Development of a Staphylococcus aureus reporter strain with click beetle red luciferase for enhanced in vivo imaging of experimental bacteremia and mixed infections

Abstract

In vivo bioluminescence imaging has been used to monitor Staphylococcus aureus infections in preclinical models by employing bacterial reporter strains possessing a modified lux operon from Photorhabdus luminescens. However, the relatively short emission wavelength of lux (peak 490 nm) has limited tissue penetration. To overcome this limitation, the gene for the click beetle (Pyrophorus plagiophtalamus) red luciferase (luc) (with a longer >600 emission wavelength), was introduced singly and in combination with the lux operon into a methicillin-resistant S. aureus strain. After administration of the substrate D-luciferin, the luc bioluminescent signal was substantially greater than the lux signal in vitro. The luc signal had enhanced tissue penetration and improved anatomical co-registration with infected internal organs compared with the lux signal in a mouse model of S. aureus bacteremia with a sensitivity of approximately 3 × 10⁴ CFU from the kidneys. Finally, in an in vivo mixed bacterial wound infection mouse model, S. aureus luc signals could be spectrally unmixed from Pseudomonas aeruginosa lux signals to noninvasively monitor the bacterial burden of both strains. Therefore, the S. aureus luc reporter may provide a technological advance for monitoring invasive organ dissemination during S. aureus bacteremia and for studying bacterial dynamics during mixed infections.

Figure 1

Genetic maps of lux and luc plasmid constructs. (A) Plasmid pHC125 lux used to generate chromosomally integrated luxCDABEG construct. Expression is driven by the P_CP25 and P_CAP promoters (prom). Plasmid stably integrates at the phage 11 attachment site in S. aureus. (B) Plasmid pHC123 CBR-luc carrying constitutively expressed click beetle red luciferase (CBR-luc). Expression is driven by the P_lgt promoter.

Figure 2

In vitro bacterial growth curves and bioluminescence signals. AH4807 (lux), AH4775 (luc), AH4826 (lux + luc), USA300 LAC::lux, LAC4303 (lux) and Xen36 (lux) (all 1 × 10⁴ CFU) were cultured in 96-well plates without (none) and with the addition of different concentrations (0.03–1.2 mg/240 µL TSB/well) of D-Luciferin for 0–14 hours and absorbance at 600 nm (A₆₀₀) (A) and bioluminescent signals (photons/0.1 second acquisition time) (B) were measured every 5 minutes on a plate reader (BioTek H1 Synergy) (n = 4 replicates with 2 iterations).

Figure 3

In vitro lux versus luc BLI signals through tissue. (A) AH4826 (lux + luc) was cultured on bacterial culture petri dishes overnight and BLI was performed for a 30 second acquisition time at 37 °C without (top panels [lux signals]) and with the addition of D-Luciferin (600 ng/200 µL PBS pipetted directly onto the surface of the plates) (bottom panels [lux + luc signals]) on an IVIS Lumina III (PerkinElmer) (n = 3 replicate plates with 2 iterations). Representative BLI signals are provided with no filter (open) and with 520, 570, 620, 670, 710 and 790 nm emission filters. (B) AH4807 (lux), AH4775 (luc), AH4826 (lux + luc), USA300 LAC::lux, LAC4303 (lux) and Xen36 (lux) (all 1 × 10⁹ CFU) were cultured in 96-well plates (n = 3 replicate wells for each condition with 2 iterations). BLI was performed for a 1 minute acquisition time at 37 °C in an IVIS Lumina III (PerkinElmer) without (none) or with different thicknesses of tissue (sliced cooked ham: none, 5.25, 10.5, 15.75 and 21.0 mm) placed on top of the plate covers ± different concentrations of D-Luciferin (0.03–2.4 mg/280 µL TSB) added to the wells. Representative BLI signals are shown. n.t. = not tested.

Figure 4

In vivo BLI of lux versus luc in a S. aureus bacteremia mouse model. A high inoculum (1 × 10⁷ CFU, 20% lethal dose) of AH4826 (lux + luc) or LAC4303 (lux) was injected intravenously in mice (n = 3 mice/group). On day 3, in vivo BLI signals were acquired before (−5 minutes) and at the indicated time points (up to 65 minutes) after administration of D-Luciferin (150 mg/kg s.c.) (IVIS Lumina III, PerkinElmer). (A,D) Representative images of in vivo BLI of the ventral (A) and dorsal (D) sides of the mice. (B,E) In vivo BLI signals (photons/s) ± SEM of the ventral (B) and dorsal (E) sides of mice inoculated with AH4826 ± administration D-Luciferin (150 mg/kg s.c.). (C,F) In vivo BLI signals (photons/s) ± SEM of the ventral (C) and dorsal (F) sides of mice inoculated with AH4826 (lux + luc) or LAC4303 (lux) ± administration of D-Luciferin (150 mg/kg s.c.) at 15–25 minutes prior to imaging the mice. ^†P < 0.01 between indicated groups, as calculated by a 2-tailed Mann–Whitney U test (C,F).

