Specific biological responses of the synovial membrane to carbon nanotubes

Abstract

Biological evaluation of carbon nanotubes (CNTs) is typically performed in the lung or abdominal cavity; however, biological reactions to CNTs are predicted to be markedly different in other tissues. In applications of CNTs as reinforcement for artificial joints and drug delivery systems, including their use in bone regeneration, the intra-articular synovial membrane makes contact with the CNTs. Herein, we analyzed the reaction of the synovial membrane with multiwalled CNTs (MWCNTs). Injection of MWCNTs into rat knee joints revealed their dose-dependent incorporation into deep synovial membranes and the formation of granulation tissue, without long-term inflammation. MWCNTs were incorporated into human fibroblast-like synoviocytes (HFLSs), with less cytotoxicity than that observed in macrophages (RAW264 cells). Moreover, MWCNTs inhibited the release of cytokines and chemokines from HFLSs. The reaction of the synovial membrane with MWCNTs differed from that observed in other tissues; thus, detailed biological evaluation at each target site is necessary for clinical applications.

Figure 1. Analysis of the effects of a single intra-articular dose of MWCNTs using hematoxylin-eosin staining.

(a) Histological images of the control group captured at 1, 4, and 12 weeks after physiological saline injection. Superficial tissue and deep adipose cells of synovial tissue were normal at all time points, and no inflammatory response was observed. (b) Histological images of 0.3 mg carbon black (CB) injection. After 1 week, CB invaded the synovial surface and was incorporated by macrophages, and a mild inflammatory response was observed primarily in the lymphocytes. CB was not observed in adipose tissue. After 4 weeks, the inflammatory response was improved, although CB was still incorporated in macrophages. At 12 weeks, the inflammatory response was resolved. (c) Histological images of 0.003 mg MWCNTs into the knee joint. After 1 week, MWCNTs invaded the synovial surface, which became mildly thickened. MWCNTs were incorporated into macrophages, and a mild inflammatory response was observed primarily in lymphocytes. After 4 weeks, the inflammatory response was alleviated, although MWCNTs were still incorporated in macrophages. At 12 weeks, the surface was covered with normal synovial tissue. (d) Histological images of 0.03 mg MWCNTs into the knee joint. After 1 week, MWCNTs had invaded the deep synovial tissue, and inflammatory cells (macrophages and lymphocytes) had replaced a portion of the adipose tissue. MWCNTs were aggregated and incorporated into macrophages. After 4 weeks, the inflammatory area was reduced, and a multinucleated foreign body giant cell made of fused macrophages was observed. After 12 weeks, the inflammatory response was resolved, and granulation tissue was formed in the normal synovial cells. (e) Histological images of 0.3 mg MWCNTs. MWCNTs invaded a larger area than that of the 0.03 mg group; however, no severe inflammatory response was observed. The inflammatory response was reduced after 4 weeks and resolved after 12 weeks. Granulation tissue was formed in a wider area than that in the 0.03 mg group; however, the synovial surface was covered with normal cells.

Figure 2. Evaluation of equivalency between multiple and single administrations of MWCNTs by hematoxylin-eosin stain.

(a) Synovial histological images captured 1 and 4 weeks after administration of 150 μL physiological saline to rat knee joints three times with a 1-week interval (control group). Normal synovial tissue, similar to that in the single-dose group, was observed. (b) Synovial histological images captured 1 and 4 weeks after administration of 0.003 mg MWCNTs (0.001-mg doses three times with a 1-week interval) to the rat knee joint. After 1 week, MWCNTs had invaded the synovial surface and were incorporated by macrophages. A mild inflammatory response, which was similar to that at 1 week after a single 0.003 mg MWCNT dose, was observed. One week after the third administration of MWCNTs, the synovial surface was mildly thickened, similar to that observed after the single dose. The inflammatory response was resolved after 4 weeks, and results were similar to those after a single administration.

Figure 3. Incorporation of MWCNTs into cells.

Human fibroblast-like synoviocytes (HFLSs) and RAW264 cells were cultured and exposed to either 10 μg/mL MWCNTs or CB for 24 h. Both HFLSs and RAW264 cells exhibited incorporation of MWCNTs and CB, with RAW264 cells showing increased incorporation per cell. Both MWCNTs and CB were observed in lysosomes of HFLSs and RAW264 cells. Blue: nucleus, Red: lysosome.

Figure 4. Proliferation of HFLSs and RAW264 cells exposed to MWCNTs.

Cultured HFLSs and RAW264 cells were exposed to 0, 0.1, 1, 10, or 100 μg/mL of MWCNTs or CB. After 24 h of exposure, Alamar Blue staining was performed, and cell numbers were counted 4 h later. (a) In HFLS cells, CB did not result in inhibition of proliferation at any concentration. MWCNTs resulted in a dose-dependent decrease in cell proliferation at 10 and 100 μg/mL at rates of 86% and 64% that of the control, respectively. The difference was statistically significant. (b) For RAW264 cells, CB did not result in inhibition of proliferation at any concentration. When treated with 100 μg/mL MWCNTs, cell proliferation was inhibited at a rate of 18% that of the control. The difference was statistically significant.

Figure 5. Cytokine and chemokine secretion from HFLSs and RAW264 cells exposed to MWCNTs.

Cultured HFLSs and RAW264 cells were exposed to 0, 0.1, 1, or 10 μg/mL MWCNTs or CB. After 24 h of exposure, detection and measurement of cytokines and chemokines were performed. (a) HFLSs secreted IL-6, IL-8, and MCP-1. IL-6 secretion was reduced by both MWCNTs and CB exposure, but the dose-dependent effects were not clear. IL-8 was unchanged by MWCNTs and showed a decreasing tendency with CB, without dose dependency, and the difference was not significant. MCP-1 secretion was significantly reduced by 1 μg/mL MWCNTs and 1 or 10 μg/mL CB, but the effects were not dose dependent. (b) RAW264 cells secreted TNF-α, MCP-1, MIP-1α, and RANTES. TNF-α secretion was significantly increased by 1 or 10 μg/mL MWCNTs and 10 μg/mL CB, and MWCNTs resulted in a larger increase than did CB. MCP-1 secretion was increased by 10 μg/mL MWCNTs and CB, and MWCNTs resulted in a larger increase than did CB. MIP-1α secretion was significantly decreased by 10 μg/mL MWCNTs and was significantly decreased in response to CB in a dose-dependent manner. RANTES secretion was significantly increased by both 10 μg/mL MWCNTs and CB.