Genomic identification of germinant receptors and phenotypic response to germinants in Parageobacillus and Geobacillus spp. strains

Highlights

• •(Para)Geobacillus spp. contain orthologs to 2–5 GRs.
• •A widely conserved GR across most strains was homologous to B. subtilis 168 GerK.
• •Most of (Para)Geobacillus spp. did not respond to single nutrients.
• •Most of (Para)Geobacillus spp. responded to complex media and CaDPA.
• •Colony formation efficiencies were low (3–53 %) for all strains.

Abstract

Bacterial spore germination is triggered by specific nutrients detected by germinant receptors (GRs) located in the inner membrane. While GR diversity and function are well-characterized in some Bacillus and Clostridium spp., they remain poorly understood in (Para)Geobacillus spp., despite the relevance of their spores in food spoilage and biotechnology. This study analyzed 105 genomes from (Para)Geobacillus strains to identify GR-encoding operons and evaluated the germination of 22 representative strains. All strains carried between two and five GRs, mostly orthologous to GRs in B. subtilis, Priestia megaterium, B. cereus, and B. anthracis, with wide variation among species and in some cases, such as G. stearothermophilus, among strains. Among GRs, the most commonly conserved GR across all strains was an ortholog of GerK from B. subtilis 168 and Priestia megaterium QM B1551. All strains germinated in rich nutrient medium (TSBYE), but none responded to common amino acids or nucleosides, and only a few (G. stearothermophilus ATCC 12980^T and ATCC 10149, G. thermodenitrificans DSM 465^T, and several P. thermoglucosidasius strains) germinated in the presence of various sugars. Notably, differences in germination responses did not align with GR diversity. Additionally, all strains displayed poor outgrowth on rich nutrient agar, with colony formation efficiencies ranging from 3 % to 53 %. These findings suggest that germination and outgrowth mechanisms in (Para)Geobacillus differ significantly from those in well-studied spore-formers, highlighting the need for further research.

Graphical abstract

1. Introduction

Sporulation is a survival strategy employed by certain prokaryotes in response to nutrient deprivation. The resulting spores can stay dormant and survive harsh environmental conditions that would be lethal to vegetative cells for extended periods, thus favoring their widespread distribution in nature (Christie and Setlow, 2020; Setlow and Johnson, 2019). Among the spore formers is the genus Geobacillus, which was established to accommodate thermophilic Bacillus isolates related on the basis of 16S rRNA gene sequences (Ash et al., 1991). Subsequently, the genus Parageobacillus was delineated from Geobacillus based on whole-genome phylogenetic analyses (Aliyu et al., 2016).

Some thermophilic species, such as Geobacillus stearothermophilus, produce extremely heat-resistant spores capable of surviving commercial sterilization processes applied to shelf-stable foods such as dairy and canned products (André et al., 2013; Burgess et al., 2010). These surviving spores can germinate and proliferate when exposed to favorable temperatures—conditions that are projected to become more frequent due to climate change—leading to food spoilage and economic losses (Kakagianni et al., 2016; Koutsoumanis et al., 2022). The challenge of controlling these food spoilers is further complicated by the tendency of culture-based quantification methods to underestimate the actual number of thermophilic spores present in food (Eijlander et al., 2019; Huesca-Espitia et al., 2016; Wells-Bennik et al., 2019). Since increasing the intensity of thermal treatment is not economically viable and negatively impacts food quality, novel food preservation strategies are needed to prevent the germination and/or outgrowth of these microorganisms.

On the other hand, (Para)Geobacillus spp. spores have relevant biotechnological applications. More specifically, they are used as biological indicators in sterilization efficacy tests and for detecting antibiotic residues in food (Guizelini et al., 2012; Wu et al., 2020; Wu et al., 2021) and as bioremediation tools (Li et al., 2021; Novik et al., 2019), where their dormant state is crucial for long-term viability (Novik et al., 2019). In each of these applications, precise control over the timing and extent of germination is essential to maximize benefits.

Germination has been extensively studied in model organisms from the Bacillus and Clostridium genera. In natural environments, the process is likely initiated by the presence of nutrients (germinants), such as amino acids, sugars, or purine nucleosides, sometimes accompanied by ions (Na⁺ or K⁺) as co-germinants (Paredes-Sabja et al., 2011; Setlow et al., 2017). In Bacillus spp., germinants are commonly recognized by their cognate GerA family germinant receptors (GRs), which act as nutrient-gated ion channels (Gao et al., 2023) and are co-localized in the germinosome of the spore’s inner membrane (IM) (Christie and Setlow, 2020; Setlow et al., 2017; Wang et al., 2020). Upon nutrient-GR recognition and irreversible activation of germination, a large fraction of the dipicolinic acid (DPA) present in the spore core, mostly chelated with Ca⁺² (CaDPA), is released through specialized channels located in the IM and replaced by water. Subsequently, cortex peptidoglycan is degraded by specific cortex-lytic enzymes (CLEs), enabling core expansion and complete rehydration. Ultimately, germination culminates in the restoration of metabolism and the outgrowth of a new vegetative cell (Setlow et al., 2003; Setlow and Johnson, 2019).

GR-encoding genes are typically organized in tricistronic operons encoding three subunits (A, B, and C) in a 1:1:1 stoichiometry (Moir et al., 2002), all of which—at least in Bacillus spp.—are essential for functional GRs (Moir, 2006). However, the genomic organization of GR genes varies between species and strains, and even among operons within the same genome, showing differences in gene number per operon and the structure of homologous subunits (Gupta et al., 2015; Paredes-Sabja et al., 2011; Ramirez-Peralta et al., 2013; Warda et al., 2017). Several GR types have been described based on genetic structure, genomic location, and germinant specificity (Carr et al., 2010b; Hornstra et al., 2006; Paredes-Sabja et al., 2011; Ross and Abel-Santos, 2010). Each GR can recognize one or more germinants, either individually or in combination, and may function independently or in cooperation with other GRs (Borch-Pedersen et al., 2016; Paidhungat and Setlow, 2000; Ross and Abel-Santos, 2010). For example, Bacillus subtilis possesses three structurally and functionally well-characterized GRs: GerA, which responds to l-alanine or l-valine; and GerB and GerK, which are both required for germination in the presence of the AGFK mixture (l-asparagine, d-glucose, d-fructose, and KCl) (Stewart et al., 2012).

Although understanding the germination of (Para)Geobacillus spp. spores is essential for developing innovative food preservation methods and enhancing their biotechnological applications, a significant knowledge gap remains in this area. While the core components and stages of germination are likely conserved—as observed in Bacillus and Clostridium (Christie and Setlow, 2020; Paredes-Sabja et al., 2011)—the genetic structure of GRs, which has been linked to phylogeny (Moir, 2006; Warda et al., 2017), as well as their germinant specificity, may differ substantially. Thus, the main objective of this study is to genetically delineate GRs in available genomes of (Para)Geobacillus spp. strains and to correlate GR diversity with their phenotypic response to germinants and colony-forming efficiency.

2. Materials and methods

2.1. Genome mining

Genomes were mined as described previously (Warda et al., 2017). A total of 105 genomes available in public databases (NCBI, ATCC, and JGI IMG) from several species and strains of (Para)Geobacillus were analyzed to delineate and compare structure of GRs (Table S1). To improve high- and low-quality draft genomes, Medusa server was used to scaffold the available contigs (Bosi et al., 2015) (Table S1).

Additionally, genomes of reference strains of B. cereus (ATCC 14579^T and ATCC 10987), B. subtilis (168), B. licheniformis (ATCC 14580^T), B. mycoides (KBAB4), B. anthracis (Ames Ancestor), Caldibacillus thermoamylovorans (B4167 and B4064), Saccharococcus thermophilus (DSM 4749^T), and Clostridium perfringens (SM101) were used for comparison (Table S2). Additionally, two orthologs of B. cereus ATCC 14579^T GerQ from Bacillus sp. N35-10-4 and B. cereus B4264, and four orthologs of B. subtilis 168 GerK from Priestia megaterium QM B1551: one designated as GerU (Christie and Lowe, 2007), which contains an additional B subunit termed GerVB (Christie et al., 2008), and three GRs designated as GerK₁, GerK₂, and GerK₃ (Gupta et al., 2013) were included.

