Abstract
Current spore detection methods rely on culture techniques, with limitations of time, efficiency, and sensitivity. The bacterial spore coat contains calcium dipicolinic acid (CaDPA) as a major constituent, which could serve as a biomarker for bacterial endospores. We report proof of concept for a rapid and sensitive technique for the detection of bacterial endospores by using ratiometric fluorescence-based sensors. This method is based on the detection of CaDPA, which enhances the luminescence of lanthanide ions when complexed with a semiconducting polymer. A CaDPA standard curve was generated at an excitation-emission wavelength (λ) of λ(275)-λ(544) by using a spectrophotometer. The intensity was recorded after chelating semiconducting fluorescent polyfluorene (PFO) dots with terbium (lanthanide) ions, sensitized by different volumes of CaDPA (0.1 μM). The resultant standard curve showed a linear relationship (R(2) = 0.98) in the experimental concentration range of 2.5 to 25 nM CaDPA, with corresponding intensity (arbitrary units) of 545 to 2,130. Endospores of the aerobic sporeformer Bacillus licheniformis ATCC 14580 were produced at 37°C for 15 d on brain heart infusion agar plates. The efficiency of sporulation was evaluated by spore staining and plating techniques. Total CaDPA content on spores was estimated after suspending decreasing concentrations of spores (logs 9.0 through 1.0 cfu/mL, at 1-log intervals) in HPLC-grade water (to serve as control) and skim milk samples. In HPLC-grade water, for higher spiking levels such as (mean ± SD) 9.2 ± 0.03, 8.4 ± 0.05, 7.1 ± 0.13, and 6.3 ± 0.02 logs, the corresponding mean CaDPA from the standard curve were 9.4, 7.2, 6.2, and 5.3 nM, respectively. For lower spiking levels of 4.2 ± 0.05, 3.1 ± 0.04, 2.0 ± 0.11, and 1.36 ± 0.09 logs, we observed mean CaDPA contents of 3.8, 3.3, 2.2, and 1.3 nM, respectively. For raw skim milk spiked with B. licheniformis ATCC 14580 spores, the mean CaDPA content on spores was approximately 2.5, 3.8, and 5.0 nM for spiking levels of 5.21, 6.39, and 9.47 log cfu/mL, respectively. Trials were conducted in replicates of 3 and means were compared. Trials conducted using HPLC-grade water showed a linear relationship for the CaDPA content of endospores and for endospore counts with the standard CaDPA concentration curve. For skim milk-spiked samples, we observed reduced fluorescence detection, which was 5 times lower than that of spiked samples in HPLC-grade water. The reduced fluorescence in skim milk could be due to the turbidity of the solution or to interference from proteins, amino acids, and other ions in milk. This study thus provides proof of concept for a potential application of this technique to rapidly detect bacterial endospores in the dairy and food industry. Further work is required to remove the interference of ionic components in milk to improve detection limits in milk and other dairy product matrices such as cheese, whey proteins, and reconstituted powders.