The strain N3^T was isolated from the feces of a river otter in Castril, Granada, southern Spain (37°52′00″N 2°45′58″W). One gram of feces was suspended in 0.9% (w/v) saline solution to a final volume of 10 mL. A volume of 0.1 mL of the sample was then plated on a tryptone soy agar (TSA) medium and incubated at 28°C for 7 days. The different isolated colonies were subsequently plated and purified on the same medium.

The strain N3^T and the phytopathogenic strains Erwinia amylovora CECT 222^T and Dickeya solani LMG 25993^T were grown in a tryptic soy broth (TSB) medium. Agrobacterium tumefaciens NTL4 (pZLR4) was grown in a Luria Bertani (LB) medium supplemented with 2.5 mmol L⁻¹ CaCl₂ 2H₂O,2.5 mmol L⁻¹ MgSO₄ 7H₂O (LB/MC), and gentamicin (Gm) to a final concentration of 50 μg mL⁻¹. Chromobacterium violaceum CV026 and C. violaceum VIR07 were grown in an LB medium supplemented with 50 μg mL⁻¹ kanamycin (Km). All the strains were grown at 28°C at 100 rpm in a rotary shaker.

The synthetic AHLs (Sigma-Aldrich, Saint Louis, USA) used were as follows: C4-HSL (N-butyryl-DL-homoserine lactone), C6-HSL (N-hexanoyl-DL-homoserine lactone), 3-O-C6-HSL (N-3-oxo-hexanoyl-DL-homoserine lactone), C8-HSL (N-octanoyl- DL -homoserine lactone), 3-O-C8-HSL (N-3-oxo-octanoyl-DL-homoserine lactone), C10-HSL (N-decanoyl-DL-homoserine lactone), 3-OH-C10-HSL (N-3-hydroxydecanoyl-DL-homoserine lactone), C12-HSL (N-dodecanoyl-DL-homoserine lactone), and 3-O-C12-HSL (N-3-oxo-dodecanoyl-DL-homoserine lactone).

The genomic DNA was isolated by using the X-DNA Purification Kit (Xtrem Biotech S.L., Granada, Spain), and the 16S rRNA gene was amplified by the universal bacterial primers 16F27 and 16R1488. The PCR product was purified and cloned into the pGEM®-T vector (Promega). Direct sequencing of the PCR-amplified DNA was determined using an ABI PRISM DyeTerminator Cycle Sequencing Ready Reaction Kit (Perkin-Elmer) and an ABI PRISM 377 Sequencer (Perkin-Elmer) according to the manufacturer's instructions. The DNA sequence obtained was compared to the reference 16S rRNA gene sequences available in the GenBank and EMBL databases obtained from the NCBI Genome Database using BLASTN software (Altschul et al., 1990) and the EzBioCloud server (Yoon et al., 2017). The phylogenetic analysis was carried out using Molecular Evolutionary Genetics Analysis (MEGA) software version X (Kumar et al., 2018) following multiple data alignments by the CLUSTAL OMEGA (Sievers et al., 2011). Distances and clustering were determined according to the neighbor-joining and maximum-likelihood methods by applying the cluster stability algorithm based on the bootstrap analysis (1,000 replications).

To describe new taxa of aerobic and endospore-forming bacteria, the recommended traits by Logan et al. (2009) were applied in the study of the type species Peribacillus. The shape, size, and pigmentation of the colonies were observed on a TSA medium after 48 h of incubation at 28°C. The motility was observed using a log phase culture according to the Hanging drop method. The oxidase (Kovacs, 1956) and catalase activities were determined as well.

