The Copy Number of the spoVA^2mob Operon Determines Pressure Resistance of Bacillus Endospores

Bacillus spp. are considered pressure-resistant microorganisms, but the resistance mechanisms remain unknown. The spoVA^2mob operon is a mobile genetic element, and it can transfer to pathogenic or spoilage organisms by horizontal gene transfer. Results in this study indicate that multiple copies of the spoVA^2mob operon mediate high-pressure resistance of Bacillus endospores, and it might contribute to the identification of the source of pressure-resistant pathogens and spoilage organisms that may contaminate the food supply. The droplet digital PCR (ddPCR) system is well suited for analysis in some human diseases due to its high efficiency and capability to provide high precision; however, no relevant studies in food microbiology have been reported so far. This study demonstrates a novel application of ddPCR in food microbiology.

ABSTRACT

The spoVA^2mob operon confers heat resistance to Bacillus spp., and the resistance correlates to the copy number of the operon. Bacillus endospores also exhibit a strong variation in resistance to pressure, but the underlying mechanisms of endospore resistance to pressure are not fully understood. We determined the effects of multiple spoVA^2mob operons on high-pressure resistance in Bacillus endospores. The copy numbers of the spoVA^2mob operon in 17 strains of Bacillus subtilis, Bacillus amyloliquefaciens, Bacillus cereus, Bacillus velezensis, and Bacillus pumilus were determined via droplet digital PCR (ddPCR) and genome sequencing. These strains contained between 0 and 3 copies of the spoVA^2mob operon; the quantification of the gene copy number by ddPCR was as accurate as whole-genome sequencing. We further tested the pressure resistance of 17 Bacillus endospores at 600 MPa and 80°C. Strains with one or no spoVA^2mob operon had significantly lower pressure resistance than strains with two or three copies of the operons (P < 0.001), indicating that redundant spoVA^2mob operons in Bacillus contributed to higher pressure resistance of endospores. The copy number of the spoVA^2mob operon was not related to the dipicolinic acid (DPA) content of endospores. Overall, the copy number of the spoVA^2mob operon contributes to pressure resistance of Bacillus endospores. This improves our understanding of the pressure resistance mechanisms in Bacillus spp. and may inform the development of high-pressure sterilization in food processing.

IMPORTANCE Bacillus spp. are considered pressure-resistant microorganisms, but the resistance mechanisms remain unknown. The spoVA^2mob operon is a mobile genetic element, and it can transfer to pathogenic or spoilage organisms by horizontal gene transfer. Results in this study indicate that multiple copies of the spoVA^2mob operon mediate high-pressure resistance of Bacillus endospores, and it might contribute to the identification of the source of pressure-resistant pathogens and spoilage organisms that may contaminate the food supply. The droplet digital PCR (ddPCR) system is well suited for analysis in some human diseases due to its high efficiency and capability to provide high precision; however, no relevant studies in food microbiology have been reported so far. This study demonstrates a novel application of ddPCR in food microbiology.

INTRODUCTION

Endospores produced by Bacillus are resistant to heat, UV irradiation, desiccation, and chemical assaults (1). They survive in nonthermal or insufficiently heated processed foods (1). In addition, heat-resistant spores may survive thermal processes that are currently used to achieve commercial sterility (2). High temperatures also negatively impact texture, flavor, and nutrient content in a variety of food products (3). Accordingly, nonthermal food processing technologies have been developed to avoid the disadvantages of traditional thermal treatments (4).

High-pressure processing is among the commercially most successful nonthermal processing technologies; pressure processes are used for the inactivation of vegetative microbial cells (4, 5). Pressure-assisted thermal sterilization to inactivate spores by using compression heating for rapid and uniform heating of the food product is still in the development stage. The mechanisms of spore inactivation by pressure are not fully understood yet and warrant further investigation (6). Spore inactivation at high pressure and ambient temperature activates physiological processes in spore germination, but spore inactivation by pressure-assisted thermal sterilization is unrelated to enzymatic or physiological enzymes (6–8). Bacillus and Clostridium spp. exhibit a large intraspecies variation with regard to the spore resistance to pressure (9–11). Spores of Clostridium botulinum group I (or C. parabotulinum [12]) are among the most pressure-resistant bacterial spores, and strains of Bacillus amyloliquefaciens have been used as pressure-resistant surrogate organisms for C. botulinum (8).

