Induced expression of the Legionella pneumophila gene encoding a 20-kilodalton protein during intracellular infection

Abstract

The eukaryotic protein synthesis inhibitor cycloheximid has been used by many investigators to selectively radiolabel intracellular bacteria. Although cycloheximide has no direct effect on bacterial gene expression, there are concerns that long-term inhibition of the host cell protein synthesis may have secondary effects on bacterial gene expression. Therefore, prior to further identification and cloning of the macrophage-induced (MI) genes of Legionella pneumophila, the effects of cycloheximide on L. pneumophila-infected U937 cells were evaluated by transmission electron microscopy. Inhibition of protein synthesis of the host cell for 6 h had no major effect on the ultrastructure of the host cell, on the formation of rough endoplasmic reticulum-surrounded replicative phagosome, or on initiation of intracellular bacterial replication. In contrast, by 15 h of cycloheximide treatment, there was profound deterioration in the host cell as well as in the phagosome. To examine protein synthesis by L. pneumophila during the intracellular infection, U937 macrophage-like cells were infected with L. pneumophila, and intracellular bacteria were radiolabeled during a 2-h cycloheximide treatment or following 12 h of cycloheximide treatment. Comparison by two-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the protein profile of radiolabeled in vitro-grown L. pneumophila to that of intracellularly radiolabeled bacteria showed that 23 proteins were induced in response to the intracellular environment during 2 h of inhibition of host cell protein biosynthesis. Twelve MI proteins of L. pneumophila were artifactually induced due to prolonged inhibition of the host cell protein synthesis. The gene encoding a 20-kDa MI protein was cloned by a reverse genetics technique. Sequence analysis showed that the cloned gene encoded a protein that was 80% similar to the enzyme inorganic pyrophosphatase. Studies of promoter fusion to a promoterless lacZ gene showed that compared to in vitro-grown bacteria, expression of the pyrophosphatase gene (ppa) was induced fourfold throughout the intracellular infection. There was no detectable induction in transcription of the ppa promoter during exposure to stress stimuli in vitro. The ppa gene of L. pneumophila is the first example of a regulated ppa gene which is selectively induced during intracellular infection and which may reflect enhanced capabilities of macromolecular biosynthesis by intracellular L. pneumophila. The data indicate caution in the long-term use of inhibition of host cell protein synthesis to selectively examine gene expression by intracellular bacteria.

FIG. 1

Electron micrographs of infected U937 cells during inhibition of host cell protein synthesis at 1 h (a), 2 h (b), 4 h (c), 6 h (d), 8 h (e), and 15-h (f) postinfection. Arrowheads indicate the smooth multilayer phagosomal membrane, arrows indicate the RER surrounded phagosome, M indicates mitochondria, and B indicates bacteria. Note the vesicles making contact and possibly fusing with the phagosomal membrane at 8 h in panel e. Note the absence of a defined membrane around the phagosome at 15 h in panel f. Bars, 0.5 μm.

FIG. 2

Electron micrographs of untreated infected U937 cells at 8 h (A) and 15 h (B) postinfection. Arrowheads indicate the RER-surrounded phagosome.

FIG. 3

Autoradiograph of two-dimensional SDS-PAGE of L. pneumophila proteins radiolabeled intracellularly in U937 cells during 2 h of cycloheximide treatment and compared to the protein profile of intracellular bacteria radiolabeled during the same growth phase after 12 h of inhibition of host cell protein synthesis and to in vitro-grown bacteria during the same growth phase (3). The alkaline-to-acid gradient is from left to right. Arrowheads indicate the proteins that were induced in response to the intracellular environment. Closed arrows indicate the proteins that were artifactually induced as a result of prolonged inhibition of host cell protein synthesis (3). The open arrow indicates the 20-kDa PPase that is uniquely induced in response to the intracellular environment. The spots in squares represent the two major spots shown in Fig. 4 on both sides of the 20 kDa PPase. The spot in the circle represents the protein to which the levels of 20-kDa PPase was normalized.

FIG. 4

Magnification of portions of two-dimensional gels to illustrate the protein level of the PPase protein expressed in vitro-grown bacteria (A) compared to that expressed by intracellular bacteria (B). The arrows indicate the 20-kDa PPase that is uniquely induced in response to the intracellular environment. Although three spots on the left of PPase seem to be slightly induced in this gel, they were not reproducible in multiple experiments.

FIG. 5

(Top) Restriction map of pJA2 insert; (bottom) Southern hybridization of chromosomal DNA of L. pneumophila digested and probed with the degenerate oligonucleotide corresponding to the N-terminus sequence of the 20-kDa MI protein. Lanes: 1, 1.8-kb BamHI; 2, 11-kb ClaI; 3, 2.2-kb EcoRI; 4, 3.2-kb EcoRV; 5, 8-kb HindIII. The second signal in lane 4 is an artifact. The sizes were based on a 1-kb ladder (BRL).

FIG. 6

Primer extension analysis of ppa transcript. A radiolabeled oligonucleotide was annealed to RNA isolated from BYE-grown bacteria, and reverse transcriptase was added to produce cDNA. The same oligonucleotide was used to prime the dideoxy sequencing reaction of pJA2 DNA. The letters above each lane indicate the dideoxy nucleotide used to terminate each reaction. The arrow indicates the 5′ end of the mRNA specie corresponding to nucleotide 140.

FIG. 7

In vitro transcription-translation of pJA2 (lane 2) and the vector pBC (lane 1). The arrow indicates the 20-kDa ³⁵S-labeled protein. Sizes were estimated based on Bio-Rad low-molecular-mass prestained protein standards.

FIG. 8

Alignment of L. pneumophila (Lpn) PPase to the E. coli (Ec) homolog. Conservative substitutions are indicated by colons. The amino acid residues that have been shown to be crucial for structural and catalytic activity of the enzyme in E. coli (see Discussion) are indicated in bold.

FIG. 9

Kinetics of β-Gal expression by in vitro-grown (A) and macrophage-grown (B) L. pneumophila harboring a ppa promoter (pPPZ) fused to the promoterless lacZ gene of pEU730. Data are expressed as the mean β-Gal activity for triplicate samples (34). L. pneumophila harboring the vector pEU730 did not show any detectable β-Gal activity (data not shown).