|Home | About | Journals | Submit | Contact Us | Français|
Spore heat resistance, germination, and outgrowth are problematic bacterial properties compromising food safety and quality. Large interstrain variation in these properties makes prediction and control of spore behavior challenging. High-level heat resistance and slow germination of spores of some natural Bacillus subtilis isolates, encountered in foods, have been attributed to the occurrence of the spoVA2mob operon carried on the Tn1546 transposon. In this study, we further investigate the correlation between the presence of this operon in high-level-heat-resistant spores and their germination efficiencies before and after exposure to various sublethal heat treatments (heat activation, or HA), which are known to significantly improve spore responses to nutrient germinants. We show that high-level-heat-resistant spores harboring spoVA2mob required higher HA temperatures for efficient germination than spores lacking spoVA2mob. The optimal spore HA requirements additionally depended on the nutrients used to trigger germination, l-alanine (l-Ala), or a mixture of l-asparagine, d-glucose, d-fructose, and K+ (AGFK). The distinct HA requirements of these two spore germination pathways are likely related to differences in properties of specific germinant receptors. Moreover, spores that germinated inefficiently in AGFK contained specific changes in sequences of the GerB and GerK germinant receptors, which are involved in this germination response. In contrast, no relation was found between transcription levels of main germination genes and spore germination phenotypes. The findings presented in this study have great implications for practices in the food industry, where heat treatments are commonly used to inactivate pathogenic and spoilage microbes, including bacterial spore formers.
IMPORTANCE This study describes a strong variation in spore germination capacities and requirements for a heat activation treatment, i.e., an exposure to sublethal heat that increases spore responsiveness to nutrient germination triggers, among 17 strains of B. subtilis, including 9 isolates from spoiled food products. Spores of industrial foodborne isolates exhibited, on average, less efficient and slower germination responses and required more severe heat activation than spores from other sources. High heat activation requirements and inefficient, slow germination correlated with elevated resistance of spores to heat and with specific genetic features, indicating a common genetic basis of these three phenotypic traits. Clearly, interstrain variation and numerous factors that shape spore germination behavior challenge standardization of methods to recover highly heat-resistant spores from the environment and have an impact on the efficacy of preservation techniques used by the food industry to control spores.
Bacillus subtilis spores, which are widely present in nature, can easily contaminate food products (1, 2). Because of their resistance to environmental stresses, such as extreme temperatures, desiccation, radiation, and exposure to different chemicals, spores are able to survive preservation treatments that are applied in the food industry (2,–4). Surviving spores can germinate and resume vegetative growth in a food product and subsequently cause spoilage (1, 5, 6). Spore germination is a necessary step in resuming vegetative growth; thus, the probability that the spores germinate in food products needs to be taken into account when assessing the risk of spoilage or outgrowth of pathogenic spore formers. Moreover, both inhibition of spore germination and induction of germination before inactivation treatments (which renders the spore sensitive) are used in the food industry to improve food safety (7,–12). However, none of these strategies is completely effective.
In nature, spores typically germinate in response to nutrients such as amino acids, sugars, and/or nucleosides. Nutrients need to permeate the spore coat through a process facilitated by the GerP proteins (13,–15) and the other outer layers of the spore to gain access to specific nutrient germinant receptor complexes (Ger receptors, or GRs) that are located in the spore inner membrane (IM) (16, 17). Germinant receptors are predominantly built up by three subunits, A, B, and C, which are encoded by tricistronic operons (18, 19). Subunit A contains five to six predicted transmembrane (TM) helices and a hydrophilic domain at both N and C termini (18, 20). Subunit B, which might play a role in recognition of germinants (21), appears to be an integral inner membrane protein, with 10 to 12 TM helices (18). Subunit C is a predominantly hydrophilic lipoprotein. It is anchored to the outer surface of the inner membrane by a diacylglycerol anchor that itself is attached to an N-terminal lipobox cysteine (22). Some ger operons encode a fourth small subunit D that consists of two TM helices and likely modulates the function of the respective GRs (23, 24).
A number of germinant receptors and the nutrients to which they respond vary between different Bacillus species and strains. The extensively studied laboratory strain B. subtilis 168 contains three functional germinant receptors: GerA, GerB, and GerK (25). The GerA complex responds to l-alanine (l-Ala). The GerB and GerK receptors cooperatively initiate germination in response to a mixture of l-asparagine, d-glucose, d-fructose, and potassium ions (AGFK). In fact, GerB responds predominantly to l-asparagine whereas GerK responds to sugars, but neither of these receptors can trigger germination alone (19, 26). Additionally, B. subtilis 168 contains two GR operons, yndDEF and yfkQRTS, that seem to be inactive and poorly transcribed (25). Moreover, certain foodborne B. subtilis strains comprise the incomplete and likely nonfunctional gerX germination operon that encodes putative GR subunits A and C but lacks a gene for the B subunit (27, 28).