The heat resistance of bacilli correlates to the copy number of the spoVA^2mob operon (13–15). The spoVA^2mob operon is present in Bacillus spp., including B. amyloliquefaciens, B. subtilis, B. cereus, B. licheniformis, and B. thermoamylovorans (2, 14, 16). Bacillus strains carrying multiple copies of the spoVA^2mob operon exhibit decimal reduction times (D values) that are up to 100-fold higher than strains of the same species that do not carry this operon (13, 14). The pressure resistance of bacterial spores does not generally correlate to their heat resistance; however, some spore properties, particularly the content of dipicolinic acid (DPA) and water, relate to their resistance to pressure and heat (6, 11). The resistance of spores to pressure-assisted thermal sterilization may depend on the spores’ ability to retain DPA and on the heat resistance of the DPA-free spores (6, 11). The likely mechanisms for spore inactivation at more than 600 MPa and more than 80°C are a compromised inner spore membrane and the inactivation of integral membrane proteins (17–19). This therefore facilitates Ca-DPA release in the absence of physiological activity (6, 20). Proteins encoded by the regular spoVA operon, SpoVAC, SpoVAD, and SpoVAEB, relate to DPA uptake during sporulation and release during spore germination (21, 22). Therefore, the copy number of spoVA^2mob operon per genome may also correlate to the spore DPA content and pressure resistance. Nevertheless, the potential link between the copy number of the spoVA^2mob operon and spore pressure resistance has not been reported.

This study therefore aimed to determine the relationship between copy number of the spoVA^2mob operon and pressure resistance of Bacillus endospores. To quantify the copy number of the spoVA^2mob operon in Bacillus spp., a fast and accurate droplet digital PCR (ddPCR) assay for the quantification of the absolute gene copy number was developed and compared to whole-genome sequencing as reference method to assess the copy number variation (23, 24). PCR-based methods are also applied as rapid methods for gene copy number determination (25). The ddPCR system has been utilized as a method that quantifies absolute copy numbers of nucleic acids with high accuracy and precision (26, 27). In the ddPCR system, an absolute target sequence is quantified from the ratio of positive and negative partitions estimated by the Poisson distribution (28) without the need for external calibration (29–31). Reports on the use of ddPCR to study the number of gene copies in bacterial chromosomes, however, are scarce.

RESULTS

Variation of the copy number of the spoVA^2mob operon determined by ddPCR.

To obtain strains of Bacillus spp. differing with respect to the copy numbers of the spoVA^2mob operon, we selected 17 isolates from spoiled foods, fermented foods, and the environment. Nine strains of Bacillus were previously isolated from ropy bread and resisted pressure treatment (11, 32), two strains were isolated from malted oats, one strain originated from the cow vagina (33), and five strains were from Daqu, a solid-state cereal fermentation (14, 34). The presence or absence of the spoVA^2mob operon was determined by PCR amplification of the last gene of the spoVA^2mob operon. Because all spoVA^2mob-positive B. cereus strains for which genome sequences are available in GenBank possess a 3′ truncated version of the operon, a second pair of primers was used to confirm that the 3′ end of the operon was present in B. cereus FUA2120. The copy numbers of the spoVA^2mob operon in these 17 Bacillus strains were determined by ddPCR (Table 1). All strains with 2 or 3 copies of spoVA^2mob belong to B. amyloliquefaciens or B. subtilis; neither of the two B. velezensis strains, which are closely related to B. amyloliquefaciens (35), carried the spoVA^2mob operon. One strain of B. cereus carried a single copy of the operon (Table 1).

TABLE 1.