To improve comparative analysis, all genomes were (re)annotated using RAST (Aziz et al., 2008; Brettin et al., 2015; Overbeek et al., 2014). From these annotated genomes, orthologous groups were studied using BLASTp in two directions (Altschul et al., 1997; Altschul et al., 2005) for each GR subunit. Multiple-sequence alignment (MSA) files were made using Clustal-Omega (https://ngphylogeny.fr/workflows/alacarte) (Madeira et al., 2024; Sievers et al., 2011), aligning the protein sequences within specific orthologous groups to facilitate the identification of pseudogenes (encoding non-complete proteins). To minimize redundancy, identical GR subunit sequences (Table S3) were consolidated, retaining only one representative in phylogenetic trees (Table S4).

2.2. Phylogenetic analysis

Phylogenetic trees of individual GR subunit A, B, C, or D based on the multiple aligned amino acid sequences of the strains were constructed using PhyML + SMS (Maximum likelihood-based inference of phylogenetic trees with Smart Model Selection) and Subtree Pruning and Regrafting (SPR) method (Capella-Gutiérrez et al., 2009; Guindon et al., 2010; Junier and Zdobnov, 2010; Lefort et al., 2017; Lemoine et al., 2019; Lemoine et al., 2018; Sievers et al., 2011; Yruela et al., 2021). The approximate Likelihood-Ratio Test (aLRT) with a seed value of 123.456 and bootstrap analyses with a value of 100 were performed. The tree and cladogram were midpoint-rooted and plotted with FigTree (http://tree.bio.ed.ac.uk/software/figtree/) (Rambaut, 2018). All main branches have bootstrap support values of > 75.

2.3. Strains and sporulation conditions

To evaluate experimentally the relationship between GR profiles and germination responses, 22 representative strains, including type strains and food and environmental isolates were selected (Table 1). All strains were collected from strain repositories (ATCC, BGSC, DSMZ, and NIZO) or researchers (Berendsen et al., 2016; Burgess et al., 2015).

Table 1.

Germinant receptors (GRs) present in the (Para)Geobacillus spp. strains included in the germination experiments. A check mark (✔) indicates the presence of a complete GR or the subunit C alone. When a pseudogene is present, it is denoted as ‘pseudogene’ along with the specific subunit containing it.

Species	Strain	GerK1	GerK2	GerX1	GerR1	GerQ1	Extra-C subunit
G. stearothermophilus	A1	✔			✔
G. stearothermophilus	D1	✔		✔	✔
G. stearothermophilus	P3	✔			✔
G. stearothermophilus	NIZO B4109	pseudogene subunit A			✔
G. stearothermophilus	NIZO B4114	✔		✔	✔
G. stearothermophilus	DSM 458	✔		✔	✔
G. stearothermophilus	ATCC 7953	✔		✔	✔
G. stearothermophilus	ATCC 12980T	✔		✔	✔
G. stearothermophilus	ATCC 10149	✔		✔	✔
G. thermodenitrificans	DSM 465T	✔	✔	✔	✔	✔
G. thermodenitrificans	G11MC16	✔		✔	pseudogene subunit C	✔
G. thermoleovorans	KCTC 3570T	✔		✔	✔
G. kaustophilus	NBRC 102445T	✔		✔	✔		✔
P. genomosp. 1	NUB3621	✔	✔	✔	✔
P. thermoglucosidasius	DSM 2543	✔	✔	✔
P. thermoglucosidasius	DSM 2542T	✔	✔	✔
P. thermoglucosidasius	C56-YS93	✔	✔	✔
P. thermoglucosidasius	M10EXG	✔	✔	✔
P. thermoglucosidasius	DSM 6285	✔	✔	✔
P. caldoxylosilyticus	DSM 12041T	✔	✔	✔	✔		✔
P. caldoxylosilyticus	NIZO B4119	✔	✔	✔	pseudogene subunit B		✔
P. toebii	DSM 14590T	✔	✔	✔	✔

Strains were stored at −80 °C in 2TY (Sigma-Aldrich, St. Louis, MO, USA) with 25 % glycerol (Panreac, Barcelona, Spain). Cells were revived by streaking on TSAYE (Tryptone Soya Agar + 0.6 % Yeast Extract; Oxoid, Basingstoke, UK) (55 °C, 24 h). A single colony was then inoculated into 20 mL of 2TY in a 250-mL flask (55 °C, 12 h, 130 rpm). Subsequently, 200 µL of this culture was transferred to 50 mL of TYE sporulation medium in a 500-mL flask (55 °C, 4 days) (Salvador et al., 2025). To examine the effect of sporulation temperature and maturation time on colony-forming efficiency, Parageobacillus thermoglucosidasius DSM 2542^T, G. thermodenitrificans DSM 465^T, and G. stearothermophilus ATCC 12980^T cultures were incubated at 50, 55, and 65 °C for 4 days, and at 55 °C for 1, 2, 4, and 7 days (Salvador et al., 2025).

For spore harvest, cultures were washed with distilled water, subjected to ethanol treatment (50 % v/v; SAEQSA, Zaragoza, Spain; Salvador et al., 2025), and purified by Nycodenz® gradient centrifugation (Freire et al., 2023; Ghosh and Setlow, 2009). Spore purity (99 % bright spores) was verified by phase-contrast microscopy (Nikon Eclipse E400, Tokyo, Japan). Suspensions were stored at −20 °C until use. Three independent spore batches were prepared to assess biological variability.

For comparison of colony-forming ability, B. subtilis 168 spores were prepared as described by Freire et al. (2023).

2.4. Germination assays

Germination kinetics were measured by the decrease in optical density at 600 nm (OD₆₀₀), resulting from DPA release and spore rehydration. Spore suspensions were adjusted to an OD₆₀₀ of 0.4–0.6 in the germination medium. Rich nutrient media included TSBYE, NBYE (Nutrient Broth + 0.6 % Yeast Extract, Oxoid), and 2TY, all with 5 µg/mL chloramphenicol (Sigma-Aldrich) to inhibit outgrowth. Individual germinants (Sigma-Aldrich unless otherwise noted) were tested in 25 mM HEPES buffer (pH 7.4) (Sigma-Aldrich) at saturating concentrations (100 mM unless otherwise specified): l-amino acids (l-alanine, l-phenylalanine (30 mM), l-tyrosine, l-tryptophan, l-glycine (VWR International Chemicals, Radnor, PA, USA), l-valine, l-leucine, l-isoleucine, l-cysteine hydrochloride, l-proline, l-methionine, l-serine, l-threonine (1 mM), l-lysine, l-arginine, l-histidine (Panreac), l-aspartic acid (10 mM), l-glutamic acid (25 mM), l-asparagine (AMRESCO, Solon, OH, USA), and l-glutamine); sugars (d-glucose (VWR International Chemicals), d-fructose (Panreac), d-mannose (VWR International Chemicals), d-galactose (Merck Millipore, Rahway, NJ, USA), d-ribose, d-arabinose (AMRESCO), d-xylose, lactose (Oxoid), maltose, sucrose (Panreac), and raffinose); nucleosides (adenosine (7.5 mM) and inosine (25 mM)); and inorganic salts (KBr, KI, and KCl (50 mM, Panreac)). The effect of common combinations of germinants was tested: FTY (l-phenylalanine, l-tyrosine, and l-tryptophan (1 or 10 mM; l-tyrosine 0.5 mM)); GVLI (l-glycine, l-valine, l-leucine, and l-isoleucine (1 or 10 mM)); CMST (l-cysteine, l-methionine, L‑serine, and l-threonine (1 or 10 mM); LAH (l-lysine, l-arginine, and l-histidine (1 or 10 mM)); DENQ (l-aspartic acid, l-glutamic acid, l-asparagine, and l-glutamine (1 or 10 mM)); casein hydrolysate or casamino acids (CA, 0.02–2.0 %, w/v; Merck Millipore); CA (0.2 %) with KI, KBr, or KCl (50 mM); l-alanine (10 or 100 mM) and inosine (10 mM) or d-cycloserine (1 mM; Sigma-Aldrich); AGFK (10 or 100 mM); GPLK (l-glycine, l-proline, l-leucine, and KBr; 0.1–100 mM, KBr at 50 mM); and each l-amino acid (10 mM; l-tyrosine 0.5 mM) with d-glucose (0.01–0.1 mM) or d-fructose (1 mM).

When indicated, spores were heat-activated (100 °C, 30 min; then 15 min on ice) or treated with NaNO₂ (0.2 M, pH 8.0, 30 °C, 17 h) (Sigma-Aldrich) (Zhou et al., 2013). GR-independent germination was induced using 60 mM CaDPA (1:1 DPA (Sigma-Aldrich) and CaCl₂ (VWR International Chemicals)).