Germinant receptors form one or two germination clusters (germinosomes) in the spore IM together with another small lipoprotein, GerD, which is required for this process (29). Formation of the germination clusters increases the local concentration of GRs and likely facilitates cooperativity and synergism between them (29,–31). Germination responses can be increased by applying a sublethal heat treatment (so-called heat activation, or HA) prior to exposure of spores to nutrient germinants (32). Heat activation is thought to act either directly on GR subunits or indirectly by altering properties of IM rather than on the GerD protein or germinosome formation (32).
After sensing germinants by the respective GRs, an unknown signal is transduced to downstream effectors. This leads to the release of monovalent cations and of dipicolinic acid chelated by Ca2+ ions (Ca-DPA) from the spore core (19, 26). Ca-DPA is transported across the IM by a channel formed by products of the conserved heptacistronic spoVA operon, which in B. subtilis 168 consists of seven genes: spoVAA-D, spoVAEb, spoVAEa, and spoVAF (19, 26, 33). Release of Ca-DPA allows for partial rehydration of the spore core and activation of the cortex lytic enzymes (CLEs), CwlJ and SleB (19, 26). Degradation of the protective peptidoglycan cortex leads to further water uptake and the completion of germination (19, 26).
Germination responses can vary strongly between Bacillus species and strains (28, 34, 35). For this reason, studies performed on the model strains, which are adapted to laboratory conditions, do not accurately reflect spore germination properties of strains that contaminate food products. This variation hinders the ability to predict the germination behavior of problematic spores and therefore complicates a risk assessment and the development of efficient spore inactivation treatments.
Recently, variation in spore germination and spore heat resistance has been partly attributed to interstrain differences in the absence or presence (and copy number) of the heptacistronic spoVA2mob operon that is usually carried on the Tn1546-like transposon element (which lost the ability of transposition) (27, 28). Tn1546 additionally carries four other operons that alone do not have apparent effects on spore heat resistance or germination (27, 28). Different B. subtilis strains have between 0 and 3 copies of spoVA2mob (two on the Tn1546-like elements integrated within the yitF gene and between the yxjA and yxjB genes and one in a chromosomal location that could not be determined in the nonclosed genome sequences) (27). Strains carrying an increasing copy number of spoVA2mob produce spores with elevated heat resistance and an affected germination phenotype in a nutrient-rich medium. The exact functions of the encoded proteins on this operon remain unclear. Three of them, SpoVAC2mob, SpoVAD2mob, and SpoVAEb2mob, are homologous to the conserved SpoVAC, SpoVAD, and SpoVAEb proteins (27, 28) that are involved in Ca-DPA transport via the spore IM (26, 36, 37). Another four, out of which two are predicted to be associated with the spore membrane, constitute putative proteins containing domains of unknown functions (27, 28). The spoVA2mob products have been hypothesized to affect spore germination (and HR) by altering properties of the spore IM in which the majority of the spore germination apparatus is located (28).
In this study, we investigated spore germination requirements and efficiencies for 17 B. subtilis strains with known genomic sequences (some of which constitute problematic food-spoilers) that contain different copy numbers (0 to 3) of the spoVA2mob operon (4, 27) (Table 1). Eight strains (168, B4055, B4056, B4057, B4058, B4060, B4061, and B4143) lack spoVA2mob in their genomes and produce low-heat-resistant spores. Nine strains have either one (B4146), two (B4068, B4069, B4071, B4072, and B4073), or three (B4070, B4067, and B4145) spoVA2mob copies and form spores with an increasing level of high heat resistance (4, 27) (Table 1). The presence and sequences of specific germination genes were analyzed to link differences in spore germination phenotype and spore germination requirements to genetic properties of the strains. This study demonstrates that the presence of the spoVA2mob operon correlates with increased spore HA requirements and, if spoVA2mob is present in multiple copies, with inefficient germination in response to AGFK. Additionally, sequences of germinant receptor proteins seemingly affect the final spore germination phenotypes and heat activation requirements.
Under laboratory conditions, spores of B. subtilis are often heat activated (HA) for 30 min at 70°C to increase their germination responses to nutrients (51,–53). In this study, germination (which, under a phase-contrast microscope, is reflected by the phase-bright to phase-dark transition of spores) in response to l-alanine and AGFK was assessed for HA and non-HA spores of various B. subtilis strains that exhibit low to extremely high heat resistance. When not heat activated, almost all low-heat-resistant spores (B4055, B4056, B4057, B4060, B4061, and 168) responded efficiently to l-alanine (means ± absolute deviations of the means were between 84.8% ± 14.1% and 98.4% ± 1.6% phase-dark spores within 3 h), with the exception of B4058 and B4143, which showed no to poor germination (2.3% ± 1.6% and 20.6% ± 6.3% phase-dark spores) (Table 2). Spores with high-level heat resistance (B4067 to B4073, B4145, and B4146), however, responded at best moderately (up to 38.0% ± 0.6% phase-dark spores) without heat activation (Table 2). When heat activated at 70°C, the germination percentage increased at least 2-fold for the poorly responding high-level-heat-resistant spores of seven strains (namely, B4068 to B4073 and B4146), whereas spores of strains B4058, B4067, and B4145 remained unresponsive to l-alanine (Table 2).
