2.1. LIVE/DEAD BacLight^TM Staining

BacLight^TM is a differential staining method used to detect viable, non-viable and total bacteria cells present in a sample. The kit uses two nucleic acid-binding stains, SYTO 9 and propidium iodide (PI), contained in a dimethylsulfoxide solution (DSMO) [22]. SYTO 9 permeates the cell membrane of both viable and non-viable cells and stains these cells green. Propidium iodide only penetrates cells with compromised membranes (i.e., non-viable cells) and reduces the SYTO 9 stain, thereby staining the cells red. The green and/or red Lm cells are typically analyzed and enumerated using epifluorescence microscopy but can also be counted with flow cytometry [20,23,24,25].

When BacLight^TM staining was run in conjunction with flow cytometry, high correlations of live and dead Lm cell counts were found (r² = 0.97 and r² = 0.99 respectively) between ratios of 10% to 100% (living to dead cells) when compared to known control populations of live and dead cells [26]. BacLight^TM stained Lm cells under microscopy also gave a more accurate estimation of viable cell counts when compared to EMA/real-time PCR [27]. By using BacLight^TM staining in conjunction with flow cytometry and comparing it with direct viable counts, the membrane integrity of VBNC Lm populations can be demonstrated, even as plate-count cultivability was reduced to nearly 0% [20].

When the correlation between the percentage of metabolically active bacterial cells added to a sample was compared to the percentage of active cells measured by the BacLight^TM kit, a statistically significant correlation, r² ≥ 0.98, was found for A. hydrophila, B. subtilis, E. coli, P. aeruginosa, and S. epidermidis [24]. When compared to 5-cyano-2,3-ditolyl tetrazolium chloride—4′,6-diamidino-2-phenylindole (CTC-DAPI) staining, BacLight^TM produced equal to or better accuracy in detecting viable and total counts of E. coli in several tests [22]. The accuracy of BacLight^TM is seen in a wide range of bacteria, and it is also effective in the analysis of Gram-positive pathogens such as Lm. BacLight^TM is commonly used to analyze Lm in biofilms [28,29] and to determine the effectiveness of antimicrobial treatments to specifically targeting cell enumeration in biofilms [29].

However, clear bimodal (green and red) staining is not always possible. Frequently, a gradient from dual staining is observed, resulting in some difficulty interpreting the results [30,31]. At the very least, mixed results warrant increased controls and verification with complementary methods; they might also suggest a gradient in cell viabilities. It also must be noted that cells with intact membranes are not always considered viable, and so an overestimation may occur [32]. This method is rapid, inexpensive, simple to perform, and is probably the most widely employed method for differentiating VBNC cells. It does, however, require an epifluorescent microscope or a flow cytometer.

2.2. EMA- and PMA-Stained Real-Time PCR

Quantitative PCR (qPCR) targets the unique DNA sequences of foodborne pathogens such as Lm and amplifies these sequences to detectable levels. However, qPCR also has the potential to amplify intact DNA from dead Lm cells [33], and thus standard qPCR analysis may lead to an overestimation of viable cell counts [34]. This is especially problematic when attempting to detect VBNC Lm because VBNC bacteria do not grow using routine enrichment and plate count techniques. Due to these limitations, an environmental swab taken from a food contact surface and analyzed by qPCR could have viable Lm, VBNC Lm, and dead Lm cell DNA all being amplified. In the case of viable Lm and VBNC Lm, DNA being amplified by qPCR, the swab would correctly confirm a true positive. However, any amplified dead Lm cell DNA would confirm a false positive, leading to expensive, erroneous recommendations for corrective actions. Fortunately, there are additional techniques to minimize these false positives.

Ethidium monoazide (EMA) and propidium monoazide (PMA) are both DNA-binding agents used in combination with Real-Time Polymerase Chain Reaction (RT-PCR) to prevent the DNA of dead bacterial cells from being amplified. EMA penetrates damaged membranes of dead cells [35] and, when exposed to light, irreversibly binds to the dead cells’ DNA [36]. The EMA-bound cell DNA will not be amplified by subsequent PCR reactions, thus preventing false positives or over-estimation of viable cells [37,38], including Lm cells [35]. The exposure to light also inactivates any remaining free EMA in the sample, preventing subsequent binding to the DNA of viable cells during the DNA extraction step [39].

However, it is possible that EMA can penetrate viable cell membranes of certain bacterial species, including Escherichia coli O157:H7 [40] and Lm [37,41], which would result in too low an estimate of viable cells. To minimize this concern, Propidium monoazide (PMA) can be used as an alternative to EMA as a non-viable cell stain. Unlike EMA, PMA has not been shown to penetrate the membranes of living cells. This increased selectivity is possibly due to the higher molecular charge on PMA [37]. In conjunction with real-time PCR, PMA prevents the DNA of dead Lm cells from being amplified while allowing viable cell DNA (including VBNC cells) to be amplified, detected, and quantified, even in complex food matrixes [42].