Origin of 17 Bacillus spp. and copy number of spoVA^2mob operon in each strain

Organism	Strain and origin	ddPCR-experimentalb (theoretical) copy no. of spoVA2mob operon/genome	CNOGpro copy no. of spoVA2mob operon/genome (GenBank accession number)	Reference
B. velezensis	Fad 94, ropy bread	0.09 ± 0.06 (0)	0 (SDKF00000000)	32
B. pumilus	FUA2024,a cow vagina	0 (0)		33
B. subtilis	FUA2114, malted oats	0 (0)	0 (VRTW00000000)	Unpublished
B. velezensis	FUA2155, Daqu	0.24 ± 0.01 (0)	0 (SDKI00000000)	14
B. cereus	FUA2120, malted oats	1.19 ± 0.10 (1)		Unpublished
		1.09 ± 0.30 (1)c
B. subtilis	FUA2148, Daqu	1.24 ± 0.11 (1)		14
		1.27 ± 0.43 (1)
B. amyloliquefaciens	FUA2149, Daqu	1.18 ± 0.28 (1)		14
		1.11 ± 0.18 (1)
B. subtilis	Fad 110, ropy bread	2.10 ± 0.16 (2)		32
		1.93 ± 0.37 (2)
B. subtilis	Fad 109, ropy bread	1.94 ± 0.31 (2)	2 (SDKE00000000)	32
		1.92 ± 0.51 (2)
B. amyloliquefaciens	Fad 108, ropy bread	1.73 ± 0.08 (2)		32
		1.72 ± 0.06 (2)
B. amyloliquefaciens	Fad 11/2, ropy bread	2.14 ± 0.21 (2)		32
		1.71 ± 0.47 (2)
B. amyloliquefaciens	FUA2153, Daqu	1.92 ± 0.11 (2)	2 (VRTV00000000)	14
		1.94 ± 0.13 (2)
B. amyloliquefaciens	FUA2154, Daqu	2.08 ± 0.18 (2)	2 (VRTU00000000)	14
		2.02 ± 0.34 (2)
B. amyloliquefaciens	Fad We, ropy bread	3.06 ± 0.18 (3)	3 (VRTX00000000)	32
		2.64 ± 0.17 (3)
B. amyloliquefaciens	Fad 97, ropy bread	2.95 ± 0.23 (3)		32
		2.81 ± 0.27 (3)
B. amyloliquefaciens	Fad 99, ropy bread	3.17 ± 0.15 (3)	3 (SDKH00000000)	32
		3.01 ± 0.31 (3)
B. amyloliquefaciens	Fad 82, ropy bread	2.69 ± 0.12 (3)	3 (SDKG00000000)	32
		2.35 ± 0.21 (2)

^aFUA number, Food Microbiology culture collection at the University of Alberta.

^bMean ± standard deviation for triplicates.

^cUnderlined data were obtained from DNA extractions applying bead beating.

Variation in ddPCR data with DNA extractions from two different methods.

Multiple copies of a gene are accounted for as a single copy in ddPCR assays if they are on the same DNA molecule. Therefore, the fragmentation of DNA during DNA isolation may influence the results. We employed bead beating as a method that produces highly fragmented DNA and a genomic DNA isolation kit that allows isolation of high-molecular-weight genomic DNA. For most strains, the two methods of DNA extraction provided the same copy numbers of the spoVA^2mob operon. For B. amyloliquefaciens Fad 82, the copy number determined after DNA isolation with bead beating was 3 spoVA^2mob copies per genome, while the corresponding result of analysis of high-molecular-weight DNA was 2 copies per genome (Table 1).

Variation of the copy number of the spoVA^2mob operon determined by genome sequencing and CNOGpro.

The genomes of 9 strains of Bacillus were sequenced to validate the ddPCR assay. Genome sequencing confirmed the absence of the spoVA^2mob operon in B. velezensis Fad 94 and FUA2155 as well as B. subtilis FUA2114. The genomes of B. subtilis Fad 109 and B. amyloliquefaciens FUA2153 and FUA2154 each harbored 2 spoVA^2mob copies; 3 copies of the spoVA^2mob operon were identified in B. amyloliquefaciens Fad 99, Fad 82, and Fad We, matching the results obtained by ddPCR (Table 1).

Variation in pressure resistance of Bacillus strains.

To determine the pressure resistance of spores from the 17 Bacillus strains, spores were subjected to treatment at 600 MPa and 80°C (Fig. 1). A high variation in pressure resistance was observed among the 17 strains, and their pressure resistance levels correlated with the copy numbers of the spoVA^2mob operon (Fig. 1). Viable spore counts of four strains lacking the spoVA^2mob operon were reduced by 6 to 8 log CFU/ml after 4 min. The pressure resistance of the three strains with a single copy of the spoVA^2mob operon was variable. B. subtilis FUA2148 and B. amyloliquefaciens FUA2149 were more resistant to pressure than B. cereus FUA2120. All 6 strains containing 2 copies of the spoVA^2mob operon, two strains of B. subtilis and four strains of B. amyloliquefaciens, had a similar resistance to high pressure. Pressure treatment for 8 min reduced viable spore counts by between 1 and 3 log CFU/ml. Inactivation by less than 2 log CFU/ml was observed for the corresponding treatment of the 4 strains of B. amyloliquefaciens with 3 copies of the spoVA^2mob operon.

FIG 1.

Viable spore counts [log(N/N₀)] of 17 Bacillus strains after treatment at 600 MPa and 80°C. The x axis represents treatment time at 600 MPa and 80°C. Strains either lacked a spoVA^2mob operon copy (4 strains) (A) or contained 1 (3 strains) (B), 2 (6 strains) (C), or 3 (4 strains) (D) copies of the spoVA^2mob operon per genome. Species are differentiated by different symbols as follows: △, B. velezensis; ∇, B. pumilus; □, B. subtilis; ◊, B. cereus; ○, B. amyloliquefaciens. Data are shown as means ± standard deviations from three independent experiments.