The OD₆₀₀ decrease was monitored in a multiwell plate reader (CLARIOstar Plus, BMG, Ortenberg, Germany) that automatically measured values each 3 min for 4 h at 55 °C with 30 s shaking between reads. The germination extent was calculated as the percentage of OD₆₀₀ decrease (OD_t/OD_i × 100, where OD_i and OD_t are the initial and 4 h value, respectively). Where indicated, phase-contrast microscopy was used to determine the proportion of dormant (phase-bright) and germinated (phase-dark/grey) spores (100–150 spores/sample), with quantification limits of 5.0 % and 97.0 %. Please note that the OD₆₀₀ decrease correlated with the proportion of microscopically germinated spores, with some variation among strains, enabling classification: insensitive (OD₆₀₀ decrease ≤ 15.0 %, ≤ 5.0 % phase-dark/grey spores), weak (> 15.0–30.0 %, ∼ 5.0–40.0 %), modest (> 30.0–45.0 %, ∼ 40.0–90.0 %), and strong (> 45.0 %, > 90.0 %). Each condition was tested with ≥ 3 biological replicates per strain.

2.5. Colony-forming efficiency

Colony-forming efficiency was calculated as N_C/N_T × 100, where N_T represents the spore titer determined microscopically using a Thoma chamber (spores/mL) and N_C corresponds to the colony count on TSAYE (CFU/mL). TSAYE yielded the highest counts across strains when compared to NAYE (Nutrient Agar + 0.6 % Yeast Extract, Oxoid) and TBAB (Oxoid) (data not shown), as previously described (Wells-Bennik et al., 2019; Eijlander et al., 2019). Plates were generally incubated at 55 °C, the optimal temperature for thermophilic spore recovery (Kent et al., 2016; McGuiggan et al., 1994), and at 45–70 °C when specified, until colony numbers plateaued (72 h at 45 °C; 48 h at 50–60 °C; 24 h at 65–70 °C). To assess the effect of CaDPA-induced germination, spores were treated with CaDPA (60 mM, 4 h, 55 °C) prior to plating (55 °C, 48 h). B. subtilis 168 spores were plated on NAYE (37 °C, 24 h). The quantification limit was 300 CFU/mL. The coefficient of variation was calculated from 20 technical replicates of plate and Thoma chamber counts for B. subtilis 168 and P. thermoglucosidasius DSM 2542^T.

2.6. Statistical analysis

Statistical analyses were performed using one-way ANOVA with Tukey's test and two-way ANOVA with Tukey's or Sidak’s test using GraphPad PRISM 8.4.2 (GraphPad Software Inc., San Diego, CA, USA). Differences with p ≤ 0.05 were considered statistically significant. Figures show the means and standard deviations from ≥ 3 biological replicates.

3. Results

3.1. Delineation of GR operons and structure in (Para)Geobacillus genomes

The genomes of 105 (Para)Geobacillus spp. strains were (re-)annotated and analyzed to identify GR operons. We found that all strains contained between two and five GR operons. Given the complexity of GR annotation, characterized by variability and a lack of uniformity across and within different species (Abee et al., 2011; Ross and Abel-Santos, 2010; Warda et al., 2017), GR sequences from reference Bacillus spp. strains (B. subtilis, B. cereus, B. anthracis, B. licheniformis, and B. mycoides), Caldibacillus thermoamylovorans, and Clostridium perfringens, as well as four orthologs of B. subtilis 168 GerK from Priestia megaterium QM B1551, and two orthologs of B. cereus ATCC 14579^T GerQ from two Bacillus spp. strains (Table S2), were included for comparative analysis. Table S5 summarizes the GR orthologs identified across all (Para)Geobacillus spp. strains, with each type labeled by appending the numbers 1 or 2 to the name of the reference ortholog to which they were most similar. Table 1 presents the GR operons found in representative strains selected for further experimental work (see below).

All (Para)Geobacillus strains harbored at least one ortholog of GerK (designated GerK1) from B. subtilis 168 and Priestia megaterium QM B1551 (Table 1, S5). A second ortholog of GerK (designated GerK2) from B. subtilis 168 and Priestia megaterium QM B1551 was identified in 41.1 % of the strains, including all Parageobacillus spp. strains—except for P. thermantarcticus M1^T—as well as G. thermodenitrificans, G. vulcani PSS1, G. subterraneus DSM 23066, and six Geobacillus sp. strains (Y4.1MC1, 44B, 44C, 46C-IIa, 47C-IIb, and PA-3). Additionally, 91.6 % of the strains carried an ortholog of GerX (GerX1) from B. anthracis Ames Ancestor. An ortholog of GerR (GerR1) from B. cereus ATCC 14579^T and ATCC 10987, was found in 83.2 % of the strains, with the exception being P. thermoglucosidasius strains, P. thermantarcticus M1^T, Parageobacillus sp. KH3-4, and Geobacillus sp. Y4.1MC1. Moreover, 25.2 % of the strains contained an ortholog of GerQ (GerQ1) from B. cereus ATCC 14579^T, including G. thermodenitrificans, G. subterraneus, G. uzenensis, G. proteiniphilus, G. vulcani, G. thermocatenulatus, G. icigianus, G. stearothermophilus DG-1, several Geobacillus sp. strains (46C-IIa, 47C-IIb, PA-3, T6, and CAMR5420), and one Parageobacillus strain (P. toebii B4110). Notably, not all GRs identified were complete; in some strains, one or more subunits appeared to be pseudogenes, encoding incomplete proteins.

The GerK1 omnipresent in all studied strains was composed of three subunits—A, B, and C. It was the only ortholog that showed > 90 % query coverage and > 40 % sequence identity across all subunits (reference: GerK orthologs from B. subtilis 168 and Priestia megaterium QM B1551). Subunit A showed the highest similarity, reaching > 57 % identity. GerK2 was less conserved: while subunit A had > 90 % coverage and > 44 % identity, subunits B and C displayed similar coverage (> 90 %) but lower identity (29–31 %). Notably, subunit D from GerK in B. subtilis 168 was absent in all (Para)Geobacillus orthologs, although all strains possessed a hypothetical protein adjacent to GerK1, similar to GerKD and GerUD from Priestia megaterium QM B1551 (coverage > 90 %, identity 30–40 %). Strains containing GerK2 also featured a nearby hypothetical protein resembling Yfk subunit D (YfkS) from B. subtilis 168 (coverage > 90 %, identity 25–41 %). Regarding GerR and GerQ orthologs from B. cereus ATCC 14579^T and ATCC 10987, subunit C was the least conserved (coverage > 90 %; identity 31 % and 29 %, respectively), compared to subunit A (coverage > 90 %; identity 52 % and 46 %, respectively) and subunit B (coverage > 90 %; identity 40 % and 41 %, respectively). The GerXC ortholog was even less conserved, with coverage of 74–78 % and identity of 29–32 %, whereas subunits A and B showed > 90 % coverage and > 42 % identity, (reference: B. anthracis Ames Ancestor GerX). Additionally, certain Geobacillus spp. strains (Geobacillus sp., G. kaustophilus, G. stearothermophilus, G. thermoleovorans, and G. zalihae) presented a single subunit C with low similarity (78 % coverage, 28 % identity) to GerXC from B. anthracis Ames Ancestor, and all P. caldoxylosilyticus strains harbored a single subunit C shorter than usual with low similarity (71 % coverage, 35 % identity) to GerKC from Priestia megaterium QM B1551.

