Unless stated otherwise, all chemicals used in this study were purchased from Sigma-Aldrich. Escherichis coli DH5α was used as a general cloning host, and E. coli S17.1 λpir for conjugal transfer of plasmids. B. methanolicus MGA3 was used as the expression host. All strains used in this study are listed in Table 1. E. coli strains were cultivated at 37 °C in either Lysogeny Broth (LB) or on LB agar plates, with supplementation of 50 μg/mL kanamycin or 25 μg/mL chloramphenicol when relevant. B. methanolicus was cultivated at 50 °C and 200 rpm in Super Optimal Broth (SOB; Difco) or SOB agar plates during strain development and otherwise cultivated in minimal media MVcM or MVcMY (0.25 g/L yeast extract) supplemented with 200 mM methanol. The MVcM medium contained, in 1 L of distilled water: 4.09 g K₂HPO₄, 1.49 g NaH₂PO₄*H₂O, 2.11 g (NH₄)₂SO₄; the pH was adjusted to 7.2 with NaOH before autoclaving. The MVcM medium was supplemented with 1 mL 1 M MgSO₄*7H₂O solution. Moreover, 1 mL trace element solution was added, containing in 1 L of distilled water: 5.56 g FeSO₄*7H₂0, 27.28 mg CuSO₄*2H₂O, 7.35 g CaCl₂*2H₂O, 40.50 mg CoCl₂*6H₂O, 9.90 g MnCl₂*4H₂O, 287.54 mg ZnSO₄*7H₂O, 48.40 mg Na₂MoO4*2H₂O, 30.92 mg H₃BO₃, and 80 mL HCl. One mL of a vitamin solution was added; it contained in 1 L of distilled water: biotin, thiamine hydrochloride, riboflavin, D-calcium pantothenate, pyridoxine hydrochloride, nicotinamide, 0.1 g each; p-aminobenzoic acid, 0.02 g; folic acid, vitamin B₁₂ and lipoic acid, 0.01 g each (Schendel et al., 1990). Media for B. methanolicus cultivation was supplemented with antibiotics (kanamycin 25 μg/mL, chloramphenicol 5 μg/mL) when relevant. The strains were cultivated either in 250 mL Erlenmeyer flasks or in 24-well Duetz system plates, each well holding 10 mL (Haupka et al., 2021). The culture volumes corresponded to 10% of the total flask volume or 30% of the well volume. The optical density at 600 nm (OD₆₀₀) of cell cultures was monitored with a WPA CO 8000 Biowave photometer. For testing H₂O₂ tolerance, B. methanolicus was cultivated in MVcM supplemented with increasing concentrations of H₂O₂ (0, 0.1, 0.5, 1, 2, or 3 mg/L). The cultivation was carried out in the BioLector Pro microbioreactor system (m2p-labs) at 50 °C, with a 48-well flower plate containing 1 mL culture and with a stirring speed of 1,200 rpm, humidity set to 85%, and a biomass gain set to 7. Cell biomass formation, based on scattered light, was monitored over 24 h.

Plasmids used in this study are listed in Table 1 and primers in Supplementary Table 1. Recombinant DNA work in E. coli was performed as described in Sambrook et al. (1989). Plasmid DNA was isolated using Promega's Wizard^® Plus SV Miniprep kit. Plasmid backbones and homologous flank regions were amplified with Takara Bio's CloneAmp^TM HiFi PCR premix. Qiaquick PCR Purification and Gel Extraction kits (Qiagen) were used for PCR purification and gel extraction. Fragments were joined by isothermal DNA assembly according to Gibson et al. (2009). Colony PCR was performed using either Phusion^® High-Fidelity DNA polymerase (New England Biolabs) or GoTaq^® polymerase (Promega). The plasmids pCasPP-spo0A, pCasPP-katA and pCasPP-ald were constructed as described by Rütering et al. (2017) by fusing 1 kb homologous arms flanking the targeted gene. The 1 kb length of homologous arms was chosen based on Rütering et al. (2017). Twenty four bp long sgRNA were inserted into the sgRNA region of pCasPP plasmid, specifically at the BbsI restriction site, by PCR overlap extension, and the plasmids were re-circularized using Gibson Assembly (Gibson et al., 2009). For the pCasPP-katA::mcherry and pCasPP-spo0A::mrfp1 constructs, the mcherry and mrfp1 genes were cloned between the homologous flanks via overlap extension PCR, and the same sgRNAs as for katA and spo0A deletions were used for gene replacement. CRISPRi plasmids were constructed by inserting 24 bp sgRNA in the XbaI/AvaI restriction sites of the piCas plasmid, following the method described by Schultenkämper et al. (2019). The genome sequence of B. methanolicus (GenBank accession numbers CP007739, CP007741, and CP007740) was used to identify 24-nucleotide target sequences on the non-translated strand using the CRISPR Guide RNA Design Tool. We selected sgRNAs with an in silico off-target score of 50, indicating high predicted specificity (Benchling, https://www.benchling.com/crispr).

