The ability of a purified human glycoprotein (HGP.92) to exert anti-tumor activity was investigated in a mouse model using long-term readout assays. In vitro, in the presence of inflammatory mouse macrophages incubated with HGP.92, the growth of the mouse Lewis-lung-tumor cells (3LL) was decreased. This effect was concentration-dependent and required direct contact between tumor targets and HGP.92-treated macrophages. In addition, if the macrophage mono-layer was depleted of HGP.92 before addition of the target cells, no more cytostatic effect was observed. This anti-tumor activity of HGP.92-treated mouse macrophage was partially abrogated by addition of catalase in the culture medium, but not by superoxide dismutase or scavengers of the hydroxyl radical and singlet oxygen. Moreover, this tumor-cell growth reduction was not dependent on nitric oxide. In vivo, multiple i.v. injections of HGP.92 (5 times, 3 days apart) during the first week and a half exerted significant anti-tumor activity, as assessed by the reduction of both the number and the size of the lung nodules 3 weeks after i.v. inoculation of 3LL cells. Int. J. Cancer, 70:0-0, 1997. r 1997 Wiley-Liss, Inc. We have shown that a human glycoprotein of 92 kDa (HGP.92), isolated and purified from human urine, protected normal and SCID mice against an otherwise lethal inoculum of Listeria monocytogenes given intravenously; in addition, inflammatory murine macrophages incubated in vitro with HGP.92 exert anti-tumoral activity, as assessed by the growth reduction of 3LL cells (Fontan et al., 1994). Sequencing of 5 tryptic peptides of HGP.92 revealed complete homology with the amino-acid sequence of the Tamm Horsfall glycoprotein (THP), a urinary oligomeric glycopro-tein consisting of 80-to 100-kDa sub-units. Treatment of THP with urea yields a monomeric molecule of 92 kDa, suggesting that HGP.92 could be a sub-unit of THP. THP is known to be produced in the kidney, but its physiological function remains unclear (Sikri et al., 1981; Kumar et al., 1985). Some authors have described immunosuppressive activity in vitro (Brown et al., 1986; Hession et al., 1987), but some have reported immunostimulatory effects on human lymphocytes (Hunt and McGiven, 1978; Yu et al., 1993) and polymorphonuclear neutrophils (Horton et al., 1990; Yu et al., 1992). In our experimental models, native oligomeric THP did not enhance the resistance of mice infected with L. monocytogenes, and mouse macrophages incubated in vitro with THP were not cytostatic against 3LL cells. The increased resistance of HGP.92-treated SCID mice to L. monocytogenes infection, as well as the anti-tumoral activity of HGP.92-treated macrophages, provide some evidence that macro-phages may be one of the targets of HGP.92 biological activity. Many studies indicate that mouse macrophages can be activated to a tumoricidal state in vitro and in vivo by various cytokines and by non-cytokine molecules such as lipopolysaccharide (LPS) acting either alone or in synergism. We therefore monitored the ability of HGP.92 to...