To compare and classify GRs from (Para)Geobacillus and reference strains, phylogenetic trees were constructed separately for each subunit using the sequences listed in Table S4 (Fig. 1, Fig. 2, Fig. 3, Fig. 4). The cladogram trees showed consistent clustering of subunits A, B, and C across all GRs (data not shown), likely due to coevolution, as previously reported in B. cereus (Warda et al. 2017). Phylogenetic analyses revealed variability in GR subunit sequences at both inter- and intraspecific levels within (Para)Geobacillus (Fig. 1, Fig. 2, Fig. 3, Fig. 4). In general, all subunits from Parageobacillus generally clustered separately (bootstrap > 75) from those of Geobacillus across all clades, with a few exceptions: Geobacillus sp. 44C and E263 grouped with P. toebii strains (clades 1–3), and Geobacillus sp. 44B with P. caldoxylosilyticus strains (clades 1–3). Within each genus, strains of the same species typically clustered together (bootstrap > 75) (Fig. 1, Fig. 2, Fig. 3, Fig. 4), except for some G. stearothermophilus strains (e.g., strain 10 in clades 1 and 2; strain 53 in clades 1–3). Geobacillus sp. strains were more dispersed: e.g., Geobacillus sp. Sah69 grouped with G. stearothermophilus strains (clades 1–3); Geobacillus sp. 46C-IIa with G. subterraneus strains (clade 1–3); and Geobacillus sp. MAS1 with G. thermoleovorans and G. kaustophilus strains (clades 1–3). In Parageobacillus spp., P. thermoglucosidasius W-2 grouped with Parageobacillus sp. KH3-4 (clades 1 and 3) rather than with other P. thermoglucosidasius strains. Notably, many subunits from Geobacillus sp. WSUCF1 were identical to those of G. zalihae SURF-189; Geobacillus sp. PA-3 and 47C-IIb to G. thermodenitrificans strains; Geobacillus sp. Y4.1MC1 to P. thermoglucosidasius strains; and Geobacillus sp. A8 to G. thermoleovorans and G. kaustophilus strains (Table S4).

Fig. 1.

Phylogenetic analysis of the A subunit of GRs. The cladogram shows the distribution of the A subunit into three major clades, highlighted in red, blue, and green. The color code is maintained in Fig. 2, Fig. 3, Fig. 4 for clarity. The cladogram illustrates both inter- and intraspecies variability of the A subunit. Orthologs of GerA, GerB, GerG, GerI/H, GerK, GerL, GerQ, GerR/Y, GerS, GerU, GerX, Yfk, and Ynd are labeled.

Fig. 2.

Phylogenetic analysis of the B subunit of GRs. The cladogram shows the distribution of the B subunit into three major clades, highlighted in red, blue, and green. The color code is consistent with that used in Fig. 1. Strains absent in Fig. 1 are highlighted in black. The cladogram illustrates both inter- and intraspecies variability of the B subunit. Orthologs of GerA, GerB, GerG, GerI/H, GerK, GerL, GerQ, GerR/Y, GerS, GerU, GerX, Yfk, and Ynd are labeled.

Fig. 3.

Phylogenetic analysis of the C subunit of GRs. The cladogram shows the distribution of the C subunit into three major clades, highlighted in red, blue, and green. The color code is consistent with that used in Fig. 1. Strains absent in Fig. 1 are highlighted in black. The cladogram illustrates both inter- and intraspecies variability of the C subunit. Orthologs of GerA, GerB, GerG, GerI/H, GerK, GerL, GerQ, GerR/Y, GerS, GerU, GerX, Yfk, and Ynd are labeled.

Fig. 4.

Phylogenetic analysis of the D subunit of GRs. The cladogram shows the distribution of the D subunit in a major clade highlighted in green. The color code is consistent with that used in Fig. 1. Orthologs of GerKD, GerUD, and YfkS are labeled.

3.2. Response of (Para)Geobacillus spp. spores to different germinants

Subsequently, we selected several (Para)Geobacillus strains with distinct GR profiles (Table 1) to evaluate their germination response to a range of germinants previously identified in Bacillus and Clostridium spp. (Fan et al., 2024; Ross and Abel-Santos, 2010; Setlow et al., 2017; Shen et al., 2019). Please note that six out of the 28 collected strains failed to sporulate in our experimental setup, including G. stearothermophilus 10, G. thermocatenulatus DSM 730^T, G. subterraneus KCTC 3922^T, G. uzenensis DSM 13551^T, G. thermoleovorans B23, and P. toebii B4110. To obtain an overview of the most effective stimuli across strains, we quantified the percentage decrease in OD₆₀₀ after 4 h of exposure at 55 °C (Table 2). Given that germination responses in (Para)Geobacillus spp. remain largely uncharacterized and that nutrient preferences vary widely within and between Bacillus and Clostridium spp. (Paredes-Sabja et al., 2011; Setlow et al., 2017), we first assessed germination of the 22 strains in a rich nutrient medium (TSBYE) and a broad range of individual amino acids, nucleosides, sugars, and inorganic salts. Germinant concentrations (all below solubility limits) and the composition of the rich medium were optimized to elicit strong responses in three representative strains—P. thermoglucosidasius DSM 2542^T, G. thermodenitrificans DSM 465^T, and G. stearothermophilus ATCC 12980^T (Table S6)—which collectively harbor all identified GR types (Table 1). Most strains exhibited modest to strong germination in TSBYE (OD₆₀₀ decrease > 30.0 %), with the exception of weak responses (> 15.0–30.0 %) observed in G. stearothermophilus D1 and B4114, and P. genomosp. 1 NUB3621. In contrast to most Bacillus and Clostridium spores (Paredes-Sabja et al., 2011; Setlow et al., 2017), none germinated (≤ 15.0 %) in the presence of l-amino acids, except G. thermodenitrificans DSM 465^T, which showed a weak response (29.5 %) to 30 mM l-phenylalanine. Since the presence of l-alanine racemase in spores can inhibit germination in this hitherto universal germinant (Dodatko et al., 2009; Omotade et al., 2013), the combination of l-alanine with the enzyme inhibitor d-cycloserine was tested in the three representative strains (P. thermoglucosidasius DSM 2542^T, G. thermodenitrificans DSM 465^T, and G. stearothermophilus ATCC 12980^T), with no positive results (Table S6). In addition, none of the 22 strains responded to nucleosides (≤ 15.0 %), which typically induce germination in B. cereus group (Hornstra et al., 2006; Warda et al., 2017).

Table 2.

Germination response of the indicated (Para)Geobacillus spp. strains to different germinants after 4 h exposure at 55 °C. The extent of germination (measured as OD₆₀₀ decrease) was scored as: insensitive (≤ 15.0 %), weak (> 15.0–30.0 %), modest (> 30.0–45.0 %), and strong (> 45.0 %).

Germinant	G. ste. A1	G. ste. D1	G. ste. P1	G. ste. B4109	G. ste. B4114	G. ste. DSM 458	G. ste. ATCC 7953	G. ste. ATCC 12980T	G. ste. ATCC 10149	G. thd. DSM 465 T	G. thd. G11MC16	G. thl. KCTC 3570 T	G. kau. NBRC 102445T	P. gen. NUB3621	P. thg. DSM 2543	P. thg. DSM 2542 T	P. thg. C56-YS93	P. thg. M10-EXG	P. thg. DSM 6285	P. cal. DSM 12041 T	P. cal. B4119	P. toe DSM 14590T
TSBYE	36.7 (4.7)	23.2 (2.4)	38.0 (6.0)	39.3 (3.5)	23.6 (3.5)	34.4 (6.0)	> 45.0	> 45.0	42.1 (7.0)	> 45.0	> 45.0	> 45.0	32.4 (7.1)	25.8 (6.6)	40.5 (4.7)	> 45.0	38.8 (4.2)	> 45.0	> 45.0	> 45.0	> 45.0	> 45.0
l-Alanine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Glycine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Valine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Isoleucine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Methionine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Leucine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Serine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Proline (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Cysteine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Asparagine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Glutamine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Threonine (1 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Aspartic acid (10 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Glutamic acid (25 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Histidine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Lysine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Arginine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Phenylalanine (30 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	29.5 (7.8)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Tryptophan (25 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
l-Tyrosine (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
CA (0.2 %)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	26.7 (6.2)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	23.9 (5.3)	≤ 15.0	≤ 15.0	≤ 15.0	34.6 (1.6)	≤ 15.0	≤ 15.0
Adenosine (7.5 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
Inosine (25 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
d-Glucose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	44.8 (1.9)	39.9 (0.7)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	40.8 (6.0)	> 45.0	> 45.0	34.6 (4.4)	32.0 (5.5)	≤ 15.0	≤ 15.0	≤ 15.0
d-Fructose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	36.9 (5.7)	35.9 (1.3)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	36.3 (1.3)	> 45.0	40.7 (8.9)	34.2 (1.7)	33.9 (6.4)	≤ 15.0	≤ 15.0	≤ 15.0
d-Galactose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	33.9 (5.8)	≤ 15.0	20.7 (5.9)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	21.5 (5.1)	≤ 15.0	32.7 (11.7)	≤ 15.0	27.5 (4.0)	≤ 15.0	≤ 15.0
d-Mannose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	> 45.0	> 45.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	44.2 (6.9)	> 45.0	> 45.0	33.2 (4.5)	30.4 (1.2)	≤ 15.0	≤ 15.0	≤ 15.0
d-Ribose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
d-Arabinose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	42.3 (5.0)	41.2 (2.4)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	40.5 (1.8)	> 45.0	42.6 (6.7)	28.2 (6.0)	30.2 (1.6)	≤ 15.0	≤ 15.0	≤ 15.0
d-Xylose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	38.3 (5.5)	37.0 (0.8)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	36.8 (3.6)	> 45.0	39.3 (10.7)	24.3 (8.2)	36.2 (6.3)	≤ 15.0	≤ 15.0	≤ 15.0
Lactose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	27.0 (6.3)	26.7 (6.1)	31.1 (1.0)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	31.0 (3.0)	39.5 (2.1)	24.3 (8.5)	≤ 15.0	29.6 (4.0)	≤ 15.0	≤ 15.0	≤ 15.0
Maltose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	> 45.0	41.7 (11.7)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	43.5 (3.2)	> 45.0	47.4 (2.0)	32.9 (4.3)	41.1 (10.9)	≤ 15.0	≤ 15.0	≤ 15.0
Sucrose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	44.4 (3.4)	38.6 (5.3)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	44.3 (3.0)	> 45.0	42.5 (5.8)	41.7 (4.9)	44.3 (2.0)	≤ 15.0	≤ 15.0	≤ 15.0
Raffinose (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	41.1 (7.2)	> 45.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	38.0 (2.2)	> 45.0	> 45.0	24.8 (9.0)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
KI (50 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
KBr (50 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	31.2 (0.6)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
KCl (50 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	30.9 (2.2)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0
AGFK (100 mM)	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	> 45.0	24.9 (1.7)	> 45.0	≤ 15.0	≤ 15.0	≤ 15.0	≤ 15.0	22.5 (2.7)	> 45.0	34.1 (1.0)	21.7 (7.9)	≤ 15.0	25.8 (5.8)	≤ 15.0	≤ 15.0
CaDPA (60 mM)	> 45.0	37.1 (4.2)	> 45.0	> 45.0	42.2 (1.5)	> 45.0	> 45.0	> 45.0	> 45.0	> 45.0	> 45.0	> 45.0	> 45.0	36.4 (4.9)	> 45.0	> 45.0	> 45.0	41.5 (3.2)	44.8 (7.6)	35.8 (6.1)	> 45.0	43.2 (10.1)