2.3. Direct Viable Count

The direct viable count (DVC) method was originally developed to enumerate bacteria in samples of seawater [44]. Its uses have expanded to the detection of VBNC cells of many bacterial species, including Lm [45]. In the DVC method, a limited level of nutrients (such as yeast extract) and an antibiotic that inhibits DNA replication are added to a water sample. To test for the VBNC state in Gram negative bacteria, nalidixic acid is usually used, but because it is not effective against Gram positive bacteria like Lm, an alternative antibiotic such as ciprofloxacin must be used [45]. Once the antibiotic and minimal nutrients are added, the water sample is incubated for 7 h at 37 °C. The antibiotic allows viable cells to begin the onset of cell growth and elongate in response to the yeast extract but prevents their cell division [44]. These elongated cells are then stained with a fluorescent dye and are subsequently directly counted with an epifluorescence microscope [45]. Any bacteria that have elongated to twice their normal cell length are considered to have been viable. The number of elongated cells can be compared to their corresponding plate counts to determine the number of cells in the VBNC state. The advantage of this method over other dye methods is that non-viable cells with intact membranes will not elongate. When analyzing microcosm water starved Lm cells for VBNC state, CTC-DAPI double staining and DVC resulted in the same average number of metabolically active Lm cells (10⁶ bacteria ml⁻¹), with DVC analyzing the cell elongation of Lm and CTC-DAPI measuring cell respiration of Lm cells through CTC fluorescence [14]. Similar counts were also obtained between DVC and CTC-DAPI when analyzing NaCl levels as a physiochemical trigger for inducing VBNC state in Lm [15]. Highmore et al. [18] used a DVC variation with green-fluorescent protein producing Lm and pipemidic acid. They demonstrated that VBNC Lm remained above 10⁶ bacteria ml⁻¹ when exposed to 50, 80, and 100 ppm chlorine, while traditional culturable plate counts were below the limit of detection [18]. Care must be used in analyzing DVC cells, as the microscopic determination of a doubling of cell length is quite subjective and prone to operator error.

2.4. CTC-DAPI Double Staining

CTC is a tetrazolium salt that is reduced to CTC-formazan by the electron transport system in actively respiring bacterial cells [46]. The CTC-formazan subsequently shows up as red/purple, fluorescent precipitant under epifluorescence microscopy. CTC has been shown to be an ideal stain due to its stability and being able to be detected at low levels because it shows a red/purple fluorescence [46]. DAPI passes through intact cell membranes, binding to the adenine-thymine (A-T) rich sequences of bacterial DNA [47], staining both living and dead cells blue. In using this double staining method, enumeration of total cell numbers, as well as respiring cell numbers, can be obtained simultaneously under epifluorescence microscopy.

2.5. CTC-DAPI

Staining used in conjunction with DVC is a common method for evaluating metabolic activity in VBNC bacterial cells and works well to detect VBNC Lm with reasonable accuracy [14,36]. CTC-DAPI staining is a precise method for analyzing viable cell counts within Lm biofilms after antimicrobial treatment and does not seem to overestimate viable cell counts when compared to EMA-qPCR [36]. With the CTC-formazan using cellular respiration to break down into a fluorescent dye, no dead cells with intact membranes should be stained. CTC-DAPI staining also gives comparable viable cell counts to DVC, with DVC measuring elongation of viable cells rather than their respiratory activity [48].

2.6. CFDA Stain

CFDA is a colorless fluorogenic ester that enters bacterial cells through diffusion. CFDA is then enzymatically cleaved by bacterial cell esterase enzymes to produce a fluorescent product, carboxyfluorescein (cF). The fluorescence is then analyzed by epifluorescence microscopy or through a flow cytometer; the intensity of the signal determines the viable cell count of the sample [36,50,51]. Due to the enzymatic breakdown of the fluorogenic ester in the dye, only viable cells with intact membranes are counted. Lm viable cell counts, analyzed by epifluorescence microscopy with CFDA, were significantly (p ≤ 0.003) higher when compared to plate counts when analyzing VBNC contamination of cheese [25]. When testing lake water samples using flow cytometry, CDFA produced significantly higher (p = 0.025) viable cell counts compared to control fluorescent labeling methods [52]. When used in conjunction with propidium iodide (PI), CFDA-based flow cytometry is an accurate method for determining viable cell counts. When analyzing the antimicrobial effects of oregano, thyme, and cinnamon essential oils on Lm, clear discrimination between viable cells and membrane-compromised cells was obtained using CFDA and PI [53]. Also, despite a strong reduction in plate counts of Lm, CFDA was retained in 40% of cinnamon oil-treated cells, suggesting that the Lm cells remained metabolically active [53]. CFDA and flow cytometry were used to evaluate the antimicrobial effectiveness of five essential oils against L. innocua [40]. The results suggested that several of the essential oils had permeated the bacterial cytoplasmic membrane and, based on the amount of CFDA fluorescence reduction caused by these oils, showed pronounced antimicrobial activity against the L. innocua [40].