Variation of spore DPA content.

Because SpoVA proteins were linked to the DPA content of spores (21), we determined whether the copy number of the spoVA^2mob operon was associated with the DPA content in spores, which is a significant factor of spore pressure resistance. The spore DPA contents differed among strains but were not related to the copy number of the spoVA^2mob operon (Fig. 2).

FIG 2.

DPA contents of Bacillus spores. Species are differentiated by different symbols as follows: △, B. velezensis; ∇, B. pumilus; □, B. subtilis; ◊, B. cereus; ○, B. amyloliquefaciens. White, light gray, dark gray, and black symbols represent strains with 0, 1, 2, and 3 copies of spoVA^2mob operon per genome. The data show results from three independent experiments; error bars refer to the standard deviations.

Variation of spore DPA released after pressure treatment.

Pressure treatment of Bacillus endospores relates to DPA release, which is observed at around the same time or earlier than spore inactivation (11). To evaluate the correlation between the spoVA^2mob operon and DPA release, we used 4 strains with different copy numbers of the spoVA^2mob operon. After treatment for 1 and 2 min at 600 MPa and 80°C, the percentage of DPA released from the spores with 2 or 3 copies was lower than for the spores containing 0 or 1 copy of the spoVA^2mob operon (Fig. 3). After treatment for 1 min, the spores with 0 or 1 copy of the spoVA^2mob operon released more than 84% of DPA, whereas the spores with 2 or 3 copies released around 65% to 72% of DPA. When treated for 4 and 8 min, all of the spores released more than 85% of DPA.

FIG 3.

Correlation of viable spore counts [log(N/N₀)] after 8 min at 600 MPa and 80°C and spoVA^2mob copy number. Species are differentiated by different symbols as follows: △, B. velezensis; ∇, B. pumilus; □, B. subtilis; ◊, B. cereus; ○, B. amyloliquefaciens. White, light gray, dark gray, and black symbols represent strains with 0, 1, 2, and 3 copies of spoVA^2mob operon per genome, respectively. Mean values for the same copy number of spoVA^2mob operon (symbols with same color) are highly significantly different (P < 0.001).

Correlation of spoVA^2mob operon copy number and endospore resistance to pressure.

Correlation of the spoVA^2mob operon copy number and the pressure resistance was assessed by comparing the cell counts of strains after 8 min of treatment at 600 MPa and 80°C. Spores with 0 or 1 spoVA^2mob operon were less resistant to pressure than spores with 2 or 3 copies (P < 0.001), but the pressure resistance of strains carrying 0 and 1 copy of the spoVA^2mob operon, or 2 and 3 copies of the operon, was not different (Fig. 4). When the analysis excluded B. cereus, the species that is phylogenetically most distant from other species used in this study, spores of strains with a single copy of the spoVA^2mob operon exhibited intermediate resistance compared to strains with 0 or 2 copies (Fig. 1).

FIG 4.

DPA release amounts of Bacillus spores after high-pressure treatment at 600 MPa and 80°C and autoclaving at 121°C for 60 min. Strains are differentiated by different colors as follows: white, B. velezensis Fad 94 with 0 copies; light gray, B. amyloliquefaciens FUA2149 with 1 copy; dark gray, B. amyloliquefaciens Fad 99 with 2 copies; black, B. amyloliquefaciens Fad 108 with 3 copies of the spoVA^2mob operon per genome.

DISCUSSION

This study reports the impact of the spoVA^2mob operon on pressure resistance of Bacillus endospores. ddPCR rapidly quantified the gene copy number of the spoVA^2mob operon in bacilli, and results obtained by ddPCR were validated by whole-genome sequencing of 9 of the 17 isolates. Spores containing 2 or more copies of the spoVA^2mob operon showed a higher level of pressure resistance than spores with 0 or 1 copy of the operon.

We used 17 wild-type Bacillus strains isolated from spoiled foods, fermented foods, and the environment to demonstrate the impact of the copy number of the spoVA^2mob operon on pressure resistance. Past studies inserted a single copy of the spoVA^2mob operon into the cloning strain B. subtilis 168 (13, 15, 16) but did not report insertion or deletion of multiple copies of the operons in wild-type strains, which are difficult to transform owing to copious exopolysaccharide production and the presence of multiple endogenous plasmids in wild-type strains. The use of a large number of closely related wild-type strains differing with respect to the copy number of the spoVA^2mob operon provides equivalent convincing conclusions with regard to the relationship between copy number and heat resistance (2, 14) and has the advantage of demonstrating that the copy number of the spoVA^2mob operon is far more relevant for heat and pressure resistance than other strain- or species-specific properties.

