Kupffer cells play important roles in the modulation of immune response, phagocytosis, and senescent cell removal 12. Hydrolytic enzymes and reactive species produce the killing effects of Kupffer cells and some degree of adjacent tissue damage 13. Liver macrophages are constantly exposed to antigens from portal circulation, to which development of full inflammatory response is useless and potentially harmful 4. Neither tissue damage nor inflammation follows senescent cell removal by Kupffer cells, due to the physiological control of inflammation events during antigen processing 25. Apolipoproteins can modulate macrophage function 6. Among them, beta₂-glycoprotein I (beta₂GPI) decreases Kupffer cells respiratory burst while increases efficiency of C. albicans killing 7. Beta₂GPI also binds phosphatidylserine (PS) residues on the surface of senescent cells, targeting them to clearance 8. In order to get an insight on the role of beta₂GPI in the silent antigen removal by Kupffer cells, perfused mouse liver was used as a model of Kupffer cell-dependent phagocytosis and related respiratory burst activity, and results were correlated with those obtained in isolated mouse non-parenchymal cells.

All reagents used were obtained from Sigma (St. Louis, MO), except for beta₂GPI that was purified from a pool of human sera 7. Livers from female CF-1 mice (20á28 g body weight) fed ad libitum were perfused with Krebs-Henseleit bicarbonate buffer pH 7.4, saturated with 95% O₂/5% CO₂, at 10 mL/min and 37 degrees C, without recirculation 9. After 15 min equilibration, O₂ uptake was measured in the effluent perfusate as it flowed past a Clark-type O₂ electrode. Total sinusoidal lactate dehydrogenase (LDH) efflux (in U/g liver) and the respective fractional LDH release (in % of the activity in the tissue) were assessed in the 30á45 min interval as described 9. Colloidal carbon (C) (0.25 mg/mL; Rotring, Germany) was infused during the 30á45 min interval, either in the absence or presence of 1á30 micrograms beta₂GPI/mL, added at 20 min, and rates of C uptake were calculated according to Cowper et al. 10. Carbon-induced O₂ consumption (in micromolar O₂/g liver/min) was calculated by subtracting the basal O₂ uptake, during the 30á45 min C perfusion interval 9. Liver samples taken after perfusion with 0.25 mg C/mL in the absence and in the presence of 30 micrograms beta₂GPI were fixed in Dubosq Brazil, embedded in Paraplast, and stained with hematoxylin-eosin. Non-parenchymal liver cell preparation was obtained by liver perfusion with collagenase 7, with viability values higher than 95%. The respiratory burst was evaluated by a luminol-dependent assay 11 after zymosan stimulation (200 particles/cell) 7, and results were expressed as relative total light emission or light emission rate.

Liver perfusion with C in the absence of beta₂GPI led to uptake of C particles and increase in O₂ consumption (Fig. 1, Table 1). The latter effect is mainly related to the respiratory burst of Kupffer cells 49, with secondary O₂ utilization in mitochondrial respiration for energy supply needed for C phagocytosis (10) and O₂ uptake induced in hepatocytes by eicosanoids released from activated Kupffer cells 12. Both Kupffer cell C uptake (Fig. 1A) and C-induced O₂ consumption (Fig. 1B) are inhibited by beta₂GPI (1á30 micrograms/mL), with significant (p < 0.05) 23% and 97% decreases being found at 30 micrograms/L beta₂GPI, respectively. C-induced O₂ consumption inhibition inversely correlates with beta₂GPI concentration (r: -0.8455; p= 0.036). In agreement with biochemical data, optical microscopy revealed that C uptake by non-parenchymal cells is diminished by infusion of 30 micrograms/mL beta₂GPI (Fig. 2). These effects by beta₂GPI are achieved without changes in liver viability, evidenced by comparable fractional LDH effluxes among experimental groups (not shown), and are not mimicked by albumin infusion (Table I). Despite the beta₂GPI-induced inhibition of C phagocytosis found in perfused liver, chemiluminescence of isolated non-parenchymal liver cells was insensitive to 30 micrograms beta₂GPI/mL (Fig. 3).

Beta₂GPI associates with membranes through annexins, PS receptor, lipoprotein receptors, and negatively charged phospholipids such as PS 513. Interference of beta₂GPI with PS availability in the phagocyte membranes may affect cellular responses, such as translocation of protein kinase C (PKC) to cell membranes 14. This effect could affect PKC and subsequent triggering of PKC-dependent events, including superoxide anion generation and particle uptake 15. From current data, it is suggested that beta₂GPI suppresses the respiratory burst response associated with Kupffer cell phagocytosis, while discretely diminishes particle uptake. The former effect is dependent on intact liver architecture, which allows interactions among different cell-types in the liver 16.