The optimum growth and growth range were determined in a TSB medium at different NaCl concentrations ranging from 0 to 25% (w/v) in 1.0 intervals adjusting the pH 7. The pH growth range and optimum pH were also determined in the TSB medium, testing from 4 to 11 in 1.0 pH unit intervals, using the following buffer systems: 0.1 M citric acid/0.1 M sodium citrate (pH 4.0–5.0,); 0.1 M KH₂PO₄/0.1 M NaOH (pH 6.0–8.0); 0.1 M NaHCO₃/0.1 M Na₂CO₃ (pH 9.0–10.0); and 0.2 M KH₂PO₄/0.1 M NaOH (pH 11.0) (Xie et al., 2012). In both tests, bacterial growth was monitored by optical density at 600 nm. The temperature range for growth and the optimum temperature were determined on the TSA plates at 4, 10, 15, 20, 28, 37, 40, 42, and 45°C. The anaerobic growth capacity was evaluated on the TSA plates by incubation in hermetic jars using the Gas Pak Anaerobic System (BBL) to generate an anaerobic atmosphere over a one-week period. The hydrolyses of casein and starch were also performed (Uttley and Collins, 1993). Other biochemical characteristics were analyzed using the API 50CH and API 20E according to the manufacturer's instructions.

The cellular fatty acids were analyzed at the Spanish Type Culture Collection (CECT) in Valencia, Spain, following the instructions of the Microbial Identification System Operating Manual (MIDI, 2008). For this, the cell mass of the N3^T strain was obtained after growing for 24 h in a TSB medium at 28°C.

The genomic N3^T strain DNA was extracted following the protocol described by Marmur (1961) for later sequencing by the Illumina Hi-Seq platform at the STAB VIDA facility (Caparica, Portugal) with 2 × 150-bp paired-end reads. The reads that were processed by BBDuk (https://sourceforge.net/projects/bbmap/) to remove the adapters and low-quality bases were then assembled using SPAdes software v. 3.11.1 (Nurk et al., 2013). Finally, the obtained contigs were blasted against the nr/nt database to remove the contigs belonging to the contaminants.

The average nucleotide identity based on BLAST (ANIb) and MUMmer (ANIm) algorithms were determined with the aid of JSpeciesWS software (Richter et al., 2015). OrthoANI was similarly calculated using OrthoANI software (Lee et al., 2016). The average amino acid identity (AAI) values were calculated from protein sequences using an online AAI calculator at the Kostas Laboratory website (http://enve-omics.ce.gatech.edu/aai/). A two-way AAI was used.

In the case of digital DNA–DNA hybridization (dDDH), it was calculated using the BLAST+ algorithm on the DSMZ Genome-to-Genome Distance Calculator (GGDC 3.0) platform (Meier-Kolthoff et al., 2013, 2021). The results presented in this study are based on the recommended Formula 2 (identities/HSP length), which, being independent of genome length, is robustly protected against the use of incomplete draft genomes.

A core genome analysis of the strain N3^T and all species of the genus Peribacillus, including a representative strain of the related genera, for which their genome was available, was also performed using Bacterial Pan Genome Analysis (BPGA) software (Chaudhari et al., 2016) with the default parameters. After obtaining the core of the 23 bacterial genomes, all protein orthologs belonging to the core genome were concatenated and aligned by MAFFT (Katoh and Standley, 2013). A phylogenomic tree of the core genes of the species was then constructed using MEGA X software according to the maximum-likelihood method.

In vitro plant-growth-promoting (PGP) traits were analyzed for the N3^T strain through the screening of the production of acid and alkaline phosphatases (Pikovskaya, 1948; Baird-Parker, 1963), 1-aminocyclopropane-1-carboxylic acid (ACC) deaminase (Poonguzhali et al., 2006), indoleacetic acid (IAA) (Gang et al., 2019), and siderophores (Alexander and Zuberer, 1991) and the ability to fix nitrogen (Matthews and Suhaimi, 2010).