CA, casamino acids; AGFK, mixture of l-asparagine, d-glucose, d-fructose, and KCl.

Regarding sugars, only five strains of P. thermoglucosidasius, two strains of G. stearothermophilus (ATCC 12980^T and ATCC 10149), and G. thermodenitrificans DSM 465^T germinated to a varying degree (> 15.0 – > 45.0 %) in the presence of all or several pentoses (d-arabinose and d-xylose), hexoses (d-glucose, d-fructose, and d-mannose), disaccharides (lactose, maltose, and sucrose), and the trisaccharide raffinose. The most effective inducers were d-glucose, d-fructose, d-mannose, maltose, and sucrose, which triggered germination (> 30.0 %) in the eight strains. In addition, most germinated in d-arabinose and d-xylose to the same extent (> 30.0 %), except for P. thermoglucosidasius M10-EXG (> 15.0–30.0 %). Notably, P. thermoglucosidasius DSM 2542^T and G. stearothermophilus ATCC 12980^T exhibited the strongest responses to most sugars, germinating efficiently even at a reduced concentration (1 mM) of d-glucose, d-mannose, and maltose with no significant difference (p > 0.05) compared to the maximum concentration tested (100 mM) (Table S6).

As inorganic salts, especially those containing K⁺, have been shown to trigger germination in some food isolates of Clostridium perfringens and Priestia megaterium via GRs (Christie and Lowe, 2008; Paredes-Sabja et al., 2008), we also evaluated the response of all strains to KI, KBr, and KCl. Only G. thermodenitrificans DSM 465^T spores germinated but weakly (∼ 30.0 %) in KBr and KCl (Table 2).

In Bacillus and Clostridium spp., the combination of certain germinants has shown to induce germination more effectively than individual components alone, even when used at lower concentrations than those required to elicit maximal responses individually (Paredes-Sabja et al., 2011; Setlow et al., 2017). Hence, we tested the effectiveness of typical combinations in stimulating germination in three representative strains—P. thermoglucosidasius DSM 2542ᵀ, G. thermodenitrificans DSM 465ᵀ, and G. stearothermophilus ATCC 12980ᵀ—at varying concentrations. These included mixtures of amino acids (such as DENQ l-asparagine, and l-glutamine) and casamino acids (CA)), amino acids and nucleosides (such as l-alanine and inosine), amino acids and inorganic salts (such as CA with K⁺), and all three components together (such as AGFK: d-glucose, l-proline, l-leucine, and KBr) (Table S6). None of them showed enhanced response compared with the single germinants, except in CA and AGFK. In CA (0.2 %), G. thermodenitrificans DSM 465^T and P. thermoglucosidasius DSM 2542^T spores exhibited an OD₆₀₀ decrease between > 15.0 % and 30.0 %. While the former strain germinated equally (p > 0.05) in CA (0.2 %) than in l-phenylalanine (30 mM), the latter one was insensitive to all individual amino acids suggesting the existence of a synergetic effect. In view of these results, we examined CA-induced germination in the 22 strains studied; however, only P. caldoxylosilyticus DSM 12041^T spores showed a moderate response (34.6 %) (Table 2).

Regarding AGFK, it was the most efficient combination tested, with the three representative strains germinating efficiently (> 45.0 %) regardless of the concentration used. In fact, G. thermodenitrificans DSM 465^T spores displayed a significantly larger (p ≤ 0.05) OD₆₀₀ decrease in AGFK (100 mM; 53.9 %) compared to d-glucose and d-fructose individually at the same concentration (39.9 % and 35.9 %, respectively) and to KCl at 50 mM (30.9 %), while the extent of germination of the other two strains in the mixture reached the same level than at least one of the single germinants (Table S6). Based on this observation and considering that AGFK is a commonly used germinant in B. subtilis (Chen et al., 2024; Freire et al., 2024; Stewart et al., 2012), its effect was further evaluated in the remaining 22 strains. Five additional strains responded to AGFK (G. stearothermophilus ATCC 10149 and four P. thermoglucosidasius strains) but to a germination extent equal or even lower than the individual components, except in P. caldoxylosilyticus DSM 12041^T which OD₆₀₀ decreased lightly (25.8 %) in the mixture while not responding to each single germinant (≤ 15.0 %) (Table 2). Interestingly, one P. thermoglucosidasius strain (DSM 6285) did not respond to AGFK, despite showing a moderate response (> 30.0 %) to both d-glucose and d-fructose individually (Table 2).

Previous research has reported that the rate and/or extent of nutrient-induced germination in G. stearothermophilus spores can be enhanced by prior heat or NaNO₂ activation (Foerster, 1983, 1985; Zhou et al., 2013). Thus, we tested the effect of heat and NaNO₂ treatments prior to exposure to relevant inducers (d-glucose, maltose, and CA) in P. thermoglucosidasius DSM 2542^T, G. thermodenitrificans DSM 465^T, and G. stearothermophilus ATCC 12980^T spores, which were insensitive to amino acids and nucleosides but responded well to some sugars and/or inorganic salts, and in five G. stearothermophilus strains (A1, D1, P3, ATCC 7953, and B4109) and one P. caldoxylosilyticus strain (B4119), which were virtually unable to germinate in any stimuli. None of the activation treatments improved the rate and final extent of OD₆₀₀ decrease in any case (data not shown).

The ability of exogenous Ca-DPA to trigger spore germination was also assessed in all strains. Ca-DPA has been shown to induce germination in a GR-independent manner by activating the CLE CwlJ in Bacillus and Clostridium spp. (Løvdal et al., 2012; Paidhungat et al., 2001; Wang et al., 2017). All strains germinated with an OD₆₀₀ decrease superior to 30.0 %, including those poorly responsive to nutrient inducers (Table 2), consistent with previous findings in G. stearothermophilus ATCC 7953 (Georget et al., 2015) and the presence of B. subtilis CwlJ orthologs in several Geobacillus strains (Paredes-Sabja et al., 2011).

