Probing the communication patterns of different chondrocyte subtypes in osteoarthritis at the single cell level using pattern recognition and manifold learning

Abstract

AbstractThe patterns of communication among different chondrocyte subtypes in human cartilage degeneration and regeneration help us understand the microenvironment of osteoarthritis and optimize cell-targeted therapies. Here, a single-cell transcriptome dataset of chondrocytes is used to explore the synergistic and communicative patterns of different chondrocyte subtypes. We collected 1600 chondrocytes from 10 patients with osteoarthritis and analyzed the active communication patterns for the first time based on network analysis and pattern recognition at the single-cell level. Manifold learning and quantitative contrasts were performed to analyze conserved and specific communication pathways. We found that ProCs (Proliferative chondrocytes), ECs (Effector chondrocytes), preHTCs (Prehypertrophic chondrocytes), HTCs (Hypertrophic chondrocytes), and FCs (Fibrocartilage chondrocytes) are more active in incoming and outgoing signaling patterns, which is consistent with studies on their close functional cooperation. Among them, preHTCs play multiple roles in chondrocyte communication, and ProCs and preHTCs have many overlapping pathways. These two subtypes are the most active among all chondrocyte subtypes. Interestingly, ECs and FCs are a pair of “mutually exclusive” subtypes, of which ECs are predominant in incoming patterns and FCs in outgoing patterns. The active signaling pathways of ECs and FCs largely do not overlap. COLLAGEN and LAMININ are the main pivotal pathways, which means they are very important in the repair and expansion of joint homeostasis. Notably, only preHTCs assume multiple roles (including sender, receiver, mediator, and influencer) and are involved in multiple communication pathways. We have examined their communication patterns from the perspective of cellular interactions, revealed the relationships among different chondrocyte subtypes, and, in particular, identified a number of active subtypes and pathways that are important for targeted therapy in the osteoarthritic microenvironment. Our findings provide a new research paradigm and new insights into understanding chondrocyte activity patterns in the osteoarthritic microenvironment.