Genetic landscape and immune mechanism of monocytes associated with the progression of acute-on-chronic liver failure

scRNA-seq and major cell typing of PBMCs from healthy controls and ACLF patients

We used scRNA-seq to analyze the PBMCs in 3 healthy controls and 6 ACLF patients (including 3 patients in ACLF survival group and 3 in ACLF death group) (Table 1). After strict quality-based screening, 83,577 high-quality sorted cells were obtained, and principal component analysis (PCA) was performed on 2,000 genes with the highest expression variability in these cells (Fig. 1A). Nineteen cell clusters were identified by unsupervised analysis (Fig. 1B). Differential expression analysis was performed among these clusters, and cell subpopulations were defined according to the gene expression patterns of published cell clusters. Consistent with the findings in healthy controls, ACLF patients were also defined as having cell subpopulations corresponding to the six leukocyte types based on differentially expressed genes (DEGs) (Fig. 1C). These cell subpopulations included T cells (clusters 0, 1, 2, 3, and 11) [15], natural killer (NK) cells (clusters 4, 8, 12, and 13) [16], monocytes (clusters 5, 7, 10, 14, and 15) [17, 18], plasmacytoid dendritic cells (cluster 16) [19, 20], B cells (clusters 6 and 9) [21], and plasma cells (cluster 18) [22], and other cellular components (Fig. 1D). To investigate the relationships of monocytes with other clusters, we analyzed pDCs, NK cells, plasma cells, and B cells. In addition, FCGR3A (CD16) was also highly expressed in NK cells. Therefore, combining these marker genes can more accurately identify monocytes. Subsequently, we mainly performed an in-depth analysis of different types of monocytes, with an attempt to explore the differences between ACLF patients and healthy controls in more detail.

Fig. 1figure 1

Unsupervised cluster analysis on single-cell RNA-seq data. A Work flow of PBMCs isolation and single-cell RNA-Seq. B Cell subpopulations detected and their UMAP plots. C Proportional bar charts of cell subpopulations in ACLF patients and healthy controls (colors are coded for the cell subpopulations identified in this study). D Violin plots showing the expression levels of marker genes for defined cell subpopulations revealed by single-cell sequencing. The distribution of cell subpopulation-specific marker genes across all clusters reflects their cell types. E Table shows the marker genes of cell types defined in D

Thus, by detecting the gene expression profiles of a large number of single-sorted peripheral blood mononuclear cells (PBMCs), we identified many featured genes of the mononuclear cell population and found that multiple genes were co-expressed in other cellular components of monocytes and other PBMC subsets, suggesting that there may be functional overlap among them.

Differences in distribution and genotypes among different human monocyte subpopulations

To further determine the heterogeneity and variation of monocytes at the single-cell level and provide markers for more precise subpopulation classification, we performed scRNA-seq on 4239 circulating monocytes from healthy controls based on the similarities and differences in gene expression profiles. To differentiate the function and activation status of each subpopulation based on differences in gene expression markers, we classified the monocytes into five subpopulations (Fig. 2A). According to the classification method of human monocytes [including classical (CD14+CD16−), non-classical (CD14dimCD16+)] [23], we identified classical monocytes with high CD14 expression and non-classical monocytes with high CD16 expression (Fig. 2B). Notably, uniform manifold approximation and projection (UMAP) also showed a monocyte subpopulation expressing both CD14 and CD16 marker genes, which belong to a small subset of CD14 monocytes. Based on the high expression of CD14 and intermediate expression of FCGR3A (CD16), these cells were labeled as intermediate monocytes. Their location on the UMAP may indicate that they are experiencing transition from classical to non-classical monocytes. We further characterized each cluster using the top 10 marker genes (Fig. 2C and D).

