Single-copy green fluorescent protein gene fusions allow accurate measurement of Salmonella gene expression in vitro and during infection of mammalian cells

Abstract

We developed a reliable and flexible green fluorescent protein (GFP)-based system for measuring gene expression in individual bacterial cells. Until now, most systems have relied upon plasmid-borne gfp gene fusions, risking problems associated with plasmid instability. We show that a recently developed GFP variant, GFP+, is suitable for assessing bacterial gene expression. Various gfp+ transcriptional fusions were constructed and integrated as single copies into the chromosome of Salmonella enterica serovar Typhimurium. A comparison of the expression levels of proU-lacZ and proU-gfp+ fusions showed that GFP+ reported proU activity in individual Salmonella cells as accurately as beta-galactosidase reported activity for entire populations. The single-copy gfp+ fusions were ideal for monitoring up- and downregulation of Salmonella virulence genes. We discovered that in vitro induction of the SPI1gene prgH occurs only in a portion of the population and that the proportion varies with the growth phase. We determined the level of expression of the SPI2 gene ssaG in bacteria released from murine macrophages. Our results demonstrate for the first time that single-copy GFP+ fusions reliably report gene expression in simple and complex environments. This approach promises to allow accurate measurement of gene expression in individual bacteria during animal infection.

FIG. 1.

Strategy used for construction of single-copy gfp⁺ fusions. Promoters of interest were amplified by PCR from the Salmonella chromosome and inserted into the pZEP08 plasmid in place of the kanamycin resistance cassette (A). The resulting plasmid was then used as a template for PCR amplification of the fragment that contained the T1 terminator, the new gfp⁺ transcriptional fusion, and the chloramphenicol resistance cassette by using primers that had 40- to 50-nucleotide tails (H1 and H2) exhibiting perfect homology with the chromosomal site of insertion (i.e., putPA locus) (B). The linear PCR product containing the new fusion was moved to the chromosome of the recipient Salmonella strain by recombination by using the Lambda Red system (13) (C).

FIG. 2.

Comparison of rpsM-gfpmut3 and rpsM-gfp⁺ expression in LB medium and after formalin fixation. E. coli strains harboring either pFPV25.1 (rpsM-gfpmut3) or pZEP02 (rpsM-gfp⁺) were grown overnight in LB broth containing ampicillin. Live or fixed (4% formalin) bacteria were immediately analyzed by flow cytometry. (A) Unfixed bacteria harboring either pFPV25.1 (shaded graph) or pZEP02 (solid line). (B and C) Fixed (solid line) and unfixed (shaded graph) bacteria harboring either pFPV25.1 expressing GFPmut3 (B) or pZEP02 expressing GFP⁺ (C). The values are the median values for fluorescence intensity for all individual bacteria in a population.

FIG. 3.

Osmotic induction of proU. Induction of proU-lacZ and proU-gfp⁺ expression was tested as follows. Salmonella strains CH946 (proU-lacZ; β-galactosidase [β-Gal] activity indicated by the dashed line), JH3017 (proU-gfp⁺; GFP fluorescence indicated by the solid black line), and JH3049 (promoterless gfp⁺; GFP fluorescence indicated by the solid grey line) were grown in LO medium containing glucose at 30°C to an optical density at 600 nm of 0.5. NaCl was then added to a final concentration of 0.06 M (A), 0.16 M (B), or 0.3 M (C). Samples were collected at 0, 5, 10, 20, 40, 60, 90, and 120 min after addition of the salt. The optical density of each culture was measured at 600 nm (panel D shows the growth curves obtained before and after addition of 0.3 M NaCl). The arrow indicates when the salt was added to each mid-log-phase culture. The median β-galactosidase activities for three independent experiments are shown (see Materials and Methods). The GFP⁺ fluorescence intensities are the median values for the intensities of all individual bacteria in a population for seven independent experiments. The error bars indicate the standard deviations.

FIG. 4.

Virulence gene induction in vitro. Salmonella serovar Typhimurium strains JH3009 (ssaG-gfp⁺) and JH3008 (promoterless gfp⁺) were grown overnight in LB broth. Both strains were subsequently diluted 50-fold in MM5.8 and were grown for 24 h. Time zero corresponded to the beginning of incubation in MM5.8. Samples were collected every hour from time zero until 10 h and once after 24 h, immediately fixed in 4% formalin, and analyzed by flow cytometry. (A) Fluorescence intensity of Salmonella strain JH3009 (grey bars) determined in triplicate. The fluorescence intensity of the negative control JH3008 strain is also indicated for each time point (solid bars). The values in the bars are the median fluorescence intensities of all individual bacteria in the populations. The error bars indicate the standard deviations. (B) Corresponding growth curve. Dashed line, strain JH3008; solid line, strain JH3009.

FIG. 5.

Differential expression of a promoter within a genetically identical population. Salmonella serovar Typhimurium strain JH3010 (carrying prgH-gfp⁺) was grown overnight in LO medium containing no salt. A culture containing 10⁴ bacteria per ml (final concentration) was then grown in either LO medium (solid line) or LO medium containing 0.3 M NaCl (shaded graph). Samples were collected after 4 h (A), 6 h (B), 8 h (C), 12 h (D), or 24 h (E) of growth with or without salt (F) (dotted line, LO medium; solid line, LO medium containing 0.3 M NaCl), fixed in 4% formalin, and analyzed by flow cytometry. The level of induction was calculated by comparing the fluorescence for the most fluorescent peak at each time point with the fluorescence at the same time obtained when strain JH3010 was grown in LO medium. The percentage of the population in each fluorescence peak is indicated on the graph. To ensure that the differential expression of the prgH-gfp⁺ fusion was genuine, the experiment was repeated 25 times, and the data from a single representative experiment are shown.

FIG. 6.

Single-copy SPI2 gene fusion accurately reports gene expression in mammalian cells. J774-A.1 murine macrophages were infected with Salmonella serovar Typhimurium strains JH3008, JH3009, and JH3016 harboring a promoterless gfp⁺ gene, an ssaG-gfp⁺ fusion, and an rpsM-gfp⁺ fusion, respectively. Six hours after infection, intracellular bacteria were released under hypotonic conditions and immediately fixed in 4% formalin. Salmonella cells were then labeled with a specific antibody, and their fluorescence was measured by flow cytometry. The results shown represent the GFP⁺ fluorescence intensity of extracellular or intracellular bacteria identified as Salmonella cells by antibody labeling. Each panel shows an overlay of the GFP⁺ fluorescence of either JH3009 or JH3016 (shaded graph) with the GFP⁺ fluorescence of JH3008 (solid line), which was used as negative control. (A and B) Levels of expression of ssaG-gfp⁺ or rpsM-gfp⁺ in extracellular Salmonella cells. (C and D) Fluorescence in bacteria of ssaG-gfp⁺ or rpsM-gfp⁺ released from inside macrophages at 6 h postinfection. The values indicate the median GFP fluorescence intensity of all individual bacteria in a population.