Moreover, rhizosphere competence traits were studied in this strain by the production of enzymes related to the hydrolyses of casein, cellulose, DNA, gelatin, starch, Tween 20, and Tween 80 (Jeffries et al., 1957; Uttley and Collins, 1993; Villalba et al., 2004). In the case of casein, cellulose, DNA, starch, and acid phosphatase, a clear halo surrounding the bacterial growth indicated a positive result for these tests, while the precipitation of calcium salts is detected in the case of Tween 20 and 80 hydrolyses. Alkaline phosphatase production is detected by a pink coloration after adding 10 mL of 30% (v/v) of ammonia to the plate. In the case of nitrogen fixation and ACC deaminase production, the growth of bacterial strain indicated a positive result by fixing gaseous nitrogen or degrading ACC, as the media do not contain any source of N except for ACC. The Siderophore test was carried out by the chrome azurol sulfonate (CAS) protocol, in which a change of color from blue to green was recorded as a positive result. Meanwhile, IAA production was detected spectrophotometrically in a TSB medium supplemented with tryptophan (500 mg/L) after the addition of the Salkowski reagent.

The N3^T strain QQ activity was assessed by a well-diffusion agar-plate assay using synthetic AHLs (Romero et al., 2011; Torres et al., 2016). Briefly, 10 μM of each AHL was added to an overnight culture of the N3^T strain and then incubated at 28°C for 24 h. A sterile TSB medium supplemented with AHLs was incubated as a negative control. The remaining AHLs were detected in the supernatant of each sample that was deposited in wells on the LB agar plates overlaid with C. violaceum CV026 or C. violaceum VIR07 or on AB agar plates supplemented with 80 μg mL⁻¹ of 5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside (Xgal) overlaid with Agrobacterium tumefaciens NTL4 (pZLR4). The plates were incubated at 28°C for 24 h to check for the development of a purple or blue color around each well. This assay was repeated three times.

For plant growth promotion assay in tomato plants, 50 seeds were surface sterilized (Molan et al., 2010) and sown in each 20 × 20 × 20 cm pot containing sterile vermiculite. When seedlings were 5 cm, each pot was irrigated with 5 mL of 10⁹ CFU mL⁻¹ of the strain N3^T-washed cells every seven days. Irrigation with sterile distilled water was used as a negative control. Pots were kept in a greenhouse under a long-day photoperiod (16:8h, light:dark) at 21°C for 4 weeks. Then, root and shoot lengths and dry weight were determined.

To test the ability of the strain N3^T to interfere in the D. solani LMG 25993^T and E. amylovora CECT 222^T virulence, experiments using cocultures between them were carried out in potato slices and pears, respectively. Briefly, pathogen (10⁷ CFU mL⁻¹)–strain N3^T (10⁹ CFU mL⁻¹) cocultures were conducted in a 1:100 ratio in a TSB medium and incubated for 24 h at 28°C. A similar concentration of each pathogen or strain N3^T was added to the cell-free TSB as controls, respectively (Torres et al., 2016). For the potato tubers (Solanum tuberosum) and pears (Pyrus communis) assays, they were tap-washed and surface-sterilized by spraying with 1% (w/v) sodium hypochlorite solution, followed by 70% (v/v) ethanol and sterile distilled water. Sliced potatoes and pears were inoculated with 5 μL of the N3^T-phytopathogen cocultures and with controls (each bacterium monocultures and the sterile distilled water) at three or four equidistant spots, respectively. Nine replicates of each treatment were performed, and the experiment was repeated three times. After 48 h of incubation at 28°C, the maceration zones were visually detected and measured using ImageJ software (Schneider et al., 2012). Plate counts in monocultures and cocultures were performed to determine the concentration of each phytopathogen and strain N3^T in the TSA medium supplemented with 5% (w/v) NaCl as a selective agent for excluding the salt-sensitive phytopathogens.

In parallel, the remaining AHLs from each culture were determined according to the well-diffusion agar-plate method to assess the QQ activity of the strain N3^T. For this, the co-cultures were centrifuged at 12,000 rpm for 10 min, and the cell-free supernatants containing the AHLs were subject to a double extraction with an equal volume of dichloromethane. The organic phase was dried and finally suspended in 20 μL of 70% (v/v) methanol. Five microliters of each extract were spotted in the sterile filter paper disks placed in AB-Xgal plates (Chilton et al., 1974) using A. tumefaciens NTL4 (pZLR4) as a biosensor. The monocultures of the strain N3^T and each pathogen were likewise extracted and tested as negative and positive controls, respectively.