Comparison of ddPCR and other methods for detecting gene copy number.

Compared to whole-genome sequencing, the reference method for the determination of the copy number of a gene in a bacterial genome, PCR methods, including quantitative PCR (qPCR) and droplet digital PCR, offer the advantage of rapid and high-throughput analysis of multiple samples. qPCR has been widely used for the quantification of nucleic acids (36), but the accuracy of qPCR is limited because the qPCR quantification requires external standards (26). The accuracy and precision of copy number estimates detected by qPCR may differ, because factors such as the physicochemical state of DNA can affect the efficiency of qPCR and consequently the cycle threshold value (26, 37). ddPCR more precisely determines the copy number, because the target gene is directly compared to a reference gene which is analyzed in the same PCR (27). The copy number of the spoVA^2mob operon in the 5 Daqu isolates was previously analyzed by qPCR (14). When matched against genome sequencing as a reference, ddPCR analyses provided a higher accuracy and precision than qPCR (13 and Table 1).

The copy number of a gene in a bacterial genome is an integer; however, ddPCR did not provide integers. The deviation is partially caused by experimental error. In addition, if two copies of a gene are located on the same DNA molecule, they are accounted for as a single copy in the ddPCR assay. Accordingly, cell lysis by bead beating, which shears DNA to fragments of less than 10 kbp, provided a higher copy number than a gentler isolation method that retains DNA strands up to 50 kbp. To validate the ddPCR data, we further used CNOGpro to analyze the copy number of the spoVA^2mob operon by the sequencing of 9 Bacillus strains, corroborating the ddPCR result. Using ddPCR to detect copy numbers thus has a comparable accuracy as whole-genome sequencing but offers the advantage of rapid analysis of multiple samples.

Relationship between the copy number of the spoVA^2mob operon and pressure resistance of Bacillus endospores.

A comparison of the pressure resistance of multiple strains of B. amyloliquefaciens and B. subtilis revealed a high intraspecies variation in pressure resistance; in particular, isolates from ropy bread were highly resistant to pressure (11). These strains were subsequently used as pressure-resistant surrogate organisms for C. botulinum in multiple studies (8). Our data on pressure resistance of B. amyloliquefaciens and B. subtilis confirm prior data obtained with the same strains; moreover, our data demonstrate that all of the rope-forming and pressure-resistant strains of B. amyloliquefaciens and B. subtilis contained 2 or 3 copies of the spoVA^2mob operon. The copy number variation thus explains most of the difference in pressure resistance between the rope-forming B. amyloliquefaciens and B. subtilis strains and other Bacillus spp. (11). It is possible that other Bacillus species with redundant spoVA^2mob operons show higher pressure resistance, but isolates are not available (this study, 11, 12).

Surprisingly, spores of B. cereus FUA2120 with a single copy of the spoVA^2mob operon were relatively pressure sensitive, particularly when resistance was assessed after 1 min of treatment at 600 MPa and 80°C. We further tested the heat resistance of this strain at 100°C for 0, 1, 2, and 4 min, showing that this strain was also heat sensitive (see Fig. S1 in the supplemental material). B. subtilis, B. amyloliquefaciens, and B. velezensis are phylogenetically closely related; in particular, B. velezensis and B. amyloliquefaciens diverge just below the species cutoff of 95% average nucleotide identity (35). In contrast, Bacillus pumilus and B. cereus are more distant to other Bacillus spp. analyzed in this study (38). Spore properties unrelated to the spoVA^2mob operon that are shared by spores of the B. subtilis group but not by B. cereus may thus account for this discrepancy. Of the 1,019 genomes of strains of B. cereus that were available on GenBank, 19 contained a truncated version of the spoVA^2mob operon, but none carried the full operon. PCR analysis targeting sequences close to the 3′ and 5′ ends of the operon confirmed, however, that B. cereus FUA2120 contains the full-length spoVA^2mob operon. The low frequency of the spoVA^2mob operon and the relatively low heat and pressure resistance of the spoVA^2mob-positive B. cereus FUA2120 suggest that the operon may be occasionally acquired from strains of the B. subtilis group but does not provide a competitive advantage to B. cereus.

Potential mechanism of spoVA^2mob operons influencing endospore germination during or after pressure treatment.

