Bactericidal/Permeability-Increasing Protein Inhibits Growth of a Strain of Acholeplasma laidlawii and L Forms of the Gram-Positive Bacteria Staphylococcus aureus and Streptococcus pyogenes

Abstract

Bactericidal/permeability-increasing protein (BPI) inhibited growth of cell wall-deficient Acholeplasma laidlawii and L forms of certain strains of Staphylococcus aureus and Streptococcus pyogenes. However, the same strains of S. aureus and S. pyogenes with intact cell walls were not susceptible to the growth-inhibitory effects of BPI.

Bactericidal/permeability-increasing protein (BPI), a 55-kDa basic protein found in azurophilic granules of polymorphonuclear leukocytes, has been reported to be cytotoxic for many gram-negative bacteria but not for gram-positive bacteria (1, 2). BPI binds to and permeabilizes the outer gram-negative bacterial lipopolysaccharide layer, initiating events that lead ultimately to cell death (9).

A recombinant N-terminal fragment (rBPI₂₃) consisting of the first 199 amino acids of human BPI has been shown to retain the bactericidal activity of neutrophil-derived BPI (3, 10). We have produced several other recombinant BPI proteins, including full-length BPI (rBPI) (4, 5), a dimer of N-terminal recombinant BPI (rBPI₄₂) (5), and rBPI₂₁, an analog of rBPI₂₃ consisting of the first 193 amino acids of BPI and lacking one of three native cysteines (at position 132, replaced with an alanine); unlike rBPI₂₃, rBPI₂₁ does not form a dimer (5).

In previous studies, BPI reportedly did not kill gram-positive bacteria (1) but was able to inhibit both oxygen consumption and energy-dependent amino acid transport by isolated membrane vesicles from both gram-positive and gram-negative bacteria (6). We hypothesized that the reported lack of sensitivity of intact gram-positive bacteria to BPI could be due to protection of the cytoplasmic membrane by the cell wall and that gram-positive bacteria growing without cell walls (i.e., as L forms [8]) might be susceptible to BPI. In this study, we examined the effects of BPI on two gram-positive bacteria growing with and without their cell walls and on Acholeplasma laidlawii, which naturally lacks a cell wall (7).

The rBPI₂₁ and rBPI proteins used for these experiments were purified as previously described (4, 5). rBPI₄₂ (dimer) was purified from a mixture of rBPI₂₃ monomer and dimer by size exclusion chromatography. Proteins were serially diluted in 5 mM sodium citrate (pH 5.0)–150 mM NaCl prior to use.

Escherichia coli J5 (9) was grown as previously described (4, 5). Staphylococcus aureus bacterial form (ATCC 19636) was grown in heart infusion (HI) broth. S. aureus L form (ATCC 19640), derived from S. aureus ATCC 19636, was grown in HI broth containing 3.5% NaCl, 10 mM CaCl₂, and 1,000 units of penicillin G per ml (to ensure maintenance of the L-form state). Streptococcus pyogenes bacterial form (ATCC 25663) was grown in brain heart infusion (BHI) broth. S. pyogenes L form (ATCC 27080), derived from S. pyogenes ATCC 25663, was maintained on BHI agar medium supplemented with 0.5% (wt/vol) yeast extract, 0.93% NaCl, 9.7% sucrose, 0.025% MgSO₄, 1% horse serum, and 1,000 units of penicillin G per ml. For broth growth, small agar blocks containing S. pyogenes L-form colonies were transferred to the above medium supplemented with 10% horse serum and were incubated for ∼24 h at 37°C under 5% CO₂. A. laidlawii (ATCC 23206) was grown in HI broth supplemented with 10% (vol/vol) yeast extract and 1% PPLO serum fraction (Difco, Detroit, Mich.).

For radial diffusion assays, cultures were grown in their respective broth media and added to molten 0.8% agarose medium at ∼3 × 10⁵ cells/ml. The agarose media were the same as those used for broth growth except that the S. aureus L-form medium lacked CaCl₂, the S. pyogenes L-form medium contained 1% horse serum, and the E. coli J5 medium consisted of HI broth. Five microliters of serially diluted rBPI₂₁, rBPI₄₂, or rBPI or of dilution buffer was added to 3-mm-diameter wells. The plates were incubated at 37°C for 24 h. Inhibition zone diameters were plotted versus BPI concentration. Each assay was performed at least three times for E. coli J5 and the gram-positive bacterial forms and L forms and twice for A. laidlawii.

For the broth microdilution assay, E. coli J5 and S. aureus L-form and bacterial-form cells were grown to log phase in their respective media and diluted to ∼3 × 10⁵ cells/ml in the same media. Five microliters of rBPI₂₁, rBPI₄₂, and rBPI dilutions were added to 95 μl of cells in 96-well plates and were incubated for ∼18 h at 37°C. The results from duplicate samples are expressed in relation to the control (buffer only), for which the value was set at 100%.

Radial diffusion assays (Fig. 1) demonstrated that the tested BPI proteins inhibited growth of A. laidlawii and the L forms of S. aureus and S. pyogenes at concentrations as low as 1 pmol/well (Fig. 1A, B, and C). By comparison, none of the tested BPI proteins, up to ∼470 pmol/well, inhibited growth of the tested S. aureus and S. pyogenes bacterial forms (not shown). As expected, the BPI proteins inhibited growth of E. coli J5 (Fig. 1D). The L forms, the mycoplasma, and E. coli J5 were similarly sensitive to the various BPI proteins. On a molar basis, rBPI₄₂ was most potent against A. laidlawii and the L forms but least potent against E. coli J5. Conversely, rBPI₂₁ was most potent against E. coli J5 but was least potent against A. laidlawii and the S. aureus L form. rBPI was more potent than rBPI₂₁ against A. laidlawii and the S. aureus L form but was least potent against the S. pyogenes L form.

FIG. 1.

Growth inhibition of a mycoplasma and the L forms of gram-positive bacteria by BPI in a radial diffusion assay. The measured diameter includes the 3-mm hole to which the samples were added. (A) S. aureus L form. (B) S. pyogenes L form. (C) A. laidlawii. (D) E. coli J5.

We next examined the effects of rBPI₂₁, rBPI₄₂, and rBPI on S. aureus and its L form in a broth microdilution assay. The tested BPI molecules inhibited growth of the S. aureus L form (Fig. 2A) but not the bacterial form (Fig. 2B). rBPI₄₂ was the most potent while rBPI₂₁ was the least potent protein against this L form. Against E. coli J5 (Fig. 2C), rBPI₂₁ and rBPI were the most potent while rBPI₄₂ was the least potent protein.

FIG. 2.

Growth inhibition of S. aureus and E. coli J5 in a broth microdilution assay. (A) S. aureus L form. (B) S. aureus bacterial form. (C) E. coli J5.

In previous studies, neutrophil-derived BPI did not kill gram-positive bacteria even at concentrations up to 1,000-fold higher than those typically effective against gram-negative bacteria (1). Under the assay conditions of this study, several recombinant human BPI forms were not cytotoxic for S. aureus and S. pyogenes, whereas they inhibited growth of the L forms of these gram-positive bacteria. They also inhibited growth of A. laidlawii. The L forms and Acholeplasma were approximately as sensitive to BPI as the gram-negative bacterium E. coli J5. Based on these results, we conclude that BPI must have similar direct cytotoxic effects on the cytoplasmic membranes of these organisms. These results indicate further that the cell walls of the tested gram-positive bacteria protect them from the BPI proteins used in this study.