B. methanolicus strains were transformed by utilizing conjugal transfer of plasmids from E. coli S17.1 λpir, described by Irla et al. (2025). Recipient and donor cultures of B. methanolicus and E. coli S17.1 λpir were cultivated overnight in 25 mL of SOB and LB media, respectively. Fourty mL 1% cultures of the donor and recipient cells were cultivated the next day for 4 hours before mixing them in two dilutions: 9 mL of B. methanolicus with 3 mL of E. coli S17.1 λpir, and 900 μL of B. methanolicus with 300 μL E. coli S17.1 λpir. The mixes were centrifuged at 8,000 rpm for 5 min and resuspended gently in 100–200 μL of leftover supernatant. The cell mixtures were then transferred without spreading to a non-selective SOB agar plate and incubated overnight at 40 °C. The mixed cell colonies were then collected and resuspended in 200 μL of prewarmed SOB media and plated on selective SOB agar at 50 °C. Transconjugant colonies (n = 20) were replated, and leftover cell material was boiled in water for 10 min at 98 °C and vortexed. The cells were then centrifuged at 13,000 rpm for 5 min in a Mini Spin centrifuge (Eppendorf). The supernatant was then diluted ten times before being used as a template in PCR for screening deletion clones. If still present, the pCaspp-based plasmids were curated through 4 successive cultivations in antibiotic-free SOB liquid medium. To verify plasmid curing, deletion clones were plated on selective SOB plates, and the absence of growth confirmed successful plasmid loss. Furthermore, a Monarch Spin gDNA Extraction kit (New England Biolabs) was used to extract genomic DNA of positive colonies, according to the manufacturer's recommendations. The correctness of all cloning and gene deletions was confirmed via Sanger sequencing by a third-party (Eurofins Genomics).

For whole genome sequencing, one representative of each deletion clone was cultured overnight in SOB medium and genomic DNA was isolated using the Monarch Spin gDNA Extraction kit. Whole genomes were sequenced by a third-party (Eurofins Genomics). Illumina sequencing was conducted by constructing two paired-end libraries with average insertion lengths of 500 bp and 2,000 bp. Sequences were generated with an Illumina GA IIx (Illumina Inc., San Diego, CA, USA). Raw data was processed in four steps, including removing reads with 5 bp of ambiguous bases, removing reads with 20 bp of low quality (≤Q20) bases, removing adapter contamination, and removing duplicated reads. Finally, 100× libraries were obtained with clean paired-end read data.

Sequencing reads were mapped to the reference genome sequence, composed of the chromosome and the two native plasmids pBM19 and pBM69 (GenBank accession numbers CP007739, CP007741, and CP007740, respectively). Before the alignment, the reads underwent trimming with a minimum length of 36 bp using the Trimmomatic v0.33 tool (Bolger et al., 2014). The trimmed reads were then aligned to the reference sequences using the short-read alignment software Bowtie 2 (Langmead et al., 2009). Mapped reads were visualized employing the ReadXplorer v2.2.2 tool (Hilker et al., 2016). Variant calling was conducted using the Bcftools mpileup pipeline, with quality score cut-off set to 50 (Li et al., 2009).

B. methanolicus strains cultured in MVcMY media were harvested at OD₆₀₀ 1–2 by centrifugation at 8,000 rpm and 4 °C for 10 min in an Eppendorf 5430 R centrifuge. Cell pellets were washed twice and resuspended in 50 mL Tris-HCl pH 7.5 before storing at -80 °C. The cells were then thawed on ice and transferred to 2 mL screw-cap tubes containing 200 μL 0.1 mm glass beads (Carl Roth) and disrupted in three rounds using a Retsch MM400 bead beater at 30 ms for 1 min and resting on ice for 5 min. The crude extract was isolated after centrifugation at 17,000 rpm for 1 h at 4 °C in an Eppendorf 5424 R centrifuge. Protein concentrations were determined with Bradford assay protocol (Bio-Rad). To determine catalase activity, the crude extracts underwent catalase assay using the Catalase Colorimetric Activity Kit (Thermo Fisher), according to the manufacturer's protocol. Furthermore, H₂O₂ drop assay was performed by adding 3 μL of 3 % (v/v) H₂O₂ to 150 μL of a cell culture (OD₆₀₀ ≈ 2) in a Thermo Scientific™ Nunc MicroWell 96-Well optical-bottom microtiter plates. The production of oxygen could be detected by the formation of bubbles over the mixture (Schultenkämper et al., 2019).