The Shapiro–Wilk test was used to verify data normality, and the data were statistically analyzed with the aid of the ANOVA (P ≤ 0.05) and Tukey tests using the SPSS software.

The cloned 16S rRNA gene of the strain N3^T resulted in a virtually complete 1558 bp-long sequence. The strain N3 ^T showed the highest sequence identity to Bacillus frigoritolerans DSM 8801^T (99.93%), Peribacillus simplex DSM 1321^T (99.80%), P. muralis DSM 16288^T (99.66%), P. butanolivorans DSM 18926^T (99.59%), and P. loiseleuriae (97.96%), while identities below 97% were obtained with other species from Peribacillus genus. Further phylogenetic analysis of its 16S rRNA gene sequences and other related strains through a phylogenetic tree reconstruction using the maximum-likelihood algorithm showed that this strain is a member of the genus Peribacillus and forms a cluster with the B. frigoritolerans species, which showed the highest sequence similarity (Figure 1). A similar phylogenetic distribution was achieved when neighbor-joining and maximum parsimony algorithms were applied (Supplementary Figures 1, 2, respectively).

The strain N3^T forms creamy and convex colonies of 4–5 mm diameter after 48 h of incubation in the TSA plates. The cells were rod-shaped, sometimes chained and motile when single cells, Gram-positive, catalase-positive, and oxidase-negative. The strain N3^T produced oval endospores in swollen sporangia. The strain N3^T was strictly aerobe, with growth temperatures ranging from 10 to 42°C and with an optimum temperature at 28°C. It grew in a pH range from 6 to 8, with pH 7 as optimum. This strain proved to be halotolerant, as it grew from 0.5 to 7.5% (w/v) of NaCl, with 1% (w/v) being the optimum concentration.

Its phenotypic characteristics are shown in Table 1 and in the species description section. The differential characteristics of the strain N3^T with respect to the most closely related species, B. frigoritolerans DSM 8801^T, P. muralis DSM 16288^T, P. butanolivorans DSM 18926^T, P. loiseleuriae DSM 101776^T, and the type strain of the genus P. simplex DSM 1321^T are also shown in Table 1. The strain N3^T mainly differs in relation to the following features: the inability of growing anaerobically and at 4 or 45°C, hydrolyze casein, reducing nitrate to nitrite, and resulting negative for oxidase, ONPG reaction, citrate utilization, and urease production, and none of the features were able to produce acids from any sugar tested in API 50CH except for aesculin, in which a weak result was detected. By contrast, this strain can grow in a wider range of NaCl and hydrolyze starch and gelatin and is also positive for arginine dihydrolase, tryptophan deaminase, and indole production.

The analysis of the strain N3^T fatty acids indicated a predominance of anteiso-C_15:0 (68.07%) and iso-C_15:0 (8.71%) (Table 2). This profile was similar to that of the most closely related strains and the type strain of the genus; however, a marked increase in the relative abundance of anteiso-C_15:0 and a decrease of iso-C_15:0 made a difference with respect to its relative profiles.

The draft genome of strain N3 ^T was manually curated, and it resulted in over 5.7 Mbp with 88 contigs. The quality of the assembly was assessed using Quality Assessment Tool for Genome Assemblies (QUAST) software. The whole-genome sequence was of sufficient quality, with an N50 value of 263,554, an L50 value of 9, and approximately 200X coverage. The PGAP (Tatusova et al., 2016) annotation of the draft genome showed a total of 5,319 protein-coding genes (PCGs), 4,825 of which were assigned at least to a single functional COG category in the EggNOG 5.0 database (Huerta-Cepas et al., 2018; Cantalapiedra et al., 2021); categories K (transcription) and E (amino acid metabolism and transport) were the most abundant, with 455 and 585 proteins assigned to those categories, respectively. This genome sequence, which was deposited in the GenBank/EMBL/DDBJ database under accession number JAJNAF000000000, was used for further analysis.

