid arthritis (RA) is a progressive autoimmune disease characterized by synovial inflammation, hyperplasia, and structural damage to cartilage and bone ([60]1). It affects 1% of the population, reduces mobility and quality of life, and is associated with substantial comorbidities including atherosclerosis, diabetes, cardiovascular disease, and obesity ([61]2–[62]4). A significant proportion of patient’s are nonresponders to current therapeutic targets, and it is currently impossible to predict who will develop severe, erosive disease and who will respond to treatment. Therefore, better understanding of the disease at the site of inflammation will allow the development of new treatment strategies or predictive biomarkers. Macrophages are pivotal players in joint destruction, with increased numbers of sub-lining macrophages being a hallmark of disease activity and response to treatment in RA ([63]5–[64]9). Recent research has revealed a diversity of macrophage phenotypes in health and disease, in both function and origin ([65]10). This has further emphasized the need to explore and characterize the phenotype and ontogeny of macrophage populations in RA. Immunohistology analysis of RA synovial tissue suggests that macrophages residing in the synovial lining layer differ from those in the sub-lining layer ([66]11). Mature resident macrophages are abundant in the intimal lining layer, displaying an interleukin-10 (IL-10) phenotype ([67]12). In contrast, the synovial sub-lining has a more heterogeneous phenotype, displaying a mixture of both inflammatory M1-like and resolving M2-like markers, possibly due to active infiltration of monocyte-derived macrophages from the periphery ([68]12, [69]13). A conceptual revolution has occurred in recent years, challenging our understanding of the origin of macrophages, demonstrating that many macrophages are tissue-derived during embryonic development ([70]14–[71]18). Many tissue-resident macrophages arise from embryonic precursors before birth independent of hematopoiesis and subsequently self-sustain their numbers throughout adulthood ([72]19–[73]21). Evidence suggests that the relative contribution of infiltrating versus resident macrophages in establishing a pool of mature macrophages differs from one tissue to another ([74]14, [75]15, [76]22–[77]24). Moreover, the prevailing dogma that activated macrophages exist as one of two phenotypical states, M1 or M2, while useful, fails to reflect the remarkable plasticity and diversity of these cells in human disease. Macrophages in vivo are subject to a plethora of stimuli, capable of changing their phenotype in response to environmental cues, such that the real-time phenotype most likely does not fit this rigid binary nomenclature ([78]25, [79]26). Evidence now suggests that this system should be extended to encompass a wide spectrum of macrophage activation states, the characteristics and ontogeny of which remain largely unknown ([80]27, [81]28). A recent study has reported in a mouse model of arthritis that CX[3]CR1^+ macrophages in the synovial lining layer form an unusual protective barrier-like layer to shield the joint from inflammation associated with arthritis ([82]29). Another recent study described the presence of a specific subset of tissue macrophages capable of re-establishing joint homeostasis in RA ([83]30). This study described differential enrichment of two general subpopulations of macrophages, MerTK^+ and MerTK^−, with risk of flare determined by the ratio of these synovial tissue macrophages ([84]30). This highlights the plasticity of macrophages in adapting to the needs of their microenvironment. Local macrophage differentiation is determined by tissue-specific cues with tissue imprinting a dominant factor in shaping the phenotype and function of macrophages ([85]31–[86]34). Undeniably, the role of macrophages in the pathogenesis of RA is widely recognized, yet this has been largely based on in vitro monocyte-derived macrophage analysis, synovial histology, or animal studies ([87]35–[88]37). Recent advances in transcriptomic analysis have improved our understanding of macrophage biology in RA synovium with some studies investigating human synovial tissue macrophages using methods such as single-cell RNA sequencing (scRNA-seq) and mass cytometry ([89]30, [90]38). Recently, it has been established that circulating autoantibodies can precede clinical onset of symptoms in patients with RA and so studying those “at risk” of developing disease may provide important clues in understanding RA disease pathogenesis ([91]39). However, the relative contribution of synovial tissue macrophages in the evolution of RA, especially before disease, has yet to be fully elucidated. Therefore, in this study, we used multiparameter flow cytometry, bulk RNA-seq and scRNA-seq, and noninvasive fluorescent lifetime imaging microscopy (FLIM) metabolic imaging and functional analysis to fully explore the spectrum of macrophage activation states residing within the synovium of patients with RA, individuals at risk of RA (IAR), and healthy controls and determine their role in driving RA pathogenesis. RESULTS CD206^+CD163^+ macrophages are the dominant macrophage subset in RA synovial tissue Within the inflamed microenvironment of the joint in vivo, macrophages are exposed to a plethora