Beta-lactamase can function as a reporter of bacterial protein export during Mycobacterium tuberculosis infection of host cells

Abstract

Mycobacterium tuberculosis is an intracellular pathogen that is able to avoid destruction by host immune defences. Exported proteins of M. tuberculosis, which include proteins localized to the bacterial surface or secreted into the extracellular environment, are ideally situated to interact with host factors. As a result, these proteins are attractive candidates for virulence factors, drug targets and vaccine components. Here we describe a beta-lactamase reporter system capable of identifying exported proteins of M. tuberculosis during growth in host cells. Because beta-lactams target bacterial cell-wall synthesis, beta-lactamases must be exported beyond the cytoplasm to protect against these drugs. When used in protein fusions, beta-lactamase can report on the subcellular location of another protein as measured by protection from beta-lactam antibiotics. Here we demonstrate that a truncated TEM-1 beta-lactamase lacking a signal sequence for export ('BlaTEM-1) can be used in this manner directly in a mutant strain of M. tuberculosis lacking the major beta-lactamase, BlaC. The 'BlaTEM-1 reporter conferred beta-lactam resistance when fused to both Sec and Tat export signal sequences. We further demonstrate that beta-lactamase fusion proteins report on protein export while M. tuberculosis is growing in THP-1 macrophage-like cells. This genetic system should facilitate the study of proteins exclusively exported in the host environment by intracellular M. tuberculosis.

Fig. 1. Schematic representation of signal sequence-‘BlaTEM-1 fusion constructs

Mycobacterial shuttle plasmids were designed to encode fusion proteins of M. tuberculosis peptide sequence (open boxes) with a truncated ‘BlaTEM-1 protein (gray boxes) lacking its native signal sequence. The hatched boxes indicate plasmid-derived peptide sequence that is present as a result of the cloning process. The constructs were driven off the constitutive M. tuberculosis hsp60 promoter for (a) pJES102/‘blaTEM-1, (b) pJES103/ssmpt63-‘blaTEM-1, and (d) pJES101/ssplcB-‘blaTEM-1. The native M. tuberculosis promoter located upstream of the mpt83 operon was used to drive expression of (c) pJES129/ ssmpt83-‘blaTEM-1 (promoters indicated by arrows). Signal peptidase cleavage sites are indicated by arrowheads and by the AxA/G recognition motif for PlcB and Mpt63, and the LAGC lipobox recognition motif for Mpt83. Diagram not to scale; ss, signal sequence.

Fig. 2. 'BlaTEM-1 does not provide β-lactam-resistance to ΔblaC M. tuberculosis

Plasmids encoding the indicated 'blaTEM-1 fusions were electroporated into M. tuberculosis ΔblaC. The resulting strains were then plated on 7H10 plates supplemented with either kanamycin and 0.05% tween or kanamycin and carbenicillin without tween. Plates were inspected for growth following 21–25 days of incubation. Not shown are colonies expressing ssMpt83-‘BlaTEM-1 and ssPlcB-‘BlaTEM-1; growth on plates containing carbenicillin for these strains was similar to that conferred by ssMpt63-‘BlaTEM-1.

Fig. 3. ‘BlaTEM-1 fusion proteins are detected at different amounts in M. tuberculosis whole cell lysates

Protein present in whole cell lysates (WCL) from each of the indicated ΔblaC strains were separated by SDS-PAGE and immunoblotted using primary antibody specific for BlaTEM-1. Comparative signal was quantified by measuring pixel density of an equal area for each blotted lysate in duplicate. Average signal intensity per µg of WCL is reported as the amount relative to protein detected in the ‘BlaTEM-1 expressing strain. Due to the different amounts of protein in each strain, it was necessary to load dilutions of the ‘BlaTEM-1 and ssPlcB-‘BlaTEM-1 expressing lysates so that signal from less abundant protein fusions could be simultaneously detected. There was no detectable signal with the WCL from the ΔblaC mutant carrying empty pMV261 plasmid.

Fig. 4. The ΔblaC mutant of M. tuberculosis does not have a growth defect and is sensitive to β-lactam antibiotic in human THP-1 macrophage-like cells

(a) THP-1 cells were seeded into 8 well chamber slides, and triplicate wells were infected with either WT H37Rv or ΔblaC M. tuberculosis at a m.o.i. of 0.1 bacilli per macrophage. At 4 hours (Day 0), 1, 3 and 5 days post infection, infected wells were washed, lysed, and plated for intracellular bacteria. Error bars represent standard error of the mean of c.f.u. in triplicate wells. (b) THP-1 cells were infected with ΔblaC M. tuberculosis as in (a). Following a 4-hour uptake period, wells were washed and indicated concentrations of carbenicillin were added to infected wells. Infected cells were lysed and plated at 4 hours (Day 0) and 5 days post infection to enumerate intracellular bacteria. Dashed line represents average intracellular CFU at 4 hours post infection. Error bars represent standard error of the mean of quadruplicate wells combined from two replicates.

Fig. 5. M. tuberculosis signal sequences fused to ‘BlaC and ‘BlaTEM-1 protect intracellular bacilli from β-lactam antibiotics

THP-1 macrophage like cells were infected in triplicate wells with (a) M. tuberculosis ΔblaC expressing either ssPlcB-‘BlaC or ‘BlaC, (b) M. tuberculosis ΔblaC expressing either ‘BlaTEM-1 or ssMpt63‘BlaTEM-1 or (c) M. tuberculosis blaC expressing ‘BlaTEM-1or ssMpt83-‘BlaTEM-1. Infected cells were then left untreated or treated with 1 mg/ml carbenicillin (carb). Wells were washed, lysed and plated 4 hours (d 0), 1, 3 and 5 days post infection. Each experiment was replicated 3 times in the case of (a) and (b), and 4 times in (c), with similar results.