All strains were purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ), except for Enterococcus faecium SF68 which was from the Istituto Superiore di Sanità culture collection. The strains were stored at −80°C in cryogenic vials (Prolab Diagnostics).

L. acidophilus (DSM 20079) and B. bifidum (DSM 20456) were used as reference strains for vPCR. L. delbrueckii subsp. bulgaricus (DSM 20081); L. delbrueckii subsp. lactis (DSM 20072); Lactiplantibacillus plantarum (DSM 20174); Lacticaseibacillus paracasei (DSM 5622); B. animalis subsp. lactis (DSM 10140); B. breve (DSM 20213); Bacillus clausii (DSM 8716); and Streptococcus thermophilus (DSM 20617), which are the most frequently found in LBCPs available in Italy, and E. faecium SF68 were used to confirm the specificity of the primers/probe sets used in qPCR and vPCR.

Bacteria were grown in de Man, Rogosa, Sharpe (MRS) broth or agar (Oxoid), supplemented with 0.05% L-cysteine HCl for culturing the more strictly anaerobic bifidobacteria strains, and incubated at 37°C under aerobic or anaerobic conditions, depending on the bacterial species requirements. Anaerobic jars and gas generating kits (Oxoid) were used to simulate the anaerobic conditions.

Plate count (PC) enumeration of L. acidophilus and B. bifidum in pure broth cultures was performed as previously described (Aureli et al., 2008). Briefly, the test samples were 10-fold diluted in 0.9% saline, and 100 μl of three consecutive dilutions were spread in duplicates on MRS agar or MRS agar supplemented with 0.05% L-cysteine HCl for enumerating L. acidophilus and B. bifidum, respectively. The plates were then incubated at 37°C for 48–72 h under anaerobic conditions.

Subsequently, the total number of L. acidophilus or B. bifidum was enumerated in plates containing 30–300 presumptive colonies, and counts were recorded as colony-forming units (CFU) per milliliter of broth culture. All experiments were repeated three times for each reference microorganism, and the results expressed as mean ± standard deviation.

Genomic DNA was isolated from pure microbial broth cultures grown overnight using the Qiagen DNEasy Blood and Tissue kit, following the manufacturer's instructions. The quality and quantity of isolated DNA were estimated using an ultraviolet spectrophotometer (Biophotometer, Eppendorf). The DNA samples were stored at −20°C until use.

The two primer-probe sets used in separate qPCR reactions to detect L. acidophilus and B. bifidum were selected from the literature (Haarman and Knol, 2006; Singh et al., 2013) and their nucleotide sequences are reported in Table 1. The probes were labeled with the reporter molecule 6-carboxyfluorescein (FAM) and quencher tetramethylrhodamine (TAMRA) at the 5′-end and 3′-end respectively.

The qPCR reaction mixtures (20 μl final volume) were prepared in duplicates and consisted of 10 μl of 2X TaqPath qPCR Master Mix (Applied Biosystems), 1 μl of 20X primer-probe set (Integrated DNA Technologies, IDT), 5 μL of template DNA, and 4 μl of DNAse/RNAse-free water (Bioline). Two control replicates without a DNA template were included in each run.

Real time qPCR amplification was performed in a MicroAmp optical 96-well reaction plates sealed with optical adhesive covers (Applied Biosystems), using a 7,500 Real-Time PCR system (Applied Biosystems). The thermal cycling conditions included pre-incubation at 50°C for 2 min, an incubation step at 95°C for 10 min to activate the AmpliTaq Gold polymerase, 45 cycles at 95°C for 15 s and 60°C for 30 s, and a final incubation step at 60°C for 1 min. The fluorescence signal was measured at the end of each 60°C step. The threshold cycle (Ct) value, corresponding to the PCR cycle number at which fluorescence was detected above the threshold, was calculated using the 7,500 System software (Applied Biosystems). All the above assays were performed twice.

The specificity of the two primer-probe sets for L. acidophilus and B. bifidum was tested using genomic DNA isolated from all the strains described above.

To determine the absolute quantities of L. acidophilus and B. bifidum in unknown samples, the Ct value of each sample was compared to the corresponding standard curves, which were constructed using 10-fold serial dilutions of genomic DNA at known concentrations, isolated from the reference strains L. acidophilus DSM 20,079 and B. bifidum DSM 20,456. The number of microorganisms in the original broth cultures was determined using the PC method and expressed as CFU/ml. The DNA dilutions used for the standard curves were selected to represent at least five bacterial concentrations, ranging from 10² to 10⁷ CFU/ml. The DNA extracts were aliquoted undiluted and stored at −20°C before subsequent single use for standard curve construction.

For each reference strain, live and dead cells in known amounts were prepared as described above and subjected to qPCR, either separately or in different combinations, with or without PMA pretreatment.