Fig. 2figure 2

Genotypes and functions of monocyte subpopulations in healthy subjects. A UMAP shows transcriptional heterogeneity in circulating monocytes. The 4239 monocytes are further divided into 5 subpopulations, whose names are annotated on the left, and different colors are used to distinguish each cluster. B Expression patterns of two marker genes CD14 and FCGR3A (CD16) in human monocytes. Purple represents the expression of these two marker genes and each dot represents an individual cell. C Dot plot curve showing the top 10 marker genes revealed by single-cell sequencing of monocyte subpopulations defined in B. D Table shows the top 10 marker genes of monocyte subpopulations defined in C

The monocytes subpopulations classified in this way exhibited distinct transcriptional and functional characteristics. However, whether all cell subpopulations of these types and other cell subpopulations in the same inflammatory microenvironment have similar functions or whether further sub-classifications exist has not been determined. It was found that pro-inflammatory mediators S100A8 and S100A9 were also highly expressed in a subpopulation with high CD14 expression, so this subpopulation was classified as pro-inflammatory monocytes. In addition, the pro-inflammatory monocytes have unique marker genes including chemokines (CCL3 and CCL4) [9], interleukins (IL-1B), and NLRP3 inflammasome and are closely related with immune regulation, promotion of inflammatory responses, and induction of apoptosis. The non-classical monocytes (CD14dimCD16++) are a subpopulation with a high expression of FCGR3A (CD16). In addition to the expression of FCGR3A, we also found that the transcription factor TCF7L2, the inflammation-inducible gene HES4, and the interferon-inducible genes IFITM2, IFITM3, MS4A7, MTSS1, and SMIM25 played key roles in the signaling pathways, suggesting that this subpopulation is enriched in genes associated with signal transcription, inflammation induction, and induction of interferon-stimulated antiviral responses. Another subpopulation was HLA monocytes, which expressed the major histocompatibility complex (MHC) class II molecules HLA-DRB1 and HLA-DPB1. While other subpopulations also expressed human leukocyte antigen (HLA)-related genes, the HLA monocytes had the most diverse and highest levels of leukocyte-related gene expressions, including high levels of the unique leukocyte antigens HLA-DMA and HLA-DMB, suggesting that this subpopulation has the strongest antigen processing and presentation capabilities [24]. The gene expression of megakaryocyte-like monocytes was similar to that of megakaryocyte progenitors (including PPBP and PF4), suggesting that megakaryocyte-like monocytes play an important role in activating platelets and mediating chronic inflammation. NK cell-like monocytes expressed genes (e.g., GZMA and GNLY, in addition to KLRB1 and NKG7) similar to those expressed by NK cells, indicating that this subpopulation is involved in cytotoxicity regulation and inflammatory responses and has functional overlap with NK cells.

Thus, through gene expression profiling of PBMCs, we identified five monocyte subpopulations with unique gene expression signatures, which indicated that each monocyte subpopulation has specific functions.

ACLF monocytes had unique immune characteristics and inflammatory responses

Monocytes are associated with ACLF progression, but their exact functions, subpopulations, cellular and molecular characteristics, and activation status in ACLF remain unclear. To identify the heterogeneity and variations in cellular components, we performed scRNA-seq on 13,071 high-quality circulating monocytes from ACLF patients and compared their gene expression patterns with monocytes from healthy controls. In accord with findings in the healthy population, the monocytes from ACLF patients were still clustered into five subpopulations based on the expressions of DEGs (Fig. 3A and B). Abundant evidence suggests that monocytes play an important role in the early stages of disease by mediating both pro- and anti-inflammatory responses. In fact, monocytes are well known for their dual roles in coordinating inflammatory responses and regulating tissue repair [25, 26]. In the present study, Based on the differences in gene expressions among monocyte subpopulations between the healthy controls and the ACLF patients, the most significant difference between up-regulated and down-regulated genes among all monocyte subpopulations was found in pro-inflammatory monocytes, followed by CD16 monocytes (non-classical monocytes) (Fig. 3C). Subsequently, we further subdivided the ACLF group into the ACLF survival group (n = 3) and ACLF death group (n = 3). The most significant change in gene expression among all monocyte subpopulations was also observed in the pro-inflammatory monocytes (Fig. 3D). Therefore, our analysis focused on pro-inflammatory monocytes and CD16 (FCGR3A) monocytes.

