Our study aimed to evaluate the genetic correlation and causal relationship between OA and human blood metabolites. Firstly, by using GWAS summary data of OA and 529 human blood metabolites, we conducted an LDSC analysis. We found 19 human blood metabolites had genetic correlations with OA. Secondly, we analyzed the causal relationship between the 19 human blood metabolites and OA by MR analysis. We found 1,6-anhydroglucose had a causal relationship with OA. Thirdly, through using the GEO2R tool, we screened FMO3 (associated with methylcysteine) as the DEG and the LIPC (associated with 1-palmitoylglycerophosphoinositol) as the DMG associated with OA.
LDSC identified 19 candidate human blood metabolites genetically correlated with OA, such as 1,6-anhydroglucose, L-Carnitine, Citrate, Proline, methyl cysteine, 1-palmitoylglycerophosphoinositol, and Gamma-glutamyl-leucine. MR analysis identified a significant causal relationship between OA and 1,6-anhydroglucose. 1,6-anhydroglucose is a dehydrated form of glucose and it is the raw material for the synthesis of glucosamine. Glucosamine is used for the biosynthesis of glycoproteins and glycosaminoglycans. Glucosamine is a natural compound present in the majority of human tissues, with the highest concentrations detected in cartilage . A study showed that OA is associated with the degradation of articular cartilage and the death of chondrocytes . This study also found glucosamine could promote the proliferation of chondrocytes via the Wnt/β-catenin signaling pathway . Combined with our results, we thought the pathogenesis of OA is related to the decrease of glucosamine in chondrocytes and cartilage. Another study showed the combination of glucosamine and chondroitin sulfate might have some efficacy in patients with moderate-to-severe symptoms of OA . To sum up, the pathogenesis of OA could lead to the decrease of 1,6-anhydroglucose and the specific pathogenesis mechanisms need to be further studied.
L-Carnitine (LC) is the biologically active form of carnitine, synthesized from the essential amino acids lysine and methionine . The carnitine pool is mainly found in skeletal muscle . LC is a potent antioxidant molecule with an anti-inflammatory effect shown in different experimental models . Some studies have found that LC has a protective effect through anti-inflammatory and antioxidant mechanisms [25, 26]. The nonsteroidal anti-inflammatory drugs (NSAIDs) are frequently chosen as fist-line therapy for OA . In addition, LC can relieve pain at central and peripheral levels. A previous study showed LC plus diclofenac sodium is more effective than diclofenac sodium alone in improving the pathogenesis and symptoms of OA . In summary, our study found a genetic correlation between OA and LC, which provides new insight into the role of LC in the pathogenesis of OA.
Citrate, as a part of the citric acid cycle, is the most important metabolic pathway for energy supply . A recent study on immune cells suggested that citrate can regulate the production of pro-inflammatory cytokines by modulating the transactivation of inflammatory genes . Furthermore, citrate can affect the solubility of calcium crystals, and the deposition of calcium crystals in human articular cartilage is a common phenomenon of OA [31, 32]. These crystals can stimulate articular chondrocytes to secrete matrix metalloproteinase (MMP) proteins such as MMP-13, which are thought to be a significant contributor to the degenerative process during OA pathogenesis [33, 34]. These studies indicating the calcium crystals have a direct pathogenic effect on OA. However, the relationship between OA and citrate needs further research.
In the results of our study, proline also had a genetic correlation with OA. Proline is a structurally and functionally unique imino acid among 20 natural proteinogenic amino acids. Therefore, proline is a marker of important regions of protein structures and biological functions. A previous study showed the proteins with a high content of proline and proline-containing peptides are some of the most important proteins in the processes of metabolism, cell cycle, and intracellular signal transduction . Tonachini L et al. identified chondrocyte proteins with poly-proline regions, which may be relevant to the late stage of chondrocyte differentiation . The occurrence and development of OA could attribute to the catabolism and the state balance of cartilage and chondrocytes .
By using the GEO2R tool, a DEG was found which codes flavin-containing monooxygenase 3 (FMO3). FMO3 can impair multiple aspects of cholesterol homeostasis . Choi W-S et al. found cholesterol plays a significant role in the pathogenesis of OA, and the level of cholesterol increased in OA chondrocytes . The study also found CH25H or CYP7B1 regulated by cholesterol overexpressed in mouse joint tissues can cause experimental OA, whereas knockout or knockdown of these hydroxylases abrogated the pathogenesis of OA . Combined with our results, these findings suggest FMO3 may play roles in the pathogenesis of OA. A DMG encodes Hepatic lipase (HL) was also found by using the GEO2R tool. Previous studies had demonstrated that HL plays significant biological roles in cholesterol transporting and the metabolism, composition, and level of several lipoproteins . The effect of cholesterol on the pathogenesis of OA has been discussed in this study. So, there is a potential genetic correlation between LIPC and OA through regulating cholesterol expression. In addition, HL also plays a significant role in the pathogenesis of OA through enhancing circulating monocyte chemotactic protein 1 (MCP1) levels and activation of stress-induced SAPK/JNK- and p38-MAPK pathways . However, due to the limited studies, the relationship between LIPC and OA needs further research.
By, using the latest GWAS summary data of OA and human blood metabolites, we observed multiple genetic correlations between OA and human blood metabolites through LDSC and MR analysis [14, 15]. The large sample size of GWAS data ensures the accuracy of our research results, which is one strength of our study. Then MR results identified a significant causal relationship between OA and 1,6-anhydroglucose, making up for the lack of observational research. At last, we verified whether human blood metabolites related genes were DEG and DMG.
This study also has some limitations. Given the complexity of the pathogenesis of OA, we are unable to determine the specific role of these identified metabolites in the pathogenesis of OA. More mechanism-based experiments are needed to further confirm the biological rationality and clarify the biological mechanism of the identified metabolites, which expect to participate in the development of OA.