Using AGFK as a germination trigger, all but one low-heat-resistant spore types germinated almost entirely without the need for HA (Table 3). The exception (B4058) reached complete germination after activation at 70°C in response to AGFK (but not to l-alanine with or without HA) (Table 3). In contrast, almost all high-heat-resistant spores germinated very poorly (0.8% ± 0.2% to 10.6% ± 4.9% phase-dark spores) in response to AGFK, even after HA at 70°C. Only spores of strain B4146, which exhibit moderately elevated heat resistance (4, 27), germinated moderately (52.0% ± 16.9% germinated spores) under these conditions (Table 3).
More severe heat activation has been proposed to improve germination of highly resistant spores of certain species and strains (54,–56). Thus, to test whether optimizing HA conditions can improve germination of the nine high-heat-resistant B. subtilis spore types (Table 1), we preincubated these spores at 80, 87, 95, and 100°C for 30 min.
In general, activation at temperatures exceeding 70°C increased the yields of germinated high-heat-resistant spores in response to both germinants, with l-alanine-induced germination (Table 2) being optimally supported by somewhat lower HA temperatures (predominantly 80°C, with 87°C giving similar results in a few cases) than those required for AGFK-induced germination (mostly 87 to 95°C and/or 100°C) (Table 3). The only exception was l-alanine-induced germination of B4146 spores, for which the best results (41.7% ± 11.2% germination) were obtained after treatment at 70°C. In l-alanine, the pretreatment at 87°C gave slightly lower germination levels than at 80°C for the majority of the high-heat-resistant spores. A temperature of 95°C severely reduced l-alanine spore germination for all strains except for B4068 and B4073. Regardless of the activation temperature, l-alanine did not trigger responses of B4067 and B4145 spores.
AGFK-induced germination of high-heat-resistant spores was rather poor (between 14.2% ± 4.1% and 42.7% ± 17.7%), even when the most optimal HA treatments at 87, 95, and/or 100°C were applied (Table 3). The only exception was B4146 spores that germinated efficiently in AGFK, yielding 81.7% ± 9.2% and 87.7% ± 1.8% phase-dark spores after pretreatment at 87°C and 95°C, respectively (Table 3). In some cases, preheating at 100°C led to a phase change of up to 10.9% ± 5.3% spores in the absence of germinants (Table 3).
In comparison, heat activation of low-heat-resistant spores of the reference strain B. subtilis 168 was not required for efficient germination in our experiments, as similar results were obtained for non-HA spores and for spores preheated at 70°C and 80°C (Tables 2 and and3).3). In contrast, preheating at 87°C caused a phase-bright to phase-dark transition of 18.4% ± 1.5% of strain 168 spores in the absence of germinants (Tables 2 and and3),3), possibly due to spore damage and an increase in permeability of outer spore layers, leading to a rehydration of the core.
Besides the differences in germination efficiency (Tables 2 and and3),3), the high-heat-resistant spores also germinated less rapidly than the low-heat-resistant spores in response to both l-alanine and AGFK, even after optimal heat activation (Fig. 1). These differences in germination kinetics observed between the various spore types are consistent with the previously described negative effect of the spoVA2mob operon on the spore germination rate (28).
The results described above indicate a correlation between the presence of the spoVA2mob operon, increased spore HA requirements, and low germination efficiencies, especially in AGFK (Tables 2 and and3).3). Nevertheless, strong variations in spore germination capacity and required HA were seen between different strains that contain the same spoVA2mob copy numbers and between different germination pathways (l-alanine and AGFK). Hence, factors other than spoVA2mob likely contribute to the eventual germination behavior to a greater extent in l-alanine and to a lesser extent in AGFK.
To find genetic features other than components of the spoVA2mob operon that tune spore germination properties, we investigated genomic sequences of the 17 B. subtilis strains in regard to genes that play key roles in nutrient-induced spore germination (19, 57, 58). In general, all key germination genes are present in all B. subtilis strains investigated in this study, including those encoding the three main germinant receptors, GerA, GerB, and GerK (see File 1 in the supplemental material); six GerP A-F proteins; the SpoVAA-AB-AC-AD-AEb-AEa-AF proteins; and the CwlJ and SleB cortex lytic enzymes (data not shown). Of the two operons yndDEF and yfkQRST, which encode putative and probably nonfunctional germinant receptors in B. subtilis 168 (25), yndDEF was present in all strains except B4057 (Table S1). In comparison, the intact yfkQRST operon occurs in 7 (168, B4055, B4056, B4057, B4060, B4061, and B4143) out of the 17 strains (Table S1), although its residues could be found in the form of pseudogenes in genomes of the remaining strains (Table S1). As reported recently, nine strains (B4067 to B4073, B4145, and B4146) additionally contain the gerXA and gerXC genes, which encode the putative GR subunits A and C (Table S1) (27, 28). However, the encoded GerXA and GerXC are very likely nonfunctional due to severe truncations (especially at the N and C termini) that result in a decreased number of predicted TM helices in GerXA and a loss of the N-terminal lipobox in GerXC in all strains except B4146 (Fig. S1).