Regarding the preparation of bacterial mixtures, aliquots (250 μl) of live cells from broth cultures at OD600 ~ 1 (concentrations of ~10⁸ CFU/ml and ~10⁷ CFU/ml for L. acidophilus and B. bifidum, respectively) were placed in 1.5 ml tubes. Subsequently, equal volumes of dead cells, which were thermally treated under optimal conditions as described above to ensure zero viability, were added to each 1.5 ml tube containing the live cells at decreasing concentrations (i.e., 10⁸, 10⁶, and 10⁴ CFU/ml for L. acidophilus and 10⁷, 10⁵, and 10³ CFU/ml for B. bifidum).

A complementary experiment was performed using the same approach, except that decreasing concentrations of live cells were added at fixed amounts of dead cells.

Individual dead and live cells were used as controls. The cell mixtures and individual cells were treated then with 25 μM PMA as previously described. Two replicates were used for each cell mixture and control.

Finally, DNA was isolated from all samples using a Mag-Bind cfDNA kit and subjected to qPCR. All the experiments were repeated twice for each reference microorganism.

Commercial LBCP capsules containing at least 10⁹ cells of both L. acidophilus and B. bifidum per capsule, according to the product label, was purchased from the market, stored at 4°C, and analyzed within the expiration date. Five LBCP capsules were analyzed. Before analysis, each LBCP capsule was dissolved in 10 ml of a 0.9% NaCl solution.

PC enumeration of L. acidophilus and B. bifidum was performed as described previously. The two microorganisms were differentiated based on colony morphology on MRS agar plates and representative colonies were confirmed at the species level using 16SrRNA sequencing, as previously described (Boye et al., 1999). For B. bifidum, selective counting was also performed on Bifidum Selective Medium (BSM) agar plates (Millipore).

For vPCR, 1 ml of the capsule suspensions were mixed with 25 μM PMA and photoactivated at the conditions above described. A DNeasy kit was used to isolate DNA from 1 ml of PMA-untreated and -treated capsule suspensions. DNA samples were 100-fold diluted and subjected to qPCR using L. acidophilus and B. bifidum specific primer/probe sets in separate reactions. Negative controls without templates were included in each run. Each reaction was performed in duplicate. Serial dilutions of the DNA standards were performed in duplicate for each qPCR run. Concentrations of the individual species were plotted against the corresponding standard curve, with the slope and linear correlation of the curves automatically calculated using the AB 7,500 system software.

Statistical analysis was performed using the GraphPad Prism 10 software (GraphPad Software). Student's t-test and analysis of variance (ANOVA) were used to compare treatment pairs. Differences between treatments were considered statistically significant at p < 0.05.

The specificity of each primer/probe set used in this study has already been assayed, with positive qPCR reactions using L. acidophilus and B. bifidum, and no cross-reactions reported using several non-target bacteria (Haarman and Knol, 2006; Singh et al., 2013). Here, we confirmed the specificity of the primer/probe sets by testing 11 non-target microorganisms other than those previously evaluated (data not shown).

The standard curves of each microorganism showed a strong linear correlation (r² = 0.9983 and r² = 0.9770 for L. acidophilus and B. bifidum, respectively) between the Ct values and cell counts in the tested range (10²-10⁷ CFU/ml) (95% confidence interval), confirming the high accuracy of the qPCR assays. The amplification efficiencies (E) calculated using the formula E = 10^(−1/slope) – 1 (Rasmussen, 2001) were 98% for L. acidophilus and 97% for B. bifidum.

Two of the applied thermal treatments (i.e., at 90°C for 15 min and 100°C for 10 min) efficiently inactivated both L. acidophilus and B. bifidum, as confirmed by the absence of bacterial growth on MRS agar plates following treatments. Although exposures to 75°C and 80°C for 25 min and 20 min, respectively, were also lethal for B. bifidum, they did not ensure 100% mortality of L. acidophilus, as indicated by the growth of a few colonies on solid media.

For both microorganisms, none of the heat treatments modified the qPCR Ct values compared with the corresponding untreated samples (data not shown).

Based on the above results, heat treatment at 100°C for 10 min was selected as the ideal thermal inactivation treatment to ensure zero viability of both target bacteria.

An optimal PMA concentration should allow the exclusive detection of viable microbial cells, while causing the minimal cytotoxic effects.

For selecting the PMA concentration that adequately distinguished between viable and non-viable target bacteria, the Ct values generated from PMA-treated live and dead cells after qPCR were compared to those of the corresponding PMA-untreated controls (Figure 1). For both microorganisms, treatment with 25 μM PMA concentration caused the lowest inhibition of qPCR from viable cells, as indicated by the minimum increase in the Ct value of DNA derived from PMA-treated live cells compared to the PMA-untreated controls (Figures 1A, B); and the highest inhibition from non-viable cells, as deduced by maximum increase in the Ct value of DNA from PMA-treated dead cells compared to the PMA-untreated controls (Figures 1C, D).

Hence, for both L. acidophilus and B. bifidum, the 25 μM PMA concentration allowed better detection of live cells with the lowest interference from dead cells.

Concerning the cytotoxic effects of PMA, the proportion of viable cells of both L. acidophilus and B. bifidum decreased as the PMA concentration increased (Figure 2). Compared to the PMA-untreated controls, the cytotoxic effects of PMA were significant at 50 μM (p = 0.0008) and 100 μM (p = 0.0026) for L. acidophilus (Figure 2A), and 100 μM for B. bifidum (p = 0.0101) (Figure 2B). The 25 μM PMA produced the least cytotoxic effects on both microorganisms, with no significant differences observed compared to the PMA-untreated samples.