3.3. Colony-forming efficiency of (Para)Geobacillus spp. spores

The colony-forming efficiency of the 22 (Para)Geobacillus strains was assessed on TSAYE at 55 °C (48 h). The percentage of total spores counted microscopically that were able to form a visible colony on agar plates varied from 2.6 % to 52.5 % (Fig. 5). The colony-forming efficiency of all (Para)Geobacillus spp. strains differed significantly (p ≤ 0.05) from that of the easily cultivable B. subtilis 168 spores (99.5 %, 37 °C, 24 h; Fig. 5), considering technical variability, primarily due to variations in plate counts (data not shown; coefficient of variation of 0.06). (Para)Geobacillus spp. strains could be divided into two significantly different (p ≤ 0.05) groups: one composed of five strains with ≤ 10.0 % of their spores forming colonies (G. stearothermophilus ATCC 7953, G. thermoleovorans KCTC 3570^T, P. genomosp.1 NUB3621, P. caldoxylosilyticus DSM 12041^T, and P. toebii DSM 14590^T), and the other composed of two strains with an average between > 40.0 % to 60.0 % of their spores forming colonies (P. thermoglucosidasius DSM 2542^T and DSM 2543). The remaining 15 strains showed an intermediate ability to form colonies (> 10.0 – ≤ 40.0 %).

Fig. 5.

Colony-forming efficiency (expressed as the percentage of total spores, determined microscopically, that form a colony on TSAYE plates (55 °C, 48 h)) of the indicated (Para)Geobacillus strains. All strains were sporulated at 55 °C for 4 days. For comparison, the colony-forming efficiency of B. subtilis 168 spores was included (NAYE, 37 °C, 24 h). Data represent mean values and standard deviations calculated from three biological replicates. Different letters indicate statistically significant differences (p ≤ 0.05) among the strains. Different colors categorize strains according to their colony-forming efficiency as low (≤ 10.0 %, light blue), medium (> 10.0 % - 40.0 %, medium blue), high (> 40.0 %, dark blue), and extremely high (> 97.5 %, black).

Interestingly, species from each genus were distributed across all three groups, with considerable variability among strains of certain species. For instance, G. stearothermophilus ATCC 7953 and P. caldoxylosilyticus DSM 12041^T showed a lower (p ≤ 0.05) percentage of recovered spores compared to the modest response of other strains within the same species, and the colony-forming ability of P. thermoglucosidasius C56-YS93 differed significantly (p ≤ 0.05) from that of the strains displaying the highest values (DSM 2542^T and DSM 2543). Intriguingly, colony-forming efficiency (TSAYE, 48 h) was significantly lower (p ≤ 0.05) than germination efficiency determined microscopically (TSBYE, 4 h) in most strains (Table S7). For instance, Fig. 6 illustrates the colony-forming and germination efficiencies of six selected strains, one representative strain producing spores with high colony-forming efficiency (> 40.0 %, P. thermoglucosidasius DSM 2542^T), two strains producing spores with intermediate colony-forming efficiency (> 10.0 – ≤ 40.0 %, G. stearothermophilus ATCC 12980^T and G. thermodenitrificans DSM 465^T), and three producing spores with low colony-forming efficiency (≤ 10.0 %, G. stearothermophilus ATCC 7953, G. thermoleovorans KCTC 3570^T, and P. genomosp. 1 NUB3621). Remarkably, G. stearothermophilus ATCC 7953 and G. thermoleovorans KCTC 3570^T spores, with a colony-forming efficiency of ca. 3.0 % (± 1.2 %), exhibited a fraction of germinated spores in TSBYE of ≥ 97.0 % (Fig. 6, Table S7). This pattern was observed in four out of the five strains categorized as low colony formers (≤ 10.0 %), except for P. genomosp. 1 NUB3621. In addition, the six strains were treated with CaDPA prior to plating on TSAYE. Although 4 h of exposure to CaDPA led to ∼ 80.0 – ≥ 97.0 % germination, it did not increase colony-forming efficiency in any strain; in fact, it further reduced spore counts, especially in strains with intermediate and high colony-forming efficiencies (Fig. 6). It has been observed that the lack of both CLEs CwlJ and SleB in B. subtilis spores reduced the ability to form colonies by > 5 log cycles, and that spores exposed to l-alanine retain some refractility (Ishikawa et al., 1998). The fact that both GR- and CLE-inducers resulted in a relatively high fraction of apparently dark and hydrated spores in a short time window suggests that the inability of a large subpopulation of (Para)Geobacillus spores to form colonies may be associated to a defect in outgrowth (i.e., from initiation of metabolism to proliferation to form a visible colony) rather than germination completion (i.e., from activation of germination to complete rehydration).

Fig. 6.

Comparison of the percentage of total spores recovered on TSAYE plates (55 °C, 48 h), with (light blue) or without (light green) previous exposure to CaDPA (60 mM, 55 °C, 4 h), and the percentage of germinated spores determined by phase-contrast microscopy in TSBYE (55 °C, 4 h; dark green) and in CaDPA (60 mM, 55 °C, 4 h; dark blue) in the indicated (Para)Geobacillus spp. strains. All strains were sporulated at 55 °C for 4 days. Data represent mean values and standard deviations calculated from three biological replicates. An asterisk above the bars representing the percentage of germination in TSBYE (dark green) indicates statistically significant differences (p ≤ 0.05) compared to the colony-forming efficiency on TSAYE without CaDPA exposure (light green), while an asterisk above the bars representing germination in CaDPA (dark blue) indicates statistically significant differences (p ≤ 0.05) compared to colony-forming efficiency on TSAYE after CaDPA exposure (light blue). The dotted line indicates the limit of quantification (≥ 97 %).

Spore outgrowth and/or colony-forming efficiency depends on recovery conditions such as incubation temperature (Eijlander et al., 2019; McGuiggan et al., 1994; Wells-Bennik et al., 2019), with the optimal and permissive temperature range for growth varying widely among (Para)Geobacillus species (Logan et al., 2015). Furthermore, limits for growth inhibition are commonly more sensitive to temperature changes than the ones for germination inhibition, so that spores may germinate at extreme temperatures without resuming growth (Freire et al., 2024; Smoot and Pierson, 1982). Therefore, we tested whether incubation temperatures (45–70 °C) affected spore recovery in three selected strains (P. thermoglucosidasius DSM 2542^T, G. thermodenitrificans DSM 465^T, and G. stearothermophilus ATCC 12980^T) with varying ranges of growth temperatures (BacDive, 2025). Although incubation temperature affected the proportion of recovered spores differently in each strain, there were no statistically significant differences (p > 0.05) between spore counts obtained at 55 °C and those obtained at 50 °C and 70 °C in any strain (Fig. S1). In addition, sporulation temperatures far from the optimum and extended maturation times influence spore outgrowth (Liang et al., 2019; Nguyen Thi Minh et al., 2011; Trunet et al., 2020). Varying sporulation temperature (50–65 °C) or maturation time (1–7 days) did not affect (p > 0.05) the recovery of P. thermoglucosidasius DSM 2542^T, G. thermodenitrificans DSM 465^T, and G. stearothermophilus ATCC 12980^T (Fig. S2 and S3). These results indicate that the low colony-forming efficiency of all (Para)Geobacillus spores, as well as their inter- and intraspecific variability, may not be explained by suboptimal physiological requirements for growth. Nevertheless, it is possible that the TSAYE medium lacks one or more nutrients specifically required for spore outgrowth, or that differences related to the physical state of the medium may influence the efficiency of germination.

4. Discussion

As (Para)Geobacillus spores are relevant food spoilers while also offering biotechnological benefits upon germination (Alonso et al., 2021; Berendsen et al., 2016; Kochhar et al., 2022), identifying and characterizing their nutrient-sensing GRs is essential. Here, we conducted, for the first time, an in silico analysis of GRs in 105 publicly available genomes and combine it with experimental germination assays on 22 representative strains using various inducers.

All strains harbored between two and five GR operons, orthologs to different GR types found in Bacillus spp. and Priestia megaterium: GerK orthologs from B. subtilis 168 and GerK and GerU from Priestia megaterium QM B1551, GerX from B. anthracis Ames Ancestor, GerR from B. cereus ATCC 14579^T and ATCC 10987, and GerQ from B. cereus ATCC 14579^T. Remarkably, the GerK1 ortholog showed the highest similarity among all identified GR types and was present in all strains, suggesting a key role in germination. In contrast, the second GerK ortholog (GerK2), found in some (Para)Geobacillus strains, was more distantly related to the references and may have originated from a duplication of GerK1 followed by evolutionary adaptation. Some Geobacillus spp. and P. caldoxylosilyticus strains also exhibited an additional subunit C, present in the absence of the other subunits, which showed similarity to GerXC from B. anthracis Ames Ancestor and to GerKC from Priestia megaterium QM B1551, respectively. The presence of single GR subunits has also been described in B. cereus food isolates (Warda et al., 2017), which presented a GerXC ortholog, although its function remains unknown.