Even though the three main GR operons (gerA, gerB, and gerK) are present in all of the investigated strains, the foodborne isolates B4067 and B4145 contain strong deviations in the structure of the gerB operon (see File 1 in the supplemental material). In the laboratory strain B. subtilis 168, this operon encodes two integral membrane proteins, GerBA and GerBB, and a lipoprotein, GerBC (see File 1). In the strains B4067 and B4145, however, a stop codon at nucleotide positions 538 to 540 divides the gerBA gene into two predicted open reading frames (ORFs), gerBA-N and gerBA-C (Fig. S2A). Moreover, gerBB and gerBC are fused into one ORF (gerBB-BC) (Fig. S2A). A topology prediction suggests that the GerBB and GerBC parts of the encoded GerBB-BC fusion protein are arranged in the IM similarly (the N terminus of the GerBB part localizes inside the spore core, followed by 11 TM helices, and the majority of the GerBC part localizes outside the spore core) to the regular nonfused GerBB and GerBC proteins (Fig. S2B). The encoded GerB receptor of B4067 and B4145 might retain partial functionality, as spores produced by these two strains were capable of moderate germination in AGFK (Table 3), which in B. subtilis 168 requires both GerB and GerK (25).
Germinant receptors are known to play a fundamental role in spore responsiveness (specificity and affinity) to nutrient germinants (25, 59, 60). Moreover, besides the spore IM, GRs are the most probable spore component targeted when heat is applied to activate spore germination (32). Thus, in addition to the chromosomal presence of the spoVA2mob operon, properties of GR proteins likely contribute to the diversity in germination efficiencies and HA requirements observed among the tested spore types. To find features distinguishing GRs of the spores with strongly affected germination phenotypes (very poor/no germination or high HA requirements), we analyzed multiple-amino-acid sequence alignments of the GR subunits (see Data Sets S1 and S2 in the supplemental material).
The analysis revealed an 82.5% to 100% amino acid sequence identity between the corresponding GR proteins of individual strains, with the GerB receptor being the most variable (Data Set S1). Spores with the highest heat resistance level and high HA requirements for AGFK germination (B4067 to B4073 and B4145) contained several distinctive amino acid residues found in various regions of the GerBB, GerKA, GerKB, and GerKC proteins (Table 4). The same amino acid substitutions (Table 4) also concurred with the weak responsiveness of spores to AGFK, as the high-level-heat-resistant spores harboring 2 or 3 spoVA2mob copies in their genome germinated rather poorly in this cogerminant mixture (Table 3). Three of the distinct amino acids additionally were found in GerKB (V147; I181) and GerKC (D398) of slightly high-heat-resistant B4146 spores (Table 4), which needed moderate HA and responded strongly to AGFK (Tables 2 and and33).
No distinctive features typical for all high-heat-resistant spores were found in sequences of the GerAA, GerAB, GerAC, GerBA, and GerBC proteins (Table 4, Data Set S2). However, a few common amino acid substitutions were present in GerAA, GerAB, and GerBC of B4068 and B4073 spores (Table 4), whose responses to l-alanine were not decreased by heating at 95°C (Table 2); L373I in GerAA and S92L in GerAC (Table 4) are present in the protein regions that have previously been shown to play a role in GR functionality (61, 62). For the spores that germinated weakly (B4143, B4146, B4071, and B4070) or not at all (B4067, B4145, and B4058) in l-alanine, no common alternations in the GerA subunits were observed (Table 4, Data Set S2). Nevertheless, B4058, B4143, B4067, and B4145 contained various unique strain-specific amino acid residues in GerAA, GerAB, and GerAC that distinguished them from the spores that responded efficiently to l-alanine (Table 4).
Besides genetic content and specific protein sequences, the expression levels of germination genes during sporulation play an important role in the eventual germination behavior of the spores. Thus, we analyzed transcription of germination genes during sporulation for B. subtilis 168, B4143, B4146, B4072, and B4067 that each shows a distinct germination phenotype: (i) 168 germinates well in both l-alanine and AGFK; (ii) B4143 responds only to AGFK; (iii) B4146 germinates efficiently in AGFK but only moderately in l-alanine; (iv) B4072 germinates only in l-alanine; (v) B4067 germinates rather poorly in AGFK and not at all in l-alanine (Tables 2 and and33).