of stimuli; therefore, we examined both M1/M2-like and non-M1/M2 macrophage phenotypes within RA synovial tissue. Patient demographics are outlined in [92]Table 1. RA synovial tissue single-cell suspensions and synovial fluid mononuclear cells (SFMCs) were stained using a panel of human macrophage polarization–associated cell surface markers to assess the phenotype and frequency of macrophages ([93]Fig. 1, A to C, and fig. S1A). The frequency and median fluorescence intensity (MFI) of pan macrophage markers CD68 and CD64 were significantly increased in RA synovial tissue compared to SFMC, indicating that a greater number of macrophages reside in synovial tissue in comparison to the synovial fluid (fig. S1B, P < 0.05). Analysis of specific macrophage markers identified a spectrum of activation states rather than the classic paradigm of M1 and M2 ([94]Fig. 1). On the basis of this spectrum, we identified a dominant population of synovial tissue macrophages in the RA inflamed joint expressing high levels of CD206 and CD163, markers typical of an M2-like phenotype ([95]Fig. 1A and fig. S1C). This double-positive population also expressed high levels of the activation marker CD40 ([96]Fig. 1, A and B, and fig. S1C). Table 1. Patient demographics for RA synovial tissue. Clinical information of patients with RA for synovial tissue samples included in this study (n = 32). Rheumatoid factor (RF), anti-citrullinated protein antibodies (ACPA), C-reactive protein (CRP), disease activity score-28 (DAS28), disease duration, and mediation at the time of arthroscopy are shown. Values reflect either the group average or the range in each group with the given criteria. Characteristics Mean Percentage or range Age (years) 54.8 24–80 Gender, female 10 31% RF positive 16 50% ACPA positive 19 59.3% Disease duration (years) 9.2 <1–41 DAS28 3.9 1.68–7.74 CRP (mg/liter) 25.5 <1–156.7 On MTX 12 37.50% Treatment naïve 10 31% [97]Open in a new tab Fig. 1. Synovial tissue macrophage phenotypic characterization. [98]Fig. 1. [99]Open in a new tab (A) Representative flow cytometric dot plots of four independent RA synovial tissue samples demonstrating the frequency of CD206^+CD163^+ macrophages and coexpression of CD40. (B) Dot plots indicating percentage frequency of CD206^+CD163^+, CD206^−CD163^−, and CD206^+CD163^+CD40^+ macrophages and median fluorescence intensity (MFI) of CD40 on CD206^+CD163^+ macrophages in RA synovial tissue (n = 9) in comparison to synovial fluid (n = 9). Data are presented as mean ± SEM with each symbol representing a different sample. Statistical analysis using Mann-Whitney U test, *P < 0.05, **P < 0.01, ***P < 0.005, significantly different from synovial tissue. (C) Visual representation of multidimensional flow cytometric data. SPICE analysis was performed for the identification of distinct macrophage subsets in an average of RA synovial tissue (n = 9) and fluid (n = 6). Each pie segment indicates the different combinations of marker expression as denoted by the legend below. The surrounding pie arcs indicate the specific macrophage markers produced by each pie segment. High-quality RNA was isolated from sorted CD206^+CD163^+ and CD206^−CD163^− synovial tissue macrophages and bulk RNA-seq was performed. (D) PCA was performed on the total dataset of RA synovial tissue sorted CD206^+CD163^+ and CD206^−CD163^− macrophage subsets (n = 9). (E) Hierarchical clustered heatmap displaying DEGs involved in adhesion/cell growth, cytoskeletal rearrangement, cytokine/chemokines, macrophage markers/phagocytosis, and metabolism in RA synovial tissue CD206^+CD163^+ macrophages compared to CD206^−CD163^− macrophages (n = 9). (F) Representative multiphoton microscopy FLIM analysis of flow sorted RA CD206^+CD163^+ and CD206^−CD163^− synovial tissue macrophages. Representative FLIM images whereby a red/green cell is predominantly using OXPHOS, while a blue cell that indicates glycolysis is being used as the main energy source. (G) Summary of macrophage emission lifetime (τ[avg]) (n = 7) and optical redox ratio (ORR) (n = 3). Data expressed as mean ± SEM using Wilcoxon signed rank or paired t test, *P < 0.05, ***P < 0.005 significantly different from each other. A marked increased expression of CD206^+CD163^+ macrophages residing in RA synovial tissue compared to fluid was observed as indicated by representative flow plots and accompanying frequencyquantification ([100]Fig. 1B and fig. S1D). In contrast, the double-negative CD206^−CD163^− subset of macrophages is significantly enriched in synovial fluid in comparison to tissue ([101]Fig. 1B). Next, we used the Simplified Presentation of Incredibly Complex Evaluations (SPICE) algorithm. SPICE analysis facilitated further visual confirmation of site-specific macrophage phenotypes ([102]Fig. 1C), where we demonstrate, as indicated by the deep red pie segment, that the CD206^+CD163^+CD40^+ macrophage subset is markedly greater in RA synovial tissue compared to fluid ([103]Fig. 1C). The specific enrichment of CD206^+CD163^+CD40^+ macrophages in RA synovial tissue compared to fluid is further highlighted and confirmed by representative flow plots, histograms, and relative proportion bar graph quantification ([104]Fig. 1B and fig. S1, D and E). CD206^+CD163^+ macrophages are transcriptionally distinct with unique metabolic preferences