The spoVA^2mob operon decreased the release of Ca-DPA through SpoVA channels in the inner membrane and slowed the germination of spores (15, 16). Slower germination is likely caused by the inner membrane proteins SpoVAC, SpoVAD, and SpoVAEB, which are encoded by the regular spoVA operon (14, 15, 20). These proteins mediate DPA uptake during sporulation as well as Ca-DPA release during germination (1, 15, 22). It remains unclear, however, whether the presence of multiple copies of the spoVA^2mob operon incrementally impacts DPA uptake and release. The present study demonstrated a role of the copy number of the spoVA^2mob operon in pressure resistance and pressure-induced DPA release. SpoVAC^2mob, SpoVAD^2mob, and SpoVAEB^2mob share 55%, 49%, and 59% amino acid identity with the respective proteins encoded in the regular spoVA operon and attenuated spore germination (16).

The mechanisms of spore resistance to pressure depend on the pressure level that is applied, and the spore response to processes used for pressure-assisted thermal sterilization is unrelated to physiological germination (6, 11, 17). Pressure of less than 300 MPa at ambient temperature induces spore germination via the nutrient-triggered germination pathway by activating the nutrient germinant receptors (6, 39–41). A pressure of 400 to 800 MPa at ambient temperature triggers release of Ca-DPA to induce germination independent of nutrient receptors (6, 7). Treatment at 600 MPa and 80°C or higher inactivates spores without triggering physiological processes involved in spore germination (6, 8, 19, 41). Previous studies hypothesized that the insertion of spoVA^2mob operon into B. subtilis 168 resulted in higher DPA concentrations in spores and consequently a high-level heat resistance (13, 16). Our results, however, do not confirm a correlation of the copy number of the spoVA^2mob operon with the DPA concentration of spores but suggest that the spoVA^2mob operon reduced the pressure-induced but germination-independent DPA release.

In conclusion, this study demonstrated that the pressure resistance of Bacillus spores is dependent on the copy number of the spoVA^2mob operon. Some of the strains used in this study were previously identified as pressure-resistant surrogates for validation of the lethality of pressure-assisted thermal pasteurization against C. botulinum (8). Our study contributes to the understanding of pressure resistance at the molecular level; it provides a rationale why rope-forming bacilli exhibit high-pressure resistance (10, 29); thus, the spoVA^2mob operon can be used as a genetic marker to identify the risky spores and improve risk assessments in actual food production. This progress enables further developments of pressure-assisted thermal sterilization in the food industry.

MATERIALS AND METHODS

Bacterial strains, preparation of spore suspensions, and culture conditions.

The Bacillus strains used in this study and their origins are listed in Table 1. All strains were grown aerobically on Luria-Bertani (LB) agar (Difco agar; BD Biosciences, Franklin Lakes, NJ, USA) plates at 37°C for 18 h. Spores were prepared by plating aliquots of 0.1 ml from fresh overnight cultures on LB agar plates. Spore suspensions were obtained after 5 days of incubation at 37°C when 90% to 99% of the cells contained spores. Sporulation was assessed microscopically after staining with the Schaeffer-Fulton method (malachite green solution of 5% and 0.1% safranin) which results in green staining of spores and red staining of vegetative cells (40). Spore suspensions were obtained as described elsewhere (11). In short, the surface of the plate was flooded with 5 ml cold sterile distilled water twice to collect the spores. After being harvested, the spore suspensions were washed four times by centrifugation at 3,000 × g for 15 min at 5°C and resuspended in sterile distilled water. To kill all the vegetative forms, the spore suspensions were pasteurized at 80°C for 10 min between the second and third wash cycles. The spore suspensions were stored at −80°C until used. Cell counts were determined on LB agar. Appropriate dilutions were surface-plated on LB agar, and plating was carried out in triplicate from independent proper dilution series. The plates were incubated aerobically for 18 h at 37°C. CFU values were calculated using Microsoft Excel (Microsoft Corp., Redmond, WA, USA).

Extraction of DNA and restriction digestion of the DNA sample.

We applied two methods for DNA extraction. The first method was according to the protocol of the Wizard Genomic DNA purification kit (Promega Corp., Madison, WI, USA). In the second method, prior to following the protocol of the kit, approximately 1 ml of overnight culture of each strain was placed into a 2-ml microcentrifuge tube filled with 0.5 g of silica beads. The samples were smashed by bead beating for 30 s, repeating 8 times. The concentration of DNA was analyzed by a NanoDrop One Microvolume UV-visible (UV-Vis) spectrophotometer (Thermo Fisher Scientific).

Genomic DNA samples were further treated by Fast Digest restriction enzyme HindIII (Thermo Fisher Scientific) at 37°C for 45 min with subsequent heat kill at 65°C for 10 min to cut the intact genomic DNA into fragments and separate linked copies of the spoVA^2mob operon efficiently. Then, DNA samples were stored at −20°C until further use. Prior to ddPCR analysis, DNA was encapsulated into droplets (42, 43).

