Separation of sublethal and lethal effects of the bactericidal/permeability increasing protein on Escherichia coli.

Abstract

Binding of the bactericidal/permeability increasing protein (BPI) of granulocytes to Escherichia coli promptly produces several discrete outer envelope alterations and growth arrest without major impairment of bacterial structure or biosynthetic capabilities, raising the question whether these early effects of BPI are sufficient to cause bacterial death. In this study, the bactericidal action of BPI was examined more closely. We have found that bovine or human serum albumin blocks bacterial killing without preventing BPI binding or an increase in outer membrane permeability. Moreover, addition of serum albumin after BPI results in growth resumption without displacement of bound BPI and without (early) repair of the envelope alterations. These effects are opposite to those produced by Mg2+ (80 mM), which displaces greater than 85% of bound BPI and rapidly initiates outer envelope repair without restoration of bacterial growth. The extent of rescue by serum albumin depends on the time and pH of preincubation of BPI with E. coli: e.g., for E. coli J5 treated with human BPI, t1/2 = 79 min at pH 7.4 and 10 min at pH 6.0. The serum albumin effects on BPI action are the same in wild-type E. coli and in a mutant strain lacking an activatable phospholipase, indicating that serum albumin does not act by sequestering membrane-damaging products of bacterial phospholipid hydrolysis. The progression from reversible to irreversible growth arrest, revealed by the subsequent addition of serum albumin at different times, is paralleled by a decrease in amino acid uptake and an increase in the permeability of the cytoplasmic membrane to o-nitrophenyl-beta-D-galactoside. These findings demonstrate at least two stages in the action of BPI: (a) an early, reversible, sublethal stage in which BPI has effects on the outer envelope and causes growth arrest, and (b) time- and pH-dependent progression to a lethal stage, apparently involving cytoplasmic membrane damage, possibly caused by penetration of a small subpopulation of BPI.