As can be seen in File 1 in the supplemental material, the three main germination operons (gerA, gerB, and gerK), as well as other important germination genes, produced transcripts with similar lengths and mostly comparable expression levels in all investigated strains. Also, transcription of gerB in B4067 was comparable to that in the other strains despite differences in the genetic organization of this operon (Table S2, Fig. S3). The putative GR operons, yndDEF and yfkQRST (25), also were transcribed, with yndDEF exhibiting somewhat higher expression in B4143 and B4067 than in 168, B4146, and B4072 (Table S2, Fig. S3). Moreover, the part of the yfkQRST transcripts coding for the yfkT gene appeared to be unstable as visualized in JBrowse (Fig. S3). The potential germination genes located on the Tn1546 transposon present in the B4146, B4067, and B4072 strains also were expressed (Table S2). The gerXA and gerXC genes, which encode likely nonfunctional GR subunits, were transcribed 1.8- ± 0.1- to 8.0- ± 0.3-fold more strongly in B4146 and B4072 than the gerAA and gerAC genes of these strains, respectively. The spoVAC2mob, spoVAD2mob, and spoVAEB2mob genes of the spoVA2mob operon were expressed similarly to the corresponding genes from the regular nontransposon spoVAA-AF operon (spoVAC, spoVAD, and spoVAEB, respectively). Transcription of the first gene of the operon, which encodes a putative protein of unknown function, was exceptionally high (57- to 99-fold higher than transcription of the spoVAC gene in the respective strains), while the second and the last genes of spoVA2mob, which also code for hypothetical proteins (27, 28), were expressed to a level similar to that of spoVAC. The second copy of the spoVA2mob operon present in the B4067 strain in an unknown genomic context was transcribed around 2.3- to 8.5-fold more weakly than the operon on the Tn1546-like transposon element (Table S2). Again, expression of the first hypothetical gene was the strongest. Whereas the last hypothetical gene, which encodes a putative membrane protein, has been suggested to affect spore germination and heat resistance, the role of the second and the (highly expressed) first genes of spoVA2mob remain unknown (27, 28).
This study shows that the spore germination capacities in response to nutrients and accompanying heat activation requirements are likely shaped by the presence (and copy number) of the spoVA2mob operon in the genome and sequences of germinant receptor proteins. In contrast, the transcription of germination genes during sporulation, which was similar in the five analyzed strains (168, B4067, B4145, B4072, B4143, and B4146) (see File 1 in the supplemental material), does not correlate with spore germination phenotypes.
The spoVA2mob operon has been shown to (i) increase spore heat resistance (27), (ii) prolong the time required for spore germination, and (iii) decrease germination rates (28). Consistently with our previous work in rich LB medium (28), the current study shows that the isolates containing the spoVA2mob operon exhibited slower l-alanine- and AGFK-induced spore germination (Fig. 1). Here, we additionally demonstrate that these differences in germination kinetics between spores with and without spoVA2mob persisted even after optimal heat activation (HA) treatments (Fig. 1). Importantly, the present work extends former findings by revealing additional correlations between spoVA2mob presence and two other spore properties, namely, higher spore HA requirements and lower germination efficiencies (Tables 2 and and3).3). Furthermore, we show that two copies of the spoVA2mob operon are transcribed but to different levels (Table S2).
Overall, low-level-heat-resistant spores germinated well in response to l-alanine and AGFK, with or without HA (at 70°C). High-level-heat-resistant spores generally germinated poorly in l-alanine without HA and in AGFK after relatively mild HA (70°C). Interestingly, an optimized HA treatment strongly increased germination efficiency of high-level-heat-resistant B. subtilis spores that contain the spoVA2mob operon only in response to l-alanine (usually after HA at 70 to 80°C) and for strain B4146 (harboring only one copy of spoVA2mob) in response to AGFK (87 to 95°C) (Tables 2 and and3).3). In contrast, AGFK-induced germination of spores with 2 or 3 spoVA2mob copies remained relatively poor, even though HA at very high temperatures (95 to 100°C) improved yields of germinated spores up to ~40% (Table 3).
A few important implications arise from these results. First of all, gene products encoded by the spoVA2mob operon (and possibly other operons on Tn1546) are likely directly or indirectly responsible for inefficient germination of non-HA spores and for an increase in spore HA requirements (Tables 2 and and3),3), with these effects being amplified when spoVA2mob is present in multiple copies. This implies a direct genetic link between spore high heat resistance (27), slow germination kinetics (28), low germination efficiencies, and elevated requirements for HA. The exact mechanism by which spoVA2mob affects these processes is not fully understood. Three out of seven products of the spoVA2mob operon, namely, SpoVAC2mob, SpoVAD2mob, and SpoVAEb2mob, display 55%, 49%, and 59% amino acid sequence identity, respectively (27, 28), to the SpoVAC, SpoVAD, and SpoVAEb proteins encoded by the regular conserved heptacistronic spoVA (spoVAA-spoVAF) operon (63, 64). The latter conserved SpoVA proteins are required for DPA uptake during sporulation (36, 63) and DPA release during germination (26, 36, 37, 64). The products of spoVA2mob seem to elevate spore heat resistance partly via an auxiliary role in DPA uptake, as indicated by significantly higher (~1.5-fold) DPA concentrations in spores of the B. subtilis 168 strain engineered to harbor the Tn1546 transposon with the spoVA2mob operon (strain B4417) and of B. subtilis 168 amyE::spoVA2mob than in the wild-type 168 spores (27). However, this phenomenon does not correlate with improved efflux of DPA upon germination. In fact, during the 2-h exposure to 10 mM AGFK, DPA was released less efficiently from B4417 spores (31.6% ± 0.7% DPA released) than from B4417ΔspoVA2mob (45.8% ± 3.1%) and 168 (67.1% ± 6.3%) spores, as monitored by fluorescence of released DPA with terbium (Tb3+-DPA) (data not shown) measured in a fluorescence plate reader, as similarly described before (65). In contrast to the regular conserved spoVA operon, spoVA2mob does not encode homologs of the SpoVAEa and SpoVAF proteins, which are reportedly important for DPA release during GR-dependent spore germination but not for DPA uptake during spore formation (66). For this reason, the spoVA2mob products may not be able to support DPA transport during germination. Instead, they might compete with the regular SpoVA channels or interfere with the SpoVA proteins and/or GRs (37, 67), either directly or indirectly by altering the spore IM properties. Such interference could be caused not only by the three SpoVA homologs but also by the four proteins of unknown function, as suggested by a somewhat improved spore germination after the deletion of the last gene that encodes a putative membrane protein from strain B4417 (28).