Figure 5

In vivo BLI of lux versus luc in additional models of S. aureus infection. (A,B) A mouse model of S. aureus skin infection was performed with inoculation of AH4826 (lux + luc) (1 × 10⁸ CFU) i.d. into the back skin of mice (n = 10 mice/group). In vivo BLI imaging was performed ± administration of D-Luciferin (150 mg/kg s.c.) at 15–25 minutes prior to imaging the mice (n = 10/group). (A) Representative in vivo BLI. (B) In vivo BLI signals (photons/s) ± SEM. (C–E) A rabbit model of S. aureus orthopaedic implant associated infection (OIAI) was performed with inoculation of AH4826 (lux + luc) (1 × 10⁴ CFU) into a femoral intramedullary canal prior to surgical placement of an orthopaedic-grade titanium locking peg (n = 6 rabbits/group) and in vivo BLI imaging was performed ± administration of D-Luciferin (150 mg/kg s.c.) at 15–25 minutes prior to imaging the rabbits. (C) Representative in vivo BLI. (D) In vivo BLI signals (photons/s) ± SEM. (E) Skin tissue from day 10 was homogenized and ex vivo CFU of AH4826 (lux + luc) was cultured on bacterial culture petri dishes overnight (n = 5 replicates) and BLI was performed for a 30 second acquisition time at 37 °C ± the addition of D-Luciferin (600 ng/200 µL PBS pipetted directly onto the surface of the plates) on an IVIS Lumina III (PerkinElmer) and representative BLI signals are provided with no filter (open) and with 520, 570, 620, 670 and 710 nm emission filters. *P < 0.05 between ± administration of D-Luciferin, as calculated by a 2-way ANOVA (B,D). n.s. = not significant.

Figure 6

3D localization in vivo BLI of lux versus luc in a S. aureus bacteremia mouse model. A sub-lethal inoculum (1 × 10⁶ CFU) of AH4826 (lux + luc) was injected intravenously in mice (n = 10 mice/group) and mice were imaged ± administration of D-Luciferin (150 mg/kg s.c.) at 15–25 minutes prior to in vivo BLI. (A) Representative in vivo BLI images of the dorsal sides of the mice on day 3 obtained using the 2D IVIS Lumina III imaging system (PerkinElmer). (B) In vivo BLI signals (photons/s) ± SEM of the ventral (left) and dorsal (right) sides of mice. (C) Mice were euthanized on day 3 and right and left kidneys, liver and heart were harvested and homogenized and the ex vivo CFU were enumerated after overnight culture on plates. Each symbol is an individual mouse. Light blue bars = geometric mean. LOD = limit of detection [10 CFU]). (D,E) Representative 2D and 3D in vivo BLI images of the dorsal sides of the mice (n = 10 mice) on day 3 obtained using the IVIS Spectrum-CT imaging system (PerkinElmer). *P < 0.05 between ± administration of D-Luciferin, as calculated by a 2-way ANOVA (B).

Figure 7

The sensitivity of detection of AH4775 (luc) by in vivo BLI in the S. aureus bacteremia mouse model. Sixteen hours post-intravenous inoculation of 1 × 10⁶ CFU of AH4775 (luc) in mice (n = 8 mice), D-Luciferin (150 mg/kg s.c.) was administered and after 20 minutes in vivo BLI signals was performed on the dorsal sides of the mice and in vivo BLI signals were acquired as total flux (photons/s) within region of interests corresponding to the right and left kidneys (IVIS Lumina III). Mice were then immediately euthanized and the right and left kidneys were separately isolated to determine the ex vivo CFU. (A) Representative images of in vivo BLI signals of a mouse with the highest CFU but below the level of detection (LOD) of the IVIS Lumina III (left, red arrow) and a mouse with the lowest detectable in vivo BLI signal from the left kidney that was above the LOD (right, black arrow). (B) Correlation between in vivo BLI (total flux [photons/s]) and ex vivo CFU from the right and left kidneys of the mice (n = 8 mice) with the red and black arrows denoting the representative mice shown in (A), respectively. LOD = 1 × 10⁴ photons/s (horizontal black dashed line). The pink boxed area represents the range of sensitivity of detection from the highest CFU that was not detected (4.2 × 10³ CFU) to the lowest CFU that was detected (3.3 × 10⁴ CFU) by in vivo BLI. The linear regression line and correlation coefficient of determination (R² = 0.9924) of in vivo BLI and ex vivo CFU of AH4775 (luc) are shown.

Figure 8

Dual in vivo BLI monitoring of a mixed S. aureus and P. aeruginosa in vivo wound infection mouse model. A excisional wound mixed infection model by performing a 6-mm punch biopsy on the backs of mice and the wound beds were inoculated with S. aureus AH4775 (luc) (2 × 10⁶ CFU) and P. aeruginosa Xen41 (lux) (2 × 10⁵ CFU) (n = 5 mice). The mixed wound infection was followed for 7 days and CFU from the wounds were harvested, plated, cultured and enumerated. To distinguish between the in vivo BLI signals of AH4775 (luc) versus Xen41 (lux), the respective 670 nm and 520 nm emission filters (IVIS Lumina III) were used because there was no overlap between the luc and lux bioluminescent signals. (A) Representative images of the in vivo BLI signals from AH4775 (luc) and Xen41 (lux) from the wounds. (B) Mean in vivo BLI signals of AH4775 (luc) and Xen41 (lux) quantified as total flux (photons/s) ± SEM. (C) Representative in vivo BLI of bacterial culture plates possessing CFU of AH4775 (luc) and Xen41 (lux) using the 670 nm and 520 nm emission filters, respectively. (D) Mean CFU of AH4775 (luc) and Xen41 (lux), with horizontal bars = geometric mean. *P < 0.05 between AH4775 (luc) and Xen41 (lux), as calculated by a 2-way ANOVA (C). n.s. = not significant (P = 0.095), as calculated by a 2-tailed Mann–Whitney U test (D).