Fig. 3figure 3

Unique immunological profile of monocytes in ACLF. RNA sequencing was performed on the monocytes of 3 healthy controls and 6 ACLF patients (including 3 patients in the ACLF death group and 3 in the ACLF survival group). A The 4239 ACLF monocytes on UMAP are further divided into 5 subpopulations, whose names are annotated on the left, and different colors are used to distinguish each cluster. B Dot plot curve showing the top 10 marker genes revealed by single-cell sequencing of ACLF monocyte subpopulations. C The number of up- and down-regulated genes per monocyte subpopulation when comparing gene expression changes between ACLF patients (n = 6) and healthy controls (n = 3). The number of genes with log-fold change (Fc) > 0.5 and adjusted P value < 0.05 in each subpopulation were as follows: 104 genes up-regulated and 101 down-regulated in NK-like Monocytes subpopulation; 124 genes up-regulated and 101 down-regulated in pro-like monocytes subpopulation; 109 genes up-regulated and 83 down-regulated in CD16 monocytes subpopulation; 113 genes up-regulated and 95 down-regulated in HLA monocytes subpopulation; 89 genes up-regulated and 101 down-regulated in megakaryocyte-like monocytes subpopulation. D The number of up- and down-regulated genes per monocyte subpopulation when comparing gene expression changes between ACLF death group (n = 3) and ACLF survival group (n = 3). The number of genes with log-fold change (Fc) > 0.5 and adjusted P value < 0.05 in each subpopulation were as follows: 94 genes up-regulated and 80 down-regulated in NK-like Monocytes subpopulation; 139 genes up-regulated and 158 down-regulated in Pro-like Monocytes subpopulation; 125 genes up-regulated and 109 down-regulated in CD16 Monocytes subpopulation; 54 genes up-regulated and 40 down-regulated in HLA Monocytes subpopulation; and 108 genes up-regulated and 97 down-regulated in megakaryocyte-like monocytes subpopulation. The volcano plots show all the up-regulated (red) and down-regulated (green) genes in monocytes when: E. comparing pro-inflammatory monocytes between healthy controls and ACLF patients; F comparing CD16 monocytes between healthy controls and ACLF patients; G comparing Pro-inflammatory Monocytes between ACLF death group and ACLF survival group; H comparing monocytes between ACLF survival group and ACLF death group. The top 10 biological processes of the differentially expressed genes were identified

Pro-inflammatory monocytes presented with most significant changes in gene expressions, including up- and down-regulated genes, suggesting that pro-inflammatory monocytes may be actively involved in the development and progression of ACLF. Compared to healthy controls, genes that were more than twice as highly expressed in the ACLF patients included HBB gene, a member of the hemoglobin family, and thrombospondin 1 (THBS-1), a member of the thrombospondin family (THBS). Other up-regulated genes included SAMSN1, MALAT1, PRS20, and MT-CO3, which are associated with activation of immune cells and promotion of inflammatory responses [27,28,29]. Genes down-regulated in the pro-inflammatory monocytes of ACLF patients included FOS, CCL3 and LGALS2, which are mainly associated with resistance to viral infection, promotion of apoptosis, and immunosuppression (Fig. 3E). Compared with the ACLF survival group, the ACLF death group had higher levels of pro-inflammatory cytokines and their receptors (e.g., IL-6R), and cytotoxic factors (e.g., NKG7) (Fig. 3F), further suggesting that increased inflammation may be an influencing factor for ACLF progression and poor prognosis.

The subpopulation with the second largest number of altered genes was CD16 monocytes (non-classical monocytes). Compared with those in the healthy controls, genes up-regulated in the ACLF group included scavenger receptors (CD163 and MRC1), growth factors (e.g., HGF), and inhibitory cytokines (e.g., IL10), suggesting their involvement in anti-inflammatory responses. The down-regulated genes included NLRP3, CCL3, and S100A4, showing the inflammatory effects of CD16 monocytes (Fig. 3G). In addition, the CD16 monocytes in the ACLF death group had similar biological functions to those in the ACLF survival group (Fig. 3H). Therefore, rather than simply promoting or inhibiting inflammation, CD16 monocytes have dual regulatory effects in apoptotic signaling pathways, inflammatory responses, and cytokine chemotaxis and migration in ACLF.