Second, findings from this study indicate that various GR types in the high-level-heat-resistant spores differ in their HA requirements and thermal stabilities. Regardless of the spoVA2mob copy number, germination in l-alanine, which occurs via GerA (25), was optimally activated at lower temperatures than germination in the AGFK mixture (Tables 2 and and3),3), which is mediated by a cooperative action of GerB and GerK (25). Moreover, l-alanine germination was easily diminished by too severe preheating (Table 2). Thus, consistent with a previous report on low heat resistant spores of B. subtilis 168 (32), the GerA receptor of the high-heat-resistant spores also seems to require the mildest HA treatment and appears the most thermolabile (Table 2). Differences in HA requirements of individual GRs support the notion that HA directly affects the GR proteins (32). However, the results from our current study make it plausible that HA acts in a dual manner, both indirectly on the IM and directly on the GRs. Alternatively, HA could improve spore germination by affecting the IM, but the exposure to (excessive) heat could simultaneously negatively influence GR proteins. Therefore, variation in the thermal stabilities of specific GRs could lead to differences in the way HA affects individual nutrient-induced germination pathways.
Third, our results suggest that in addition to spoVA2mob, the amino acid sequences of GR proteins likely affect GR thermal stabilities and HA requirements. Spores of B4068 and B4073, which germinate efficiently with l-alanine even after exposure to 95°C (Table 2), have several distinct amino acid residues in the three GerA subunits (Table 4). Similarly, characteristic amino acids occur in GerBB, GerBC, and GerKA-KD (Table 4) of the foodborne strains that germinate moderately in AGFK only after severe HA at 95 or 100°C (Table 3). Contradictory to the current work, our previous study (28) showed that HA at higher temperatures does not improve germination rates and efficiencies of spores of the B. subtilis 168 strain with the spoVA2mob operon introduced on the Tn1546 transposon (B4417). This discrepancy could be explained by an assumption that GRs of B. subtilis 168 are intrinsically less thermostable than GRs of high-level-heat-resistant spores; thus, more severe HA would simultaneously counteract the effect of spoVA2mob and reduce activity of B4417 GRs. This explanation further supports the hypothesis that final spore HA requirements are concurrently shaped by the spoVA2mob operon and by the (sequence-dependent) properties of the individual GRs.
Another observation from this work is that certain germination pathways, in particular involving the GerA receptor and l-alanine, may be (at least temporarily) deactivated by preheating at less severe conditions than are required for the inactivation of spores. High-level-heat-resistant spores included in this study have decimal reduction times (D values) at 100°C of 83.0 ± 32.3 min for B4146 spores and between 310 ± 137 and 4,224 ± 2,470 min for B4067 to B4073 and B4145 spores (Table 1) (4, 27), and their counts are reduced no more than 0.1 log (1.26-fold) after heating at 100°C for 1 h (4). In contrast, germination responses in l-alanine of some high-level-heat-resistant spores are substantially decreased after a 30-min heat treatment at 80 to 95°C (Table 2). Most strikingly, l-alanine-induced germination of B4072 spores decreased ~6-fold (94.9% ± 1.0% to 15.7% ± 2.8%) after a heat treatment at 95°C compared with heating at 80°C (Table 2). Although a certain low degree of spore inactivation may occur at the highest used activation treatments (95 to 100°C), the observed reduction in germination cannot be attributed solely to spore killing, since the D value for B4072 spores at 100°C reaches 340 ± 197 min (Table 1) (27). Taken together, the observed decrease in spore germination in l-alanine after such heat treatments more likely is due to specific inactivation of the GerA receptor complex than spore killing. Since the complete spore is not inactivated, germination may still occur in response to other germination triggers.