The in silico analysis of G+C content in the draft genome of the strain N3^T produced a value of 40.28 mol%, whereas the range of G+C content for the type species of the genus P. simplex DSM 1321^T is 39.5–41.6 mol% (Heyrman et al., 2005).

The ANIs, based both on BLAST (ANIb) and MUMmer (ANIm), and the average AAI for the strain N3^T and the related species are shown in Supplementary Table 1. The ANIb and ANIm values between the strain N3^T and the most closely related P. simplex DSM 1321^T were 93.21% and 93.94%, respectively. The proposed cutoff for species delimitation is 95–96%, as proposed by Richter and Rosselló-Móra (Richter and Rossello-Mora, 2009; Kim et al., 2014). In all the cases, the ANI values for the strain N3^T and all the related species were below that cutoff. The AAI values between strain N3 and phylogenetically related species ranged from 67.54 to 95.67% (Supplementary Table 1). Konstantinidis and Tiedje (2005) proposed an AAI threshold (about 95–96%) for species demarcation of prokaryotes based on 175 genomes. The low AAI values confirm that the strain N3^T represents a novel species within the genus Peribacillus.

The OrthoANI calculation between the N3^T strain and P. simplex DSM 1321^T was 93.66%, which is in line with the results obtained for ANIb and ANIm and below the proposed cutoff (Supplementary Table 2).

The dDDH of the whole-genome sequences of the N3^T strain and the closely related species was carried out, and in all the cases, the results fell below the proposed cutoff delimitation for the species (70%) (Goris et al., 2007) (Supplementary Table 2).

The data obtained with Bacillus frigoritolerans (Liu et al., 2020) also indicated that this taxon must be reclassified as Peribacillus frigoritolerans.

The concatenated alignment of the 803 core orthologous proteins of the strain N3^T and the species of the genus Peribacillus, including a representative strain of the related genera, was used to reconstruct a maximum-likelihood phylogenetic tree, confirming our previous results, as shown in Figure 2.

The strain N3^T resulted positive for DNA, gelatin, starch, and Tween 20 and Tween 80 hydrolyses. It also produced siderophores and IAA and was able to fix nitrogen and degrade ACC, while the acid phosphatase activity was variable (Supplementary Table 3). Some of the hydrolyses are associated with plant nutrient acquisition, pathogen competition, and phytohormone balance interference in the plant with the increase of plant growth and development hormones and the decrease of stress hormones (Gupta and Pandey, 2019; Miljaković et al., 2020).

Concerning the QQ activity of this strain, it was able to degrade the majority of the synthetic AHLs tested with different intensities depending on the AHL. The highest degradation was achieved against 3-O-C6-HSL and 3-O-C12-HSL, in which a total degradation was observed. By contrast, a lower degradation was detected for C6-HSL, C10-HSL, and C12-HSL (Figure 3).

Experiments to test the PGP activity were performed with tomato because it is one of the most important vegetable plants in the world. Knowledge obtained from the studies conducted on tomatoes can be easily applied to these plants, which makes tomatoes an important research material (Kimura and Sinha, 2008). In vivo experiments to determine the growth promotion of the strain N3^T in tomato plants showed an increase in the length and dry weight of the plants compared with the control. Significant increases with respect to the control plants were observed in terms of aerial and total lengths, 16.1 and 14.3%, respectively, and of aerial, radicular, and total dry weight, 110.7, 55.3, and 106.8%, respectively (Figure 4).