Primer and probe design.

In this study, we applied TaqMan hydrolysis probes as the reporter fluorophores. Each ddPCR mixture contained duplex TaqMan probes for the region of interest (ROI), the last gene on the spoVA^2mob operon, and the region of reference (REF), DNA gyrase subunit B (gyrB). In this study, we chose the gyrB gene as the REF gene, because this gene is highly conserved and a single-copy housekeeping gene (44). We extracted several nucleotide sequences of Bacillus strains from the National Center for Biotechnology Information (NCBI) database (see Table S1 in the supplemental material) and aligned them with the MUSCLE website (https://www.ebi.ac.uk/Tools/msa/muscle/). Afterwards, we examined the aligned sequences to determine conserved regions suitable as targets. Primers and probes were first designed using IDT PrimerQuest (Integrated DNA Technologies, Inc., Coralville, IA) and then adjusted manually following the instructions of the Droplet Digital PCR application guide (Bio-Rad Laboratories, Inc., Hercules, CA). Primers and probes are shown in Table 2.

TABLE 2.

Sequences of primers and probes

Gene	Primer or probe	Sequence (5′→3′)	Tma (°C)	Product size (bp)
Last gene of spoVA2mob operon	spoVA2mob-F	AACCACTAGCCACGATTG	59	169
	spoVA2mob-R	AAGGGTCTTTCTTGTGGG	59
	spoVA2mob-probe	/FAMb /ACGAAGTCGGGCTTGGCTACA/	68
gyrB	gyrB-F	ATCGTCGACAACAGTATTG	57	205
	gyrB-R	CTTTATATCCGCTTCCGTC	57
	gyrB-probe	/HEXc /CCCTGCGGTTGAAGTCATCATGA/	66
Front part of spoVA2mob operon	spo-F	AAGGTCGAGCAAAGACTG	59	1,516
	spo-R	ACCTGTAGCCACAACTAAC	59

^aT_m, melting temperature.

^bFAM, 6-carboxyfluorescein.

^cHEX, reporter dye.

Quantification of the copy number of the spoVA^2mob operon by ddPCR.

All experiments were performed using a QX200 Droplet Digital PCR system (Bio-Rad). Each ddPCR mixture consisted of 12.5 μl of 2× ddPCR SuperMix for probes (no dUTP) (Bio-Rad), 100 pg of template DNA, 840 nM each forward and reverse primers of gyrB, 720 nM each forward and reverse primers of spoVA^2mob, and 400 nM both probes. The final volume (25 μl) was attained with nuclease-free water. The entire reaction mixture was loaded into a DG8 cartridge (Bio-Rad) together with 70 μl of droplet generation oil (Bio-Rad) and placed in the QX200 droplet generator (Bio-Rad). Approximately 20,000 droplets were then generated in the droplet generator. The generated droplets were transferred to a new 96-well PCR plate, and the plate was subjected to amplification in a C1000 Touch thermal cycler (Bio-Rad). The thermal cycling conditions consisted of a 5-min activation period at 95°C, followed by 40 cycles at 95°C for 20 s and 56°C for 60 s, 1 cycle at 94°C for 10 min, and ending at 4°C. After amplification, fluorescence was detected by the QX200 droplet reader (Bio-Rad) with the following settings: ROI was detected in the 6-carboxyfluorescein (FAM) channel, and REF was detected in the 6-carboxy-2,4,4,5,7,7-hexachlorofluorescein (HEX) channel. The raw fluorescence was analyzed with QuantaSoft software (Bio-Rad). In this experiment, positive droplets which contained at least one copy of the spoVA^2mob operon exhibited a fluorescence not evident for negative droplets without the spoVA^2mob operon. The copy numbers of the ROI gene were calculated by multiplying the known copy number of the REF gene by the ratio of the ROI and REF concentrations.

Quantification of the copy number of the spoVA^2mob operon by genome sequencing and CNOGpro.

The genomes of B. subtilis Fad 109, B. amyloliquefaciens Fad 82 and Fad 99, and B. velezensis Fad 94 and FUA2155 were sequenced using a HiSeq 2500 system (Illumina) on high output. Sequencing was performed as a fee for service by Genome Quebec (Montreal, QC, Canada). B. subtilis FUA2114, B. cereus FUA2120, and B. amyloliquefaciens FUA2153, FUA2154, and Fad We were sequenced by Illumina next-generation sequencing by MicrobesNG (Birmingham, United Kingdom). The accession number of each strain is listed in Table 1. The 125-bp reads were assembled using SPAdes (45), and the resulting contigs were scaffolded with the Medusa server (46). Afterwards, the scaffold containing the spoVA^2mob was subjected to read depth analysis using CNOGpro (47), and the copy number for the region of the scaffold containing the spoVA^2mob operon was determined.