Therefore, pro-inflammatory monocytes promote inflammatory and immune responses, while CD16 monocytes may play a dual role of promoting and inhibiting inflammation in the occurrence and development of ACLF.

ACLF progression-related biomarker THBS1 and its pathways

Since the pro-inflammatory monocytes showed most significant changes in gene expression in ACLF, the phenotypic changes in this subpopulation might be associated with the poor prognosis of ACLF. Accordingly, our further analysis focused on pro-inflammatory monocytes. Although clinical measures, such as creatinine (CREA) and total bilirubin (TBil), can be used to monitor disease activity and treatment response, these assessments provide little information on the immune status or underlying prognostic mechanisms of ACLF. To determine the relevant indicators of ACLF progression, we detected the expressions of thrombospondin 1 (THBS-1) in the monocytes of ACLF patients at the single-cell transcriptional level and found that the expression of THBS1 was more obvious in the pro-inflammatory monocytes of ACLF patients (Fig. 4A). Moreover, the THBS1 expression gradually increased in the order of healthy controls, ACLF survival group, to ACLF death group. Therefore, THBS1 may be a factor influencing ACLF progression and poor prognosis (Fig. 4B).

Fig. 4figure 4

A Dot plot showing THBS1 expression in monocyte subpopulations in ACLF patients and healthy controls. B Dot plot showing THBSI expression in monocyte subpopulations in healthy controls, ACLF survival group, and ACLF death group. C Dot plot showing the expressions of SMAD2, SMAD3, TGF-β1, and NK-κB1 in monocyte subpopulations in ACLF patients and healthy controls. D Dot plot showing the expressions of IL-6, TNF-α, and IL-1β in ACLF patients and healthy controls. E THBS1 pathways in monocytes (the pathways and possible mechanisms by which THBS1 activates related transcription factors and produces inflammatory factors). F Flow cytometry shows the mean fluorescence intensity (MFI) of THBS1 in one healthy control, one ACLF survival patient, and one ACLF death patient. G Bars show the MFI of THBS1 in healthy controls (n = 3), ACLF survival group (n = 3), and ACLF death group (n = 3). IL-6 interleukin-6, TNF-α tumor necrosis factor alpha, IL-1β Interleukin-1β

THBS1 belongs to the thrombospondin (THBS) family and has biological functions, such as promoting immune activation and enhancing inflammatory response [30]. We analyzed the expressions of potential molecules associated with the THBS1 pathway using scRNA-seq. The results showed that pro-inflammatory monocytes with high expression of THBS1 gene in ACLF also had high expressions of TGF-β, NF-κB, SMAD2 and SMAD3; in addition, the gene expressions of IL-6, IL-1β, and tumor necrosis factor-α (TNF-α) also increased (Fig. 4C and D). Previous studies have suggested that THBS1 stimulates SMAD2 and SMDA3 phosphorylation through the NF-κB signaling pathway and transforming growth factor β1 (TGF-β1) pathway and ultimately regulates the secretion of inflammatory cytokines (e.g., IL-6, IL-1β, and TNF-α) [27, 31]. Therefore, we speculate that the highly expressed THBS1 in pro-inflammatory monocytes can bind to the conserved sequence (leucine–serine–lysine–leucine [LSKL]) to activate the TGF-β and NF-κB pathways and induce the upregulation of downstream transcription factors (e.g., SMAD2 and SMAD3), which drive the secretion of IL-6, IL-1β, and TNF-α from multiple effector cells, thus leading to ACLF progression and even death (Fig. 4E).

Flow cytometry further showed that the MFI of THBS1 in monocytes was significantly higher in ACLF survival patients (962.16 ± 74.69) than in healthy controls (485.16 ± 67.32), and significantly higher in the ACLF death group (1129.18 ± 74.43) than that in the ACLF survival group (Fig. 4F and G). Therefore, high THBS1 expression may be a potential biomarker of ACLF progression and poor prognosis, and THBS1 may be used as an indicator for evaluating ACLF progression.

留言 (0)

沒有登入
gif