Based on the above results, 25 μM PMA was selected to pretreat samples in the vPCR method.

Figure 3 shows the effects of pretreatment using the optimal PMA concentration (25 μM) on live and dead cells of L. acidophilus and B. bifidum, respectively. For both microorganisms, in the absence of PMA pretreatment, the Ct values of DNA derived from live and dead cells were comparable, indicating that DNA was amplified irrespective of cell viability. In contrast, PMA pretreatment resulted in a significant increase in the Ct values of DNA from dead bacteria compared with those from live bacteria (p < 0.0001 for L. acidophilus and p = 0.0210 for B. bifidum). The reduction in the qPCR signal indicated effectively inhibited DNA amplification from dead cells. This result was confirmed when the live and dead cells of both microorganisms were serially diluted (Figures 3A, B).

In Figure 4, the effects of PMA pretreatment on qPCR of DNA derived from dead cells at decreasing concentrations, either alone or in combination with a fixed quantity of live cells, are shown for L. acidophilus and B. bifidum. For both microorganisms, the addition of live cells to different amounts of dead cells always produced significantly lower Ct values (p < 0.0001), as expected, because PMA should allow PCR amplification from live cells while suppressing amplification from dead cells. The fact that, for both microorganisms, the Ct values of all tested dead/live bacterial mixtures were very similar, regardless of the different dead cell concentrations, further demonstrates that DNA was essentially amplified from the live cells present in all mixtures in the same amount (Figures 4A, B).

In the complementary experiment, for L. acidophilus the addition of a fixed quantity of dead cells to decreasing numbers of live cells followed by PMA pretreatment did not affect the Ct values compared to those of the live cells alone, confirming that PMA efficiently inhibited qPCR amplification from the dead cells. However, inhibition of DNA amplification by PMA treatment was not evident at the lowest L. acidophilus live cell concentration tested at 10² CFU/ml, as they generated a significantly higher Ct value compared to that of the relative mix with dead cells (p = 0.0055) (Figure 5A).

For B. bifidum, PMA inhibition of qPCR from dead cells was apparent when live cells in the live/dead cell mixtures were present at relatively high levels (i.e., 10⁷ CFU/ml and 10⁵ CFU/ml); in fact, at these concentrations there was no significant difference between the Ct values derived from live cells alone and the live/dead cell mixtures (Figure 5B). As for L. acidophilus, PCR inhibition by PMA pretreatment was not observable when the B. bifidum live cells quantities in the live/dead cell mixtures decreased ( ≤ 10³ CFU/ml), as demonstrated by the significant increase in the Ct values of the live cells alone compared to those of the relative mixtures with dead cells (p < 0.0001) (Figure 5B). This result could be due to the fact that, when the live cell concentration decreased in the mixtures containing high dead cell ratios, the 25 μM PMA pretreatment reduced the amplification from the viable cells while not fully inhibiting amplification from the dead cells, in accordance with what observed by other authors (Papanicolas et al., 2019).

Thus, our overall data indicate that the live cells quantification limits for the proposed vPCR assay, consisting of qPCR preceded by 25 μM PMA treatment, were approximately 10² CFU/ml for L. acidophilus and 10³ CFU/ml for B. bifidum.

Figure 6 shows L. acidophilus and B. bifidum contents in a commercial LBCP, as determined by PC, qPCR, and the newly developed vPCR protocol.

Plate counts confirmed the presence of at least 10⁹ CFU/capsule of L. acidophilus, which was consistent with the declared labeled amount for that microorganism. For B. bifidum, whose stated label claim was also ≥10⁹ CFU/capsule, both counts on MRS agar supplemented with cysteine and on BSM agar plates yielded ~10⁸ CFU/capsule (Figure 6). Because B. bifidum is a “fastidious” microorganism to grow, being strictly anaerobic and nutrient-demanding, its concentration may have been underestimated using PC (Modesto, 2018). The production of inhibitory substances by L. acidophilus (the other microorganism present in the product formulation), such as organic acids and bacteriocins, or competition for nutrients on agar plates may also have contributed to the quantitative inconsistency between the product label and PC results for B. bifidum. Alternatively, it is possible that VBNC B. bifidum cells were present in the test samples and escaped culture detection.

The presence of VBNC microbial cells might also account for the significantly lower quantitative values obtained for both microorganisms using PC compared to those estimated using qPCR and vPCR (all p values < 0.001) (Figure 6): in fact, while VBNC microbial cells fail to grow in culture media, their DNA can be amplified using PCR.

The quantity estimates of both bacteria using vPCR were lower than those determined using qPCR, as expected, since amplification of DNA from dead cells in the samples should be prevented by the PMA pretreatment step of the vPCR protocol. However, the lack of a statistically significant difference between the quantitative values obtained using qPCR and vPCR for both microorganisms suggests that the tested LBCP contained a few dead cells (Figure 6).