GRs are typically composed of three subunits, and even though all appear to be necessary for proper germination (Moir, 2006; Moir et al., 2002), the precise function of each subunit is not fully understood. Recently, it has been reported that the subunit B of B. subtilis GerA plays a role in l-alanine recognition (Artzi et al., 2021; Gao et al., 2023; Wang et al., 2020) and that the subunit A is involved in transducing the nutrient signal (Amon et al., 2022; Gao et al., 2023). Additionally, GerAB forms a water channel, reinforcing its contribution in water uptake during the initial stages of germination (Chen et al., 2024). In contrast, the role of subunit C remains less well defined, likely due to its unique protein structure (Christie and Setlow, 2020). Some GRs possess an additional subunit D, such as GerK and Yfk (termed as GerKD and YfkS, respectively) in B. subtilis 168 and GerK₁ and GerU in Priestia megaterium QM B1551 (termed as GerKD and GerUD, respectively) (Ramirez-Peralta et al., 2013), although it is not present in their orthologs found in B. licheniformis ATCC 14580^T (Borch-Pedersen et al., 2016) and Clostridium perfringens SM101 (Banawas et al., 2013). Subunits D are believed to modulate GR response to nutrients. In B. subtilis 168 and Priestia megaterium QM B1551, deletion of gerKD has been shown to improve germination, whereas knockout of gerUD in Priestia megaterium QM B1551 abolishes it (Ramirez-Peralta et al., 2013). In the case of (Para)Geobacillus spp. strains, a potential gene encoding a subunit D was found adjacent to gerK1 and gerK2 operons. The subunits A identified in (Para)Geobacillus genomes were highly conserved compared to the subunits B and C, as well as subunit D in GerK orthologs. This may reflect that GRs have evolved independently in each species, with (Para)Geobacillus spp. potentially adapting their function to nutrients and environmental conditions in their natural niches while preserving subunits A, likely due to its critical role in signal transduction. GR-type profiles and subunit sequences display both interspecific and intraspecific variability. Indeed, phylogenetic trees showed that strains from the same genus usually clustered together and the same observations were made for the different species, although this was not always the case, particularly for G. stearothermophilus strains. This agrees with previous works that observed large differences among G. stearothermophilus genomes, especially in those isolated from food (Burgess et al., 2017). However, it should be noted that the number of available genomes for G. stearothermophilus was higher than for other species.

In addition, phylogenetic trees illustrated that unclassified (Para)Geobacillus strains clustered with species from the same or even different genus, coinciding with reclassifications previously proposed by other authors. For instance, Geobacillus sp. PA-3 clustered with G. thermodenitrificans (Burgess et al., 2017), Geobacillus sp. Sah69 clustered with G. stearothermophilus (Burgess et al., 2017), Geobacillus spp. Y4.1MC1 clustered with P. thermoglucosidasius (Burgess et al., 2017; Mol and de Maayer, 2024), Geobacillus sp. 44C and E263 clustered with P. toebii (Mol and de Maayer, 2024), Geobacillus sp. 44B clustered with P. caldoxylosilyticus (Mol and de Maayer, 2024), and P. thermoglucosidasius W-2 clustered with Parageobacillus sp. KH3–4, and both were proposed to form a novel species, termed P. genomosp. A (Mol and de Maayer, 2024). The discordance in classification at the genus level could be related to the fact that Geobacillus strains were originally classified within the Bacillus genus (Ash et al., 1991; Nazina et al., 2001), and later Geobacillus was divided into two clades (Aliyu et al., 2016). Furthermore, taxonomy evaluation and reclassifications are continuously suggested. A recent study proposed to divide Parageobacillus clade into Parageobacillus and Saccharococcus, suggesting changing P. caldoxylosilyticus strains to S. caldoxylosilyticus and P. genomosp. 1 NUB3621 to S. genomosp. nov. A NUB3621 (Mol and de Maayer, 2024). In fact, phylogenetic trees constructed using the same GR sequences as in Fig. 1, Fig. 2, Fig. 3, with the addition of sequences from the Saccharococcus type strain (DSM 4749^T), showed that it clusters with Parageobacillus strains (Figs. S4–S6). However, based on GRs alone, it appears more closely related to P. toebii than P. caldoxylosilyticus.

Phenotypically, the most striking result was that most strains did not germinate in the presence of l-amino acids―even l-alanine, which is considered a universal germinant in spore-formers (Ross and Abel-Santos, 2010)―nucleosides, or their combinations. These include strains that harbored a GerR and/or GerQ-like GRs, which are involved in germination with many l-amino acids, purine nucleosides, and the combination of l-alanine and inosine in B. cereus ATCC 14579^T (Hornstra et al., 2006; Hornstra et al., 2005). The insensitivity of (Para)Geobacillus spp. strains to l-amino acids―alone or combined with inosine or sugars―even after heat activation, has also been observed in G. stearothermophilus ATCC 7953 (Georget et al., 2015). However, other authors have reported that the strain G. stearothermophilus NGB101 germinated in l-valine (1 mM), but only after prior heat activation (Foerster, 1983; Zhou et al., 2013).

Sugars were the major nutrient germinants in two G. stearothermophilus strains (ATCC 12980^T and ATCC 10149), G. thermodenitrificans DSM 465^T, and all P. thermoglucosidasius strains. Interestingly, all these strains responded to various sugars to different extents, whereas the remaining strains were unresponsive to any of them. To facilitate comparison between GR profiles and germination phenotypes, phylogenetic trees were constructed using only the (Para)Geobacillus strains tested in the germination assays, along with reference strains (Figs. S7–S10). Sugar response did not correlate with the presence of specific GR-type orthologs (Figs. S7–S10). All strains harbored one or even two orthologs of B. subtilis 168 GerK, whose homologs, including Priestia megaterium QM B1551 GerU and GerK₁, and B. licheniformis ATCC 14580^T GerK, have been demonstrated to react against glucose (Borch-Pedersen et al., 2016; Christie and Lowe, 2007; Gupta et al., 2013; Paidhungat et al., 2000). However, despite the genetic proximity among GerK orthologs, their germination phenotypes differ. For instance, while B. licheniformis ATCC 14580^T GerK shows a weak response to d-glucose (Borch-Pedersen et al., 2016) and Priestia megaterium QM B1551 GerK₁ only induces moderate germination with d-glucose plus KBr (Gupta et al., 2013), Priestia megaterium QM B1551 GerU triggers efficient germination with d-glucose, l-leucine, l-proline, KBr alone, and combined (GPLK) (Christie and Lowe, 2007, 2008; Ustok et al., 2014), and B. subtilis PS832 GerK requires GerB cooperation to germinate in AGFK (Paidhungat et al., 2000). Subtle amino acids substitutions in GR sequences may explain differences in GerK-mediated germination among (Para)Geobacillus spp. and other mesophilic species, as observed in the GerVB subunit of Priestia megaterium, leading to altered responses to GPLK and d-glucose (Christie and Lowe, 2008). Conversely, the absence of a GerB ortholog in all (Para)Geobacillus strains may account for the lack of an enhanced response to AGFK compared to d-glucose or d-fructose alone, as suggested for Caldibacillus thermoamylovorans (Berendsen et al., 2015).

Although a few strains responded exclusively to sugars, most germinated efficiently in TSBYE (∼ 45.0 – ≥ 97.0 % germinated spores) for the same exposure time, except for some G. stearothermophilus strains (D1 and B4114) and P. genomosp. 1 NUB3621 (∼ 12.0 % – 18.0 % germinated spores). This medium contained many of the individual and combined germinants tested, such as d-glucose and l-alanine (Tao et al., 2023). Therefore, although the precise identity of the inducing compounds in TSBYE remains unknown, we cannot exclude the possibility that the observed germination is due to synergistic or combinatorial effects of multiple amino acids and other nutrients present in the medium. No correlation was observed between GR diversity and germination efficiency in this medium. Strikingly, the strains showing limited germination in TSBYE possessed three to four complete GRs. Similar observations were made in poorly germinating Caldibacillus thermoamylovorans spores with two GR operons (Berendsen et al., 2015), and in certain B. cereus harboring up to nine GR operons (Warda et al., 2017).