Determination of pressure resistance of Bacillus endospores.

Spores were diluted in sterile distilled water in order to obtain the working spore suspension with a cell count of approximately 5.0 × 10⁷ to 4.7 × 10⁸ CFU/ml. We transferred 123 μl of spore suspension into a polypropylene tube (Fisher Scientific), which was sealed on both sides, avoiding inclusion of air bubbles. The samples were treated with an MP5 high-pressure micropump system (High Pressure Physics, Polish Academy of Sciences, Warsaw, Poland) at 600 MPa and 80°C. The pressure vessel was immersed in a water bath that was maintained at 80°C. The temperature profiles for treatments at 600 MPa and 80°C are shown in Table 3. Adiabatic heating during compression to 600 MPa increased the temperature in the pressure vessel by less than 6.5°C. Samples were held on ice before and after pressure treatment (8). Viable spore counts were determined after surface plating on LB agar. Surface plating of pressure treatment underestimates surviving spores compared to that with MPN techniques but nevertheless allows the comparative assessment of spore resistance (11).

TABLE 3.

Pressure-temperature profiles for treatment at 600 MPa and 80°C

Process stage	Time	Pressure (MPa)	Temp (°C)
Compression	0	0	79.0
	42 s	100	84.9
	65 s	200	85.3
	82 s	300	85.5
	96 s	400	85.6
	108 s	500	85.7
	118 s	600	84.6
Pressure holding time	1/2/4/8 min	600	80.6
Decompression	0	600	80.6
	114 s	0	73.3

Determination of spore DPA content by fluorescence spectrometry.

To evaluate the amount of DPA present in the spores of the individual bacterial species, 2 ml of each spore suspension was standardized to an optical density at 600 nm (OD₆₀₀) of 2 and autoclaved for 60 min at 121.1°C to inactivate spores and ensure complete DPA release (48). A terbium-DPA fluorescence assay was carried out as described previously (48, 49). Briefly, untreated as well as autoclaved spore suspensions were centrifuged at 10,600 × g for 4 min, and 150 μl of the supernatant was mixed with 150 μl of 20 mM terbium(III) chloride solution in nontransparent 96-well microtiter plates (Corning 96-well NBS Microplate; Corning, Inc., Corning, NY, USA). Fluorescence intensity was determined using a fluorescence spectrophotometer (Varioskan Flash; Thermo Electron Corp., Nepean, ON, Canada) with excitation and emission wavelengths of 270 and 545 nm, respectively. A calibration standard curve was recorded, using analytical-grade DPA (Sigma-Aldrich, St. Louis, MO, USA) in the range of 0 to 150 μM, and linearity was ensured (R² > 99.6%). All measurements were carried out in quadruplicate technical repeats from three independent spore suspensions. DPA release was expressed as the difference between the DPA concentration prior and subsequent to autoclaving. Furthermore, spore enumeration was performed by serial dilution in sterile distilled water and spread plating on LB agar plates. To obtain comparable values, DPA release was interrelated to spore numbers as molar DPA release per spore (48).

Determination of DPA release from spores after pressure treatment.

To evaluate the correlation between the spoVA^2mob operon and the amount of DPA released by pressure treatment, 4 strains with different copy numbers of the spoVA^2mob operon were tested: B. velezensis Fad 94 with 0 copies, B. amyloliquefaciens FUA2149 with 1 copy, B. amyloliquefaciens Fad 99 with 2 copies, and B. amyloliquefaciens Fad 108 with 3 copies of the spoVA^2mob operon per genome. Spore suspensions were standardized to an OD₆₀₀ of 2 and divided into aliquots of 200 μl. The DPA release was quantified after pressure treatment at 600 MPa and 80°C for 1, 2, 4, and 8 min. The total content of DPA was quantified after autoclaving for 60 min at 121.1°C. DPA release was expressed as the percentage of the total DPA content of spores as determined after autoclaving.

Statistical analysis.

The independent experiments were repeated at least three times (biological replicates). Statistical analysis was performed with RStudio 3.4.3 (RStudio Inc., Boston, MA, USA) software using a mixed model with copy number and strain treated as the fixed factor and random factor, respectively. A P value of ≤0.05 was considered statistically significant.

Accession number(s).

The GenBank accession numbers for the studied strains, SDKF00000000, SDKI00000000, SDKE00000000, SDKH00000000, SDKG00000000, VRTU00000000, VRTV00000000, VRTW00000000, and VRTX00000000, are listed in Table 1.