Domain structure of human plasma fibronectin. Differences and similarities between human and hamster fibronectins.

The domain structure of human plasma fibronectin was investigated by limited proteolysis with trypsin and thermolysin with special emphasis on the comparison with hamster plasma fibronectin. Both human and hamster plasma fibronectins gave closely spaced doublet bands on NaDodS04-polyacrylamide gels after reduction, but the spacing between these two bands was significantly smaller in human plasma fibronectin. Thermolysin digestion of human plasma fibronectin produced a similar set of fragments as hamster plasma fibronectin, le. Mr = 155,000 (155K), 145,000 (145K), 44,000 (44K), 24,000 (24K), and 22,000 (22K) fragments. The binding properties of these fragments are also identical with those from hamster plasma fibronectin. However, tryptic digestion of human plasma fibronectin gave a slightly different fragmentation profile than that of hamster plasma fibronectin. Human plasma fibronectin was cleaved into 215K, 185K, 32K, and 37K fragments, of which the first three fragments were also generated from hamster plasma fibronectin but the 37K fragment was not. Time course analysis of tryptic digestion of human plasma fibronectin suggested that the 37K fragment was derived from large (a) subunit together with the 32K and the 185K fragments whereas small (/3) subunit was cleaved into the 215K and the 32K fragments. The human plasma fibronectin-specific 37K fragment was shown to be composed of the “Fib- 2” domain which is represented by the 22K thermolysin fragment, because (i) the 37K tryptic fragment was degraded to the 22K fragment by thermolysin, (ii) the binding properties to fibrin, heparin, and gelatin of the 37K fragment were identical with those of the 22K thermolysin fragment, and (iii) affinity-purified anti- 37K antibodies specifically reacted with the 22K fragment as well as the 37K tryptic fragment. In addition, the anti-37K antibodies also reacted with the tryptic 215K fragment but not with the tryptic 185K fragment, suggesting that only the 215K fragment contains the Fib-2 domain of the /3 subunit, while the 185K fragment was lacking the Fib-2 domain of the a subunit which was identified as the 37K fragment. In support of this view, only the 215K fragment could bind to fibrin. Furthermore, the antibodies also reacted with the 155K thermolysin fragment but not with the 145K. This in- dicates that the 37K tryptic fragment contains not only the Fib-2 domain but also the COOH-terminal 10K region of the 155K thermolysin fragment which is not present in the 145K fragment. In conclusion, the Fib-2 domain of human plasma fibronectin is, unlike hamster plasma fibronectin, equally present in both a and p subunits, but only the domain present in the a subunit is released as the 37K fragment upon tryptic digestion, possibly because of the different susceptibility of the hinge region between the “cell/Hep-2” (which corresponds to the 155K-145K thermolysin fragments) and Fib-2 domains.