Specific inhibitors of hyaluronan (HA) biosynthesis can be valuable therapeutic agents to prevent cancer invasion and metastasis. We have found previously that 4-methylumbelliferone (MU) inhibits HA synthesis in human skin fibroblasts and in group C Streptococcus. In this paper, the inhibition mechanism in mammalian cells was investigated using rat 3Y1 fibroblasts stably expressing HA synthase (HAS) 2. Exposure of the transfectants to the inhibitor resulted in significant reduction of HA biosynthesis and matrix formation. The evaluation of HAS transcripts and analysis of cell-free HA synthesis demonstrated the post-transcriptional suppression of HAS activity by MU. Most interesting, the post-transcriptional suppression of HAS activity was also observed using p-nitrophenol, a well known substrate for UDP-glucuronyltransferases (UGT). We investigated whether the inhibition was exerted by the glucuronidation of MU using both high pressure liquid chromatography and TLC analyses. The production of MU-glucuronic acid (GlcUA) was consistent with the inhibition of HA synthesis in HAS transfectants. MU-GlcUA was also detected at a similar level in control cells, suggesting that the glucuronidation was mediated by an endogenous UGT. Elevated levels of UGT significantly enhanced the inhibitory effects of MU. In contrast, the inhibition by MU was diminished to the control level when an excess of UDP-GlcUA was added to the cell-free HA synthesis system. We propose a novel mechanism for the MU-mediated inhibition of HA synthesis involving the glucuronidation of MU by endogenous UGT resulting in a depletion of UDP-GlcUA.

4-Methylumbelliferone (MU) inhibits the cell surface hyaluronan (HA) formation, and that such inhibition results in suppression of adhesion and locomotion of cultured melanoma cells. Here, we examine the effect of MU on melanoma cell metastasis in vivo. MU-treated melanoma cells showed both decreased cell surface HA formation and suppression of liver metastasis after injection into the mice. Oral administration of MU to mice decreased tissue HA content. These HA knockdown mice displayed suppressed liver metastasis. Thus, both cell surface HA of melanoma cells and recipient liver HA can promote liver metastasis, indicating that MU has potential as an anti-metastatic agent.

No abstract

The reconstruction of glycosaminoglycan chains using the transglycosylation reaction of testicular hyaluronidase was investigated. First, the optimal conditions for the transglycosylation reaction catalyzed by the enzyme were determined by incubation with the enzyme, using hyaluronic acid (M(r) = 800,000) as a donor and pyridylaminated hyaluronic acid hexasaccharide having glucuronic acid at the nonreducing terminal as an acceptor. The carbohydrate chains as reaction products were determined by high performance liquid chromatography and mass spectrometry. The optimal pH for hydrolysis by the enzyme was found to be about 5.0, whereas that for the transglycosylation reaction was about 7.0. Sodium chloride in the reaction medium inhibited the transglycosylation reaction. Under the optimal conditions, the carbohydrate chains were sequentially transferred along with disaccharide units to the nonreducing terminal of the acceptor and elongated up to docosasaccharide from the acceptor, pyridylaminated hexasaccharide. Using a combination of hyaluronic acid, chondroitin, and chondroitin 4- and 6-sulfate as an acceptor and a donor, it was possible to reconstruct hybrid chains, which were natural or unnatural types of glycosaminoglycan chains. Therefore, it is highly likely that application of the transglycosylation reaction using testicular hyaluronidase would facilitate artificial reconstruction of glycosaminoglycans having some physiological functions.

Various oligosaccharides from hyaluronic acid, which were fluorescence-labeled and blocked by pyridylamination at the reducing terminal, were incubated as substrates or acceptors with bovine testicular hyaluronidase. Fluorescence-labeled reaction products in the reaction mixture were monitored selectively and directly by ion-spray mass spectrometry without chemical derivatization. As a result, several features of the relationship between oligosaccharides, substrates, and testicular hyaluronidase were clarified. When hexasaccharides or larger oligosaccharides having D-glucuronic acid at the nonreducing terminal were used as substrates, they were hydrolyzed sequentially to disaccharides from the nonreducing terminal, and these disaccharides were then transferred to other hexasaccharides. On the other hand, when heptasaccharides or larger oligosaccharides having N-acetyl-D-glucosamine at the nonreducing terminal were used as substrates, trisaccharides were released from the nonreducing terminal, and then also transferred to other hexasaccharides, thus forming nonasaccharides. Thus, the relationship between hydrolysis and transglycosylation reactions with testicular hyaluronidase was characterized using ion-spray mass spectrometry.