Several in vitro investigations have shown the positive effects of copper on cells that regulate bone metabolism. Li and Yu [22] showed that copper ions can suppress the resorption of osteoclasts. Several investigators have confirmed that copper exerts positive effects in a dose-dependent manner. Low concentrations (0.1% w/w) of copper increased the viability and growth of osteoblasts, while higher concentrations (2.5% and 1% w/w) were shown to be toxic to cells [23]. In addition, the existence of copper stimulates the differentiation of MSCs toward the osteogenic lineage [24]. The copper content of femoral head necrosis tissues reported in previous studies varies. Milachowski observed a decrease in copper levels in his study of the relationship between idiopathic femoral head ischemic necrosis and trace element metabolism [25]. Yamazaki's study showed the exact opposite results, with an increase in copper levels in subchondral bone and cartilage in ischemic femoral head necrosis [26]. Our analysis may be related to the imbalance between the two investigators' subjects at baseline. A study by González-Reimers et al. [27] confirmed through experiments in rats that steroids increased muscle copper, iron and zinc levels as well as bone copper levels. In conclusion, we hypothesize that disorders of copper metabolism may be one of the mechanisms underlying the pathogenesis of SONFH. To test our hypothesis, we searched for SONFH-related and copper metabolism-related genes in public databases and identified PNP and SLC2A1 as DE-SONFHGs that are associated with copper metabolism through a series of analyses, such as GO and KEGG analyses. We also discovered a significant and the strongest negative correlation between PNP expression and plasmacytoid dendritic cell infiltration. The strongest negative correlations were found between PNP expression and plasmacytoid dendritic cell infiltration, between SLC2A1 expression and T follicular helper cell infiltration, between PNP expression and CD56 bright natural killer cell infiltration, and between SLC2A1 expression and T helper 17 cell infiltration. It was experimentally verified that both PNP and SLC2A1 were significantly downregulated in the peripheral blood of SONFH patients, which was consistent with the results of our bioinformatics analysis.
Purine nucleoside phosphorylase (PNP) is a vital enzyme in purine metabolism. A missense SNP (rs1049564) in the PNP gene was found to be associated with high IFN levels in SLE. The rs1049564 T allele of PNP is a loss-of-function variant that triggers blockade of the S phase and activation of the IFN pathway in lymphocytes [28]. Few studies have been published on the association of the PNP gene with SONFH. SLC2A1 is a gene that encodes a glucose transporter protein (GLUT1) that controls glucose uptake and is encoded on 1p34.2 [29]. This gene is essential for glucose metabolism, is involved in normal and tumor cell glycolysis, and can play a key role in the cell growth and proliferation of many tumor cells [30–32]. Some scholars found that only HEMGN and SLC2A1 were downregulated in SONFH tissues by bioinformatics, indicating that HEMGN and SLC2A1 may become diagnostic biomarkers of SOFNH; however, the exact mechanism has not been elucidated [33].
The PNP gene was mainly enriched in ‘positive regulation of smoothened signaling pathway’, ‘positive regulation of cytokine production involved in immune response’, ‘glycosphingolipid biosynthesis lacto and neolacto series’, and ‘B-cell receptor signaling pathway’. SSLC2A1 was mainly enriched in ‘cgmp metabolic process’, ‘vacuolar acidification’, and ‘glycosphingolipid biosynthesis lacto and neolacto series’. The main pathways clearly associated with SONFH in this study included ‘neutrophil extracellular trap formation’, ‘PI3k-Akt signaling pathway’, ‘mTOR signaling pathway’, ‘TNF signaling pathway’, and ‘HIF-1 signaling pathway’. Drug-induced glucocorticoid administration is considered to be a risk factor for femoral head necrosis. It has been proposed that glucocorticoid-induced platelet activation leads to disruptions in the local blood flow in the femoral head. Activated platelets can trigger neutrophil extracellular trap (NET) formation, leading to the ischemic necrosis of bone cells [34]. AKT/mTOR signaling pathway components are upregulated in a glucocorticoid-induced ONFH model, and human umbilical cord MSCs reduce macrophage polarization by inhibiting the AKT/mTOR signaling pathway and thus ameliorate necrosis and osteoclast apoptosis in a GC-induced model of ONFH [35]. In contrast to normal tissues, the expression of TNF-α in ONFH bone tissues was notably upregulated, and autophagy, apoptosis and the p38 MAPK/NF-κB signaling pathway were significantly activated, suggesting a significant effect of the TNF signaling pathway in the pathogenesis of femoral head necrosis [36]. Some scholars have also confirmed the involvement of the HIF-1 signaling pathway in the pathogenesis of hormonal osteonecrosis. Animal experiments have confirmed that bone health supplements can expedite the formation of new bone, promote the resorption of damaged bone, inhibit the inflammatory response and ultimately ameliorate SONFH through the HIF-1α/BNIP3 pathway [37].
