Peripheral immune landscape for hypercytokinemia in myasthenic crisis utilizing single-cell transcriptomics

Abstract

Background: Myasthenia gravis (MG) is the most prevalent autoimmune disorder affecting the neuromuscular junction. A rapid deterioration in respiratory muscle can lead to a myasthenic crisis (MC), which represents a life-threatening condition with high mortality in MG. Multiple CD4⁺ T subsets and hypercytokinemia have been identified in the peripheral pro-inflammatory milieu during the crisis. However, the pathogenesis is complicated due to the many types of cells involved, leaving the underlying mechanism largely unexplored.

Methods: We conducted single-cell transcriptomic and immune repertoire sequencing on 33,577 peripheral blood mononuclear cells (PBMCs) from two acetylcholine receptor antibody-positive (AChR +) MG patients during MC and again three months post-MC. We followed the Scanpy workflow for quality control, dimension reduction, and clustering of the single-cell data. Subsequently, we annotated high-resolution cell types utilizing transfer-learning models derived from publicly available single-cell immune datasets. RNA velocity calculations from unspliced and spliced mRNAs were applied to infer cellular state progression. We analyzed cell communication and MG-relevant cytokines and chemokines to identify potential inflammation initiators.

Results: We identified a unique subset of monocytes, termed monocytes 3 (FCGR3B⁺ monocytes), which exhibited significant differential expression of pro-inflammatory signaling pathways during and after the crisis. In line with the activated innate immune state indicated by MC, a high neutrophil-lymphocyte ratio (NLR) was confirmed in an additional 22 AChR + MC patients in subsequent hemogram analysis and was associated with MG-relevant clinical scores. Furthermore, oligoclonal expansions were identified in age-associated B cells exhibiting high autoimmune activity, and in CD4⁺ and CD8⁺ T cells demonstrating persistent T exhaustion.

Conclusions: In summary, our integrated analysis of single-cell transcriptomics and TCR/BCR sequencing has underscored the role of innate immune activation which is associated with hypercytokinemia in MC. The identification of a specific monocyte cluster that dominates the peripheral immune profile may provide some hints into the etiology and pathology of MC. However, future functional studies are required to explore causality.

Keywords: Innate immunity; Monocyte; Myasthenia gravis; Myasthenic crisis; Single-cell sequencing.

Fig. 1

Study design and single-cell landscape of cell types identified from myasthenic crisis (MC). a. The design of the study. This is a self-comparative study. The single-cell RNA sequencing (ScRNA-seq) was performed on peripheral blood mononuclear cells (PBMCs) from three (acetylcholine receptor antibody positive) AChR + myasthenic crisis (MC) patients. The sequencing was conducted at the most acute stage (on ventilation) and three months after the crisis. Hemogram results from 22 AChR + MC patients were also analyzed to obtain an overview of the whole blood. b. A total of 33577 PBMC cells are identified in ScRNA-seq, among which 18114 are from the acute stage of MC and 15463 from three months after MC. The dimensional reduction is performed with the uniform manifold approximation and projection (UMAP). c. The single-cell landscape at the acute stage of MC. d. The single-cell landscape after three months after the crisis. Treg regulatory T; Tcm central memory T; Tem effector memory T; Temra terminally differentiated effector memory T; DC dendritic cell; NK natural killer

Fig. 2

Hemogram results and overall differentially expressed genes (DEG) analysis of all PBMCs. a. The DEG analysis of all MC PBMCs (at MC vs. three months after MC). Myeloid cell pathways activation indicated an innate immune activation, while lymphoid cell pathways suppression indicated an adaptive immune inhibition. b. Enrichment analysis of MC up-regulated genes (n = 440) in different databases. The CellMarker enrichment indicates monocytes were the most activated cell type. c. Enrichment analysis of MC down-regulated genes (n = 903) in different databases. The CellMarker enrichment indicates lymphocytes were the most repressed cell types. d. the hemogram results of 22 AChR⁺ MC patients between the acute stage at MC and three months after MC (Bonferroni adjusted p-values). Extremely neutrophilia and lymphopenia comprise the drastically higher neutrophil–lymphocyte ratio (NLR) at MC. e. The correlation relationships between NLR and MG-related clinical scores (Bonferroni adjusted p-values). Three (ADL, MMT, QMG) of four scores show the positive relationships between NLR and clinical severity. QMG, Myasthenia Gravis Foundation of America quantitative myasthenia gravis score; MMT, MG-specific manual muscle testing; ADL, Myasthenia Gravis Activities of Daily Living Scale; QOL-15, Myasthenia Gravis quality of life questionnaire

Fig. 3

Single-cell analysis of monocytes from MC PBMCs. a. The distribution of all monocytes (at MC and after MC). b. The RNA velocity tendency in different monocytes at MC, where monocytes 1 were differentiated into monocytes 2, 3 and 4. c. The cell markers used to classify monocytes. d. The DEG analysis of all monocytes (MC vs. after MC). e. The DEG analysis of monocytes 3 compared to other monocytes (including monocytes from both at MC and after MC states). f. The DEG analysis of only monocytes 3 (MC vs. after MC)

Fig. 4

The characterization of monocytes 3 from MC PBMCs. a. The expression of relevant cytokines and chemokines across all cell types at MC. The eight chemokines were in fact found surging in plasma at the acute stage of MC in our previous follow-up study [15]. The nine chemokines are representative CC (induce migration of lymphocytes and monocytes) and CXC (promote neutrophil migration) chemokines. The monocytes 3 exhibited high expression of IL1B and CXCL8, which potentially might augment the MC inflammation. b. The receptors expression of (IL-1 and CXCL8) in different cells at MC. c. Expression of CXCL8, IL1B, and RNASE2 in different monocytes. d. The inflammasome signaling scores in all cell types. The monocyte 3 exhibited the highest activation in inflammasome signaling at MC. e. The interferon signaling scores in all cell types. The interferon signaling was highest activated in monocytes at MC. f. The cell–cell communications between monocytes 3 and other cell types at MC. g. The B activation (BAFF) signaling communications among all cell types at MC

Fig. 5

Single-cell analysis of B cells from MC patients. a. The distribution of all B cells (MC and after MC). b. The RNA velocity tendency in different B cells at MC. c. The DEG analysis of all B cells (MC vs. after MC), in which the B cell receptor signaling was inhibited. d. ScBCR clonal expansion in all B cells. Age-associated B, naive B, and plasma cells were clonally expanded at MC. e. VDJ combination of BCRs in all B cells

Fig. 6

Single-cell analysis of T cells from MC patients. a. The distribution of all T cells (MC and after MC). b. The RNA velocity tendency in different T cells at MC. c. The DEG analysis of all T cells (MC vs. after MC), in which T exhaustion signaling (CTL4 and PD-1) were mostly activated. d. T activation scores in different T cells. e. T exhaustion scores in different T cells. f. ScTCR clonal expansion in all T cells. Most T cells were clonally expanded at MC