The presence of multiple spoVA2mob copies in the genomes of eight B. subtilis strains (B4067 to B4073 and B4145) correlates with modest (maximally, 42.7% ± 17.7%) spore germination in AGFK (Table 3) even after HA at 95 to 100°C. However, a causative relationship between these two factors is unclear. First of all, this was not observed for l-alanine-induced germination, which becomes efficient after proper spore HA (Table 2) for spores of strains harboring spoVA2mob. Second, the same spores that harbor 2 to 3 spoVA2mob copies (and germinate weakly in AGFK) share several changes in amino acid sequences of the GerBB, GerBC, and GerKA-GerKD subunits that distinguish them from spores that respond strongly to AGFK (and that contain 0 to 1 copies of spoVA2mob) (Table 4). Unfortunately, despite multiple attempts we were unable to introduce the respective mutations in the relevant ger genes in B. subtilis 168 in order to directly test their effect on spore germination and HA requirements. Still, it cannot be excluded that some of these distinctive amino acid residues contribute to the lower responsiveness of these spores to the AGFK mixture. Notably, the same unique amino acids might simultaneously cause reduced functionality and higher thermostability of these GRs, as a decreased conformational flexibility that is characteristic for thermostable proteins has been suggested to compromise their activities (68, 69). Potentially stabilizing amino acid changes, such as substitutions to l-alanine, which tend to stabilize α-helices (70, 71), or to hydrophobic amino acids, which can increase a degree of protein packing (72), can be found in the GR subunits of the investigated high-heat-resistant strains (Table 4). However, the prediction of mutations with thermostabilizing effects is generally challenging (73, 74) and in the case of GRs additionally is hindered by their membrane-associated localization and predominantly unknown protein structures and mechanisms of action (26, 32).
In contrast to GerB and GerK, no common distinctive residues in the GerA subunits were found for all five strains whose spores germinate barely (B4143, B4058, B4067, and B4145) or moderately (B4146) in response to l-alanine (Table 4, Data Set S2). A few identical amino acid changes were found in GerAB and GerAC of closely related B4067 and B4145 strains (Fig. 2), and one common substitution (G135C) was present in GerAC of strains B4143 and B4146 (Table 4, Data Set S2). The three GerA subunits of strain B4058 contain multiple unique amino acid residues distinguishing them from the GerA proteins of all other analyzed strains (Table 4, Data Set S2). The very poor or moderate responsiveness to l-alanine observed for B4143, B4067, B4145, B4058, and B4146 spores therefore may be caused by various strain-specific changes in the GerA protein sequences (Table 4, Data Set S2). This constitutes an interesting subject for future investigation.
In summary, we show that apart from interstrain phenotypic variation in spore HR, germination, and HA requirements that correlate to the presence of spoVA2mob, the influence of other intrinsic factors, in particular GR protein sequences, further complicates the prediction of spore properties and behavior in response to heat and nutrients. These findings have major impacts on practices in the food industry and challenge standardization of risk assessment and preservation techniques. First, in quality assessment, the recovery and enumeration of spores from food products and food processing equipment involves specific heat treatment methods and commonly applied plating techniques, which can result in an underestimation of viable spores due to poor, heterogeneous, and unpredictable spore germination properties. In addition to optimization of germinant conditions (75), this study shows that the use of optimal HA treatments, especially for high-level-heat-resistant spores, could help alleviate this problem. Furthermore, our findings suggest that some of the heat treatments that are used in food processing for spore inactivation can result in the activation rather than inactivation of high-level-heat-resistant spores and thus increase unwanted spore germination and outgrowth in food products, thereby decreasing food safety. Finally, as processing conditions in food manufacturing likely select for spores with properties that are different from those of the commonly studied laboratory strains, our study underlines the importance of extending scientific research from model laboratory organisms to industrially relevant species and strains.
Strains used in this study are listed in Table 1. Spores were prepared during a 7-day incubation at 37°C on Schaeffer-agar plates without antibiotics as described before (4, 27, 28). Harvested spores were washed for 4 days with cold sterile Milli-Q water and stored at 4°C for at least 1 month prior to the germination experiments to ensure full maturation of the spores (76). The spore crops consisted of ≥95% dormant phase-bright spores as determined by phase-contrast microscopy.
Prior to the experiments, spores were washed with cold sterile Milli-Q water. Before exposure to nutrients, spore suspensions with an optical density at 600 nm (OD600) of 10 were subjected or not (non-heat-activated spores, nH) to a 30-min heat treatment at 70, 80, 87, 95, and/or 100°C. Spores produced by individual strains had substantially different levels of heat resistance, as indicated by their distinct decimal reduction times, or D values, at 100°C (for comparison purposes, D values have been extrapolated from higher temperatures to 100°C for high-level-heat-resistant spores) (4, 27) (Table 1). For this reason, the choice of applied HA temperature was tailored for each strain, depending on the spore heat resistances (Tables 2 and and3).3). Thus, low-level-heat-resistant spores (D values were between 2.4 ± 0.7 and 12.0 ± 4.3 min at 100°C) (Table 1) produced by strains 168, B4055 to B4058, B4060 and B4061, and B4143 were heat activated in most cases at 70°C and maximally at 87°C. Moderately high-heat-resistant spores of strain B4146 (D value of 83.0 ± 32.3 min at 100°C) (Table 1) and high-level-heat-resistant spores of strains B4067 to B4073 and B4145 (D values between 310 ± 137 and 4,224 ± 2,470 min at 100°C) (Table 1) were heat activated at temperatures of up to 100°C. After heating, spores were cooled on ice and assessed by phase-contrast microscopy to ensure that the applied heat treatment did not cause spore phase transformation in the absence of germination triggers. Spores were diluted to an OD600 of 1 in 10 mM l-alanine or 10 mM AGFK mixture (l-asparagine, d-glucose, d-fructose, and KCl) in 25 mM Tris-HCl buffer (pH 7.4) with addition of 0.01% Tween 20 to prevent spore clumping and absorption to the plate wells (77). The 10 mM germinant concentrations were used, as these are known to be saturating for B. subtilis 168 spores (15, 65, 78). Germination was monitored at 37°C in a 96-well plate reader (Tecan Infinite 200; Tecan) by measuring the decrease in OD600, which corresponds with the change of dormant phase-bright spores to germinated phase-dark spores. Measurements were taken every 2 to 3 min for 3 h with shaking between steps. All spore germination experiments were performed on two independently prepared spore crops.