The enzymatic degradation of AHLs in phytopathogens seems to be a promising alternative strategy to fight bacterial infections (Helman and Chernin, 2015). Previous studies reported the top 10 plant pathogenic bacteria in molecular plant pathology (Mansfield et al., 2012). The phytopathogens tested in this study, D. solani LMG 25993^T and E. amylovora CECT 222^T (members of the top 10), produce damages in potatoes and pears and cause huge losses in the agriculture production.

To evaluate the ability of the strain N3^T to interfere in the virulence of D. solani LMG 25993^T and E. amylovora CECT 222^T, first, the QQ activity against AHL extracts from each pathogen strain was analyzed. A well-diffusion plate test showed a total degradation of the AHLs produced by E. amylovora, but a partial degradation was found on the D. solani AHLs extract when the N3^T strain was cocultured with them (Figure 5B).

Second, the cocultures of the strain N3^T and D. solani LMG 25993^T and E. amylovora CECT 222^T were carried out and tested for their virulence in potato and pear assay, respectively. Potato slices treated with D. solani LMG 25993^T in monoculture showed a tissue maceration of 26.3% (Supplementary Figure 3A), while those treated with the coculture of this pathogen and the strain N3^T showed a complete inhibition of this maceration (Figure 5A). In the case of the experiments carried out with E. amylovora CECT 222^T in pears, monoculture with E. amylovora CECT 222^T produced a 93.8% of tissue maceration in pears, and the percentage was reduced to 18.1% when pears were treated with the coculture of the strain N3^T (Supplementary Figure 3B). On the other hand, the N3^T strain did not cause any damage to the potato and pear tissues (Figure 5A). To discard the growth inhibition of the phytopathogens by the strain N3^T, a plate-counting method was carried out using different concentrations of NaCl. The results indicated that no differences in the growth of D. solani and E. amylovora were observed in the monocultures with respect to the cocultures (10⁷ CFU mL⁻¹).

These results indicated that the strain N3^T attenuated the virulence of the pathogens tested through a QQ approach. Our results are in line with other studies that also demonstrated the attenuation of phytopathogens virulence by using the QQ strains, such as Lysinibacillus sp. Gs50 (Garge and Nerurkar, 2016), Stenotrophomonas maltophilia M9-54 (Reina et al., 2019), Pseudomonas segetis P6 (Rodríguez et al., 2020), Ochrobactrum intermedium D-2 (Fan et al., 2020), and Acinetobacter sp. XN-10 (Zhang et al., 2020). In fact, heterologous expressions of QQ bacterial enzymes, such as the lactonases AiiA from Bacillus sp. or the metagenome derived HqiA (Torres et al., 2017) in Pectobacterium carotovorum subsp. Carotovorum, have demonstrated an effect on the production of pathogen virulence factors.

In many cases, the results shown in the laboratory conditions are not the same as the ones obtained under in vivo assays. Thus, in order to use the strain N3^T as an effective biocontrol agent in biotechnology, new experiments need to be done under greenhouse conditions. The preliminary results under in vivo assays suggest that the biocontrol mechanism of Pseudomonas syringae pv. tomato might be QQ Peribacillus sp. N3 (personal communication).

Other antivirulence mechanisms apart from QQ cannot be discarded in the strain N3^T, since the Bacillus species are well-known to produce numerous inhibitor compounds and some of them do not affect the viability of the pathogen (Leathers et al., 2020). An analysis in silico using the antiSMASH tool (Blin et al., 2021) indicated that the strain N3^T produced lipopeptides of the family NRPS and lassopeptides that also implicated its antivirulence mechanism.

Although the Bacillus strains with growth promotion activity (Torres et al., 2019) or AHL-degrading enzymes from the Bacillus strains that can reduce the QS-regulated virulence factors in pathogens have been described (Dong et al., 2000; Zhao et al., 2008; Zhou et al., 2008), only a few studies have reported a bacterium that combines both properties (Vega et al., 2019; Rodríguez et al., 2020). Indeed, to our knowledge, this is the first study to describe a Peribacillus strain with both the PGP and QQ activities.