Many studies have confirmed that copper plays a vital role in the function of the mammalian immune system; for example, animals with copper deficiency are more vulnerable to infections, whereas animals that consume an oral diet that is rich in copper are more resistant [38]; oral copper supplementation helps to maintain T-cell function in rats with acute spinal cord injury [39]. From the perspective of the immune system, the body needs balanced copper intake; moderate copper intake is sufficient for optimal immune function, while too much may be harmful to the organism [40]. Our study also found that the pathogenesis of SONFH is associated with multiple immune regulatory pathways. Therefore, we hypothesize that in the pathogenesis of SONFH, an imbalance in copper metabolism may cause disordered immune regulation in the body and thus induce femoral head necrosis. Recently, studies have shown that immune cell infiltration is associated with the progression of SONFH. One study showed that patients with SONFH had significantly more activated B cells in peripheral blood than healthy controls and that the extent of femoral head collapse was positively linked to the proportion of CD86 + CD19 + B cells [41]. Ma et al. [42] revealed that suppressor T lymphocytes maintain homeostasis of femoral bone mass by secreting various cytokines, such as interleukin 4 and osteoprotegerin, which are tightly associated with the development of nontraumatic ONFH. Moreover, it has also been demonstrated that the collapse of the immune response plays a role in the pathogenesis of SONFH [43]. In the present study, we concluded that PNP expression had a notable and the strongest negative association with plasmacytoid dendritic cell infiltration, while PNP expression had a notable and the strongest positive association with CD56 bright natural killer cell infiltration. SLC2A1 had a notable and the strongest negative association with T follicular helper cell infiltration, and SLC2A1 had the strongest positive association with T helper 17 (Th17) cell infiltration. In previous studies, the four types of immune cells mentioned above were closely associated with arthritis pathogenesis and the regulation of osteoblast/osteoclast coupling. Enhanced recruitment and activation of plasmacytoid dendritic cells and to arthritic joints due to topical use of the TLR7 agonist imiquimod improved arthritis in a genetic mouse model [44]. One study confirmed that compared to peripheral blood lymphocytes (PBLs), the majority of NK cells in the synovial fluid of patients with OA were CD56brightCD16 (-) cells, and the presence of CD56 (+) brightCD16 (-) cells that expressed granzyme A was associated with elevated levels of proinflammatory cytokines in the synovial fluid of patients with OA [45]. T follicular helper cells (Tfhs) are currently a research hotspot in basic immunology. Tfhs may contribute to the bone destruction that is associated with osteoporosis [45]. The balance between Treg and Th17-cell activity directly affects osteoclastogenesis and osteoblast/osteoblast coupling regulation [46]. Th17 lymphocytes typically secrete proinflammatory cytokines, such as IL-17A, IL-17F, IL-21, IL-22 and TNF-α, which can evoke NF receptor activator generation. Th17 lymphocytes can activate specific bone resorption genes directly or indirectly through the NF-κB pathway [47], and IL-17A can also downregulate osteogenic genes that are relevant to bone formation, such as osteocalcin (OCN) and Runt-related transcription factor 2 (Runx-2), and even lead to focal osteoblast death [48–50]. In summary, immune cells may influence osteoblasts and osteoclasts in the pathogenesis of SONFH. Therefore, assessing the differences in the proportions of infiltrating immune cells in SONFH is valuable for elucidating molecular mechanisms underlying SONFH and verifying molecular markers that are associated with immune infiltration.
Among the drugs predicted to target the key genes in this study and thus to affect SLC2A1, genistein (genistein) may be one of the most promising drugs. Genistein is an isoflavone that is also known as genistein, and it is an estrogen-like compound that is widely found in legumes. It can prevent bone loss in human and rat models of osteoporosis directly by acting through estrogen receptors (ERs) on bone cells and indirectly by affecting thyroid follicular cell activity [51]. Phytoestrogens prevent methylprednisolone-induced femoral head necrosis and secondary osteoporosis in rats [52]. The specific mechanism by which genistein ameliorates SONFH requires further investigation.
In the present study, we found notable differences in the expression levels of three copper death-related genes (DLD, PDHB, and MTF1) that are associated with PNP and SLC2A1 in SONFH. Dihydrothioctanamide dehydrogenase (DLD), which is also referred to as the E3 subunit of pyruvate dehydrogenase complex (PDHC) EC 1.6.4.3, is the third catalytic enzyme of PDHC, and it is a multifunctional mitochondrial matrix enzyme [53]. It has been demonstrated that DLD gene silencing prevents lipid peroxidation and iron-related death in vitro and in vivo [54]. Pyruvate dehydrogenase B (PDHB) encodes pyruvate dehydrogenase, which is a constituent enzyme of the pyruvate dehydrogenase multienzyme complex in mitochondria [55]. Recessive PDHB mutations cause pyruvate dehydrogenase complex (PDC) deficiency, which mainly affects the nervous system, such as developmental delays, seizures, and peripheral neuropathy [55]. Recent studies have found that DLD and PDHB positively regulate copper-related death [21], and we hypothesize that in the pathogenesis of SONFH, high expression of DLD and PDHB may disrupt pyruvate metabolism and thus induce copper-related death in osteoblasts. Further experimental studies are needed to elucidate the related mechanisms. Metal-regulated transcription factor 1 (MTF1) is a highly conserved zinc (Zn)-binding transcription factor in eukaryotes that responds to both metal overload and metal deficiency to protect cells from oxidative and hypoxic stress. A comparable Cu + binding center was confirmed to be present in mammalian MTF1, suggesting that it may also respond to Cu [56]; however, in vitro experiments also confirmed that copper ions enhance MTF1 expression in myogenic cells [57]. The application of hormones can increase copper contents in muscle, bone and even femurs [26, 27]. Recent studies have confirmed that MTF1 is a negative regulator of copper-related death [21]. We hypothesize that in SONFH, the increased copper content in femoral head tissues may further enhance the expression of MTF1 in bone tissues, and high expression of MTF1 negatively regulates osteoblast copper-related death, which in turn is involved in femoral head necrotic tissue repair.