A comprehensive comparison of GR-type profiles and their phylogenetic similarity with germination phenotypes among (Para)Geobacillus spp. strains also showed discrepancies. For instance, in G. stearothermophilus, strains ATCC 12980^T and ATCC 10149 shared GR types that clustered together and exhibited similar responses to nutrients. Although strain ATCC 7953 had GRs identical to those of ATCC 12980^T and/or ATCC 10149, it was unresponsive to all single nutrients. Nevertheless, all three strains germinated in TSBYE. Strains B4114 and B4109 harbored different GR profiles―the latter lacked GerX1 and possessed a pseudo subunit GerK1A, while both shared GerR1—, but B4109 germinated more efficiently in TSBYE. In the case of Parageobacillus, all P. thermoglucosidasius strains showed common GR orthologs, with all subunit sequences clustering together, and displayed virtually similar germination response. In contrast, the GRs of P. caldoxylosilyticus strains also clustered together, but strain DSM 12041^T responded weakly to CA and d-galactose, unlike strain B4119. Notably, B4119 harbored a gerR1 operon, although its subunit B was a pseudogene. The GRs of P. toebii DSM 14590^T and P. genomosp. 1 NUB3621 were closely related, but only the former responded extensively to TSBYE.

Other previous studies did not find a correlation between germination response with the presence and/or similarity of different GRs within B. cereus (Warda et al., 2017), B. licheniformis (Madslien et al., 2014), and Caldibacillus thermoamylovorans (Berendsen et al., 2015), which has been attributed to a large number of factors: (i) different requirements in the type of individual and combined nutrients and their concentrations, although varying the concentration of most common studied nutrients and its combination did not improve germination in our case; (ii) variations in the expression levels of GRs (Cabrera-Martinez et al., 2003; Chen et al., 2014; Griffiths et al., 2011), (iii) or in enzymes degrading inducing germinants, such as l-alanine racemase in B. cereus sensu lato (Dodatko et al., 2009; Omotade et al., 2013); (iv) the presence of GRs of yet unknown function such as Yfk and Ynd in B. subtilis PS832 (Paidhungat and Setlow, 2000) or GerX in B. anthracis Sterne 34F₂ (Carr et al., 2010b); (v) subtle differences in GR subunit sequences leading to changes or inhibition of their activity (Artzi et al., 2021; Christie and Lowe, 2008; Krawczyk et al., 2017; Madslien et al., 2014; Mongkolthanaruk et al., 2011; Paredes-Sabja et al., 2008); (vi) the presence of redundant GRs, as occurs in B. cereus ATCC 14579^T GerR, GerQ, and GerI (Abel-Santos and Dodatko, 2007; Hornstra et al., 2006); (vii) variations in nutrient permeability, for instance, derived from coat structure (Behravan et al., 2000; Butzin et al., 2012; Carr et al., 2010a; Isticato et al., 2020; Saggese et al., 2022; Saggese et al., 2016); (viii) variations in the surrounding IM environment of GRs affecting their function, as has been suggested in B. subtilis mutant spores with deletions in genes encoding predicted IM-embedded proteins (ylag, ychN, yetF, and ydfS) (Johnson and Moir, 2017; Yu et al., 2023), or due to the presence of the 2Duf-encoding gene within the spoVA^2mob transposable element (Berendsen et al., 2015; Berendsen et al., 2016; Korza et al., 2023; Krawczyk et al., 2017); (ix) alterations in the downstream germination steps from GR-nutrient binding, which in our case it would not involve variations in CLEs activity since all strains tested responded efficiently to CaDPA. In addition, (x) different ecological or methodological requirements among strains to respond to certain germinants, such as sporulation conditions (like composition of the medium, temperature, maturation time), spore purification method, the need of heat or chemical activation, and germination conditions (temperature, pH, and germinant concentration) may contribute to the disagreement between GR profiles and germination phenotypes (Bressuire-Isoard et al., 2018; Garcia et al., 2010; Ghosh and Setlow, 2010; Rose et al., 2007; Salvador et al., 2025). In this study, sporulation and germination conditions were optimized based on the type strains of P. thermoglucosidasius, G. thermodenitrificans, and G. stearothermophilus (Salvador et al., 2025), which may not be optimal for all strains. Despite these limitations, this approach enabled us to outline the distribution of GRs among different (Para)Geobacillus strains and compare them with well-studied spore-forming strains, while showing their ability to respond to nutrient and non-nutrient germinants.

Another relevant finding of this study was the low colony-forming rate of (Para)Geobacillus spp. spores on a standard rich cultivation medium, varying largely among species and strains (between 2.6 % and 52.5 %). This variation is unlikely to result from differences in physiological requirements for growth or sporulation. These values contrast with the generally above 95 % colony-forming efficiency in well-characterized mesophilic Bacillus spp., such as B. subtilis (Ghosh and Setlow, 2009) (Fig. 5) and B. anthracis (Carr et al., 2010b). A reduced rate of recovery was previously described in G. stearothermophilus ATCC 7953 (10 %) (Huesca-Espitia et al., 2016) and in Caldibacillus thermoamylovorans strains (≤ 5 %) (Berendsen et al., 2015). Thus, the reduced colony-forming efficiency may be an inherent trait in both (Para)Geobacillus genera, and perhaps in other thermophilic bacteria.

Our data suggest that the reduced colony-forming efficiency in most (Para)Geobacillus strains may be attributed to an outgrowth defect (i.e., the transition from initiation of metabolism to proliferation and visible colony formation), rather than germination defect (i.e., the transition from activation of germination to complete rehydration). This is supported by the observation that most strains germinated efficiently in rich growth medium within a short period, yet only a small fraction was able to form visible colonies under optimal recovery conditions. However, we cannot rule out the possibility that TSAYE medium lacks one or more nutrients essential for spore outgrowth or that the transition from liquid to solid nutrient medium negatively affected spore germination. Furthermore, exposure to CaDPA, which efficiently induced germination—in some strains even to a greater extent than liquid growth medium—failed to restore viability on plates. This contrasts with previous findings in B. subtilis mutants lacking all functional GRs and in Caldibacillus thermoamylovorans strains (Berendsen et al., 2015; Paidhungat and Setlow, 2000). Nevertheless, it has been observed that deletion of the sole GR present in Clostridium perfringens SM101 only slightly reduced the spores’ ability to germinate in rich liquid medium but severely impaired the ability to form colonies on plates, suggesting that GerK, in Clostridium perfringens, may be involved in both germination and subsequent outgrowth (Paredes-Sabja et al., 2008). Closer examination of the entire spore revival process at the cellular and molecular level will be necessary to elucidate the mechanism impeding colony formation and, consequently, to develop strategies that avoid underestimation of (Para)Geobacillus spp. spores during enumeration.

5. Conclusions

This study presents the first genotypic and phenotypic analysis of GR profiles and their germinant responses in (Para)Geobacillus spp. spores. Comparative genomics identified five GR types across all strains, including orthologs of B. subtilis, Priestia megaterium, B. cereus, and B. anthracis, though their functions in nutrient-induced germination may have diverged. GR ortholog types varied among species and, in some cases like G. stearothermophilus, among strains, with only a B. subtilis and Priestia megaterium GerK ortholog consistently present. None of the strains responded to typical amino acids and nucleosides, and only a few (G. stearothermophilus ATCC 12980^T and ATCC 10149, G. thermodenitrificans DSM 465^T, and five P. thermoglucosidasius strains) germinated in response to sugars. However, all strains germinated to varying degrees in nutrient-rich TSBYE, suggesting unidentified compounds may act as inducers. No clear correlation was found between GR gene content and germination behavior, complicating predictions based solely on genotype. Additionally, most (Para)Geobacillus spp. exhibited lower colony-forming efficiency than mesophilic Bacillus spp., a trait that, in most cases, was not linked to reduced germination, indicating that conventional plating techniques may underestimate spore loads due to defects in germination and/or outgrowth.

Funding

This study was supported by grants from MCIN/AEI/10.13039/501100011033 (PID2023-148505OB-I00), the Government of Aragón-FEDER - E35_23R for “Building Europe from Aragón”, and a Ph.D. scholarship from the Government of Aragón (to M.S.).

CRediT authorship contribution statement

Maika Salvador: Conceptualization, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Inmaculada Yruela: Formal analysis, Investigation, Methodology, Supervision, Writing – original draft, Writing – review & editing. Santiago Condón: Conceptualization, Supervision, Funding acquisition, Writing – review & editing. Elisa Gayán: Conceptualization, Formal analysis, Methodology, Project administration, Resources, Funding acquisition, Supervision, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.