Plate reader samples were investigated by phase-contrast microscopy 3 h after the addition of nutrients as described before (75). Germination efficiency was calculated as a percentage of phase-dark (germinated) spores seen using phase-contrast microscopy in a total of 100 to 700 spores per spore crop and condition, using the Fiji software (79) (http://fiji.sc/Fiji), similarly to what has been described before (75).
For all of the analyzed B. subtilis strains, the phylogenetic tree was prepared (Fig. 2) with the PhyloPhlAn program based on the sequences of the selected 400 most conserved microbial proteins (https://huttenhower.sph.harvard.edu/phylophlan) (80) and displayed with the use of the iTOL tool (http://itol.embl.de) (81). For all of the predicted protein sequences encoded by the genomes of the 17 B. subtilis strains, an orthology prediction was performed applying two programs, Ortho-MCL (82) and ProteinOrtho (83) (data not shown), the latter with the use of B. subtilis 168 as a reference. To find potential functional equivalents for a selection of germinant receptor genes (see File 1 in the supplemental material), additional protein BLAST searches (84) were made using B. subtilis 168 sequences as queries to scan genomes of the foodborne strains. For the selected GR proteins found in this manner, protein sequence alignments were made using MUSCLE (85). Additionally, the genomic context of the selected genes was manually inspected after visualization of the genomes with Clone Manager software (Clone Manager v8; Scientific & Educational Software, Denver, CO, USA). In relevant cases, to verify operon structures, operon predictions were performed with Glimmer (86, 87) and visualized using the draw context tool in the Genome2D server (http://genome2d.molgenrug.nl) (88). The membrane topologies of the selected A, B, and D GR subunits were modeled using TOPCONS (http://topcons.cbr.su.se/) (89), and the secondary structures of the C subunits were predicted with PredictProtein (http://www.predictprotein.org) (90), except for the GerBC protein, the structure of which has been solved (22).
Sporulation of five B. subtilis strains, 168, B4143, B4146, B4072, and B4067, was induced by the resuspension method similarly to that described by Nicolas et al. (91). Shortly, strains were grown with shaking (200 rpm) at 37°C in casein hydrolysate (CH) medium until an OD600 of 0.6. Subsequently, whole bacterial cultures were spun down at 6,000 rpm for 8 min and resuspended in the same volume of prewarmed Sterlini-Mandelstam (SM) medium to induce sporulation. Samples for RNA isolation and for microscopic analysis were collected at various time points of the sporulation process. For RNA extraction, 15 ml of the cultures was centrifuged (12,000 rpm, 1 min), and the cell pellets were frozen in liquid nitrogen and stored at −80°C. For the microscopic analysis, 300-μl aliquots of cultures were precipitated by centrifugation at 10,000 rpm for 2 min, washed with phosphate-buffered saline (PBS), and fixed using 4% paraformaldehyde as described before (92). The microscopic samples were used to assess the stages of sporulation at the selected time points (data not shown).
Total RNA was isolated from the samples of sporulating cells by phenol-chloroform extraction and precipitation with ethanol and sodium acetate, as described before (93). To ensure homogenization of both mother cell and forespore compartments, sporulating cells were exposed to five 45-s-long bead-beating cycles, with at least 1-min intervals of cooling on ice. The RNA samples were subjected to next-generation directional sequencing on an Ion Proton Sequencer at the PrimBio Research Institute (Exton, PA, USA), and T-REx (94) was used for the RNA-Seq analysis. The transcripts were mapped on either the reference genome of B. subtilis 168 or the genomes of the respective individual strains. Transcripts mapped on the genome of B. subtilis 168 were visualized in JBrowse (95). In order to compare expression of selected germination genes between the strains or between two different genes within one strain, average ratios of maximal gene transcription signals (expressed as reads per kilobase of transcript per million mapped reads, or RPKM) during sporulation were calculated based on the results of two independent experiments.
We thank Cyrus A. Mallon for proofreading the manuscript, Erwin M. Berendsen for providing decimal reduction values of spores, as described in references 4 and 27, and Gerwin Kamstra for technical assistance.
The research was funded by TI Food and Nutrition, a public-private partnership on precompetitive research in food and nutrition. The funding organization had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
We have no competing interests to declare.
Supplemental material for this article may be found at https://doi.org/10.1128/AEM.03122-16.