Live imaging of the genetically intractable obligate intracellular bacteria Orientia tsutsugamushi using a panel of fluorescent dyes

Abstract

Our understanding of the molecular mechanisms of bacterial infection and pathogenesis are disproportionally derived from a small number of well-characterised species and strains. One reason for this is the enormous time and resources required to develop a new organism into experimental system that can be interrogated at the molecular level, in particular with regards to the development of genetic tools. Live cell imaging by fluorescence microscopy is a powerful technique to study biological processes such as bacterial motility, host cell invasion, and bacterial growth and division. In the absence of genetic tools that enable exogenous expression of fluorescent proteins, fluorescent chemical probes can be used to label and track living cells. A large number of fluorescent chemical probes are commercially available, but these have overwhelmingly been applied to the study of eukaryotic cell systems. Here, we present a methodical analysis of four different classes of probes, which can be used to delineate the cytoplasm, nucleic acids, cell membrane or peptidoglycan of living bacterial cells. We have tested these in the context of the important but neglected human pathogen Orientia tsutsugamushi but expect that the methodology would be broadly applicable to other bacterial species.

Keywords: Fluorescence microscopy; Fluorescent dyes; Intracellular bacteria; Neglected bacterial pathogens; Rickettsia.

Fig. 1

(A) Purified bacteria labelled with CFSE (left panels, green) or CellTrace FarRed (right, red) and imaged live or fixed with paraformaldehyde (PFA) or acetone prior to imaging. (B) Graph showing the bacterial copy number in one well of a 24-well culture plate after 7 days growth in the presence or absence of CFSE/CT FarRed. Inoculum used was 1.36 × 10⁴ ± 1.9 × 10³ bacteria per well (mean ± SD). Individual replicate values are plotted and the mean ± SD is shown. The results of the one-way ANOVA statistical test using the no label condition as a comparison are shown. (C) Images of live L929 cells infected with CFSE or CT FarRed-labelled bacteria. Scale bar = 10 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

(A) Time course showing L929 cells infected with CFSE-labelled or unlabelled bacteria at various times after infection. The top row shows images of live L929 cells infected with CFSE-labelled bacteria. These images were not taken under equivalent imaging conditions (laser strength and acquisition time). The middle and lower rows show samples that have been fixed with paraformaldehyde and labelled with a fluorescently labelled antibody against Orientia TSA56 outer membrane protein (green, bacteria) or the actin-binding dye phalloidin (red, L929 cells). Scale bar = 10 μM. (B) and (C) Pixel intensity of CFSE-labelled bacteria stored in media (B) or added to L929 cells (C) at various times after labelling and infection. Images for this analysis were obtained under constant imaging conditions. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

(A) Purified bacteria labelled with the green DNA-binding dye SYTO9 and imaged live or fixed with paraformaldehyde (PFA) or acetone prior to imaging. (B) Graph showing the bacterial copy number in one well of a 24-well culture plate after 7 days growth in the presence or absence of DMSO or SYTO9. Inoculum used was 4.09 × 10⁵ ± 1.8 × 10⁴ bacteria per well (mean ± SD). Individual replicate values are plotted and the mean ± SD is shown. The results of the one-way ANOVA statistical test using the no label condition as a comparison are shown. (C) Images of live L929 cells infected with SYTO9-labelled bacteria. White arrows indicate bacteria. Scale bar = 10 μm.

Fig. 4

(A) Purified bacteria labelled with the red membrane-binding dye DiD and imaged live or fixed with paraformaldehyde (PFA) or acetone prior to imaging. (B) Graph showing the bacterial copy number in one well of a 24-well culture plate after 7 days growth in the presence or absence of DiD, DiI and DiL. Inoculum used was 1.5 × 10⁵ ± 7.7 × 10³ bacteria per well (mean ± SD). Individual replicate values are plotted and the mean ± SD is shown. The results of the one-way ANOVA statistical test using the no label condition as comparison are shown. **P ≤ 0.01; ***P ≤ 0.001. (C) Images of live L929 cells infected with DiD-labelled bacteria. Scale bar = 10 μm.

Fig. 5

(A) Purified bacteria labelled with the blue peptidoglycan-binding dye HADA and imaged live or fixed with paraformaldehyde (PFA). Arrows point to bacteria where the HADA has labelled as a clear ring (B) Graph showing the bacterial copy number in one well of a 24-well culture plate after 7 days growth in the presence or absence of HADA. Inoculum used was 3.12 × 10⁶ ± 7.6 × 10⁴ bacteria per well (mean ± SD). The results of the one-way ANOVA statistical test using the no label condition as comparison are shown. (C) Images of live L929 cells infected with HADA-labelled bacteria. Scale bar = 10 μm.