Ald assay reactions measuring the reductive amination of pyruvate contained final concentrations of 50 mM Tris-HCl, pH 7.5, 5 mM pyruvate, 0.5 mM NADH, 200 mM NH₄Cl, and 50 μL of crude extract in a total volume of 200 μL. Ald assay reactions measuring the oxidative deamination of alanine contained 50 mM Tris-HCl pH 7.5, 5 mM alanine, 0.5 mM NAD⁺, and 50 μL crude extract. The reactions were monitored at 340 nm for 6 min under a constant temperature of 40–42 °C. Calculations were performed based on a molar extinction coefficient of 3.990 L/mmol/cm for NADH. Both Ald assays were conducted using a TECAN-Infinite M200-Microplate reader and Thermo Scientific™ Nunc MicroWell 96-Well optical-bottom plates.

Supernatants from B. methanolicus strains cultured in MVcM media were collected (1 mL) by centrifugation at 8,000 rpm and room temperature for 10 min in an Eppendorf 5430 R centrifuge and stored at -20 °C until use. The supernatants were analyzed using high pressure liquid chromatography (HPLC) samples were derivatized using a Waters Alliance e2695 Separations Module. The samples were derivatized using FMOC-Cl (fluorenylmethyloxycarbonyl chloride), according to Melucci et al. (1999). Amino acids were separated in a Symmetry C18 column (100 Å, 3.5 μm, 4.6 mm × 75 mm, Waters) carried by mobile phase A, 50 mM Na-acetate, pH 4.2, and B, acetonitrile. The mobile phase flow rate was set to 1.3 mL/min and the gradient conditions were as follows: 5 min- 62% A and 38% B, 12 min- 43% A and 57% B, 14 min-24% A and 76% B, 15 min-43% A and 57% B, and 18 min- 62% A and 38% B. The detection was performed with a Waters 2475 HPLC Multi Fluorescence Detector (Waters), with excitation at 265 nm and emission at 315 nm (Brito et al., 2019).

The CRISPR-Cas9 mediated genome modification tool developed in this study is based on the pUB110-based plasmid pCasPP, carrying a Streptococcus pyogenes-derived cas9 gene under the transcriptional control of the broad-host-range surface layer protein gene (sgsE) promoter from Geobacillus stearothermophilus, and an sgRNA region under a constitutive gapdh promoter derived from Streptomyces griseus. It carries a neomycin/kanamycin resistance gene, a repU gene responsible for replication in Bacillales, BbsI flanked lacZ selection cassette, and origin of transfer (oriT) required for conjugation (Shao et al., 2013; Cobb et al., 2015). The plasmid contains a SpeI site which is used for insertion of homologous arms necessary for HDR (Rütering et al., 2017). The Cas9-carrying plasmids can be transferred to B. methanolicus cells via electroporation or conjugation with E. coli S17.1 λpir through a co-cultivation step carried out at 40 °C. The Cas9 activity threshold lies below the optimal growth temperature of B. methanolicus at 50 °C (Wiktor et al., 2016; Mougiakos et al., 2017a; Schmidt et al., 2019). Therefore, lowering the temperature to 40 °C during conjugation with E. coli not only allowed co-cultivation on the two strains but also activity of Cas9 protein. The subsequent raising of the temperature to 50 °C led to elimination of the E. coli cells and deactivation of Cas9 protein. The use of a more efficient plasmid delivery method reduces the screening burden. Therefore, this conjugation method yields approximately 600 colonies on plate, compared to ~20 colonies achieved using the classic electroporation technique (Irla et al., 2025). To validate the functionality of the CRISPR-Cas9 based pCasPP system for this thermophilic bacterium, we targeted two distinct genes located at different loci in the B. methanolicus genome: the katA gene (BMMGA3_04865; WP_004433703.1), which encodes a catalase, and the ald gene (BMMGA3_13155; WP_003347426.1), which encodes an alanine dehydrogenase.

In a previous study, the role of katA was analyzed using the CRISPRi approach. Targeting katA with CRISPRi led to a decrease in catalase activity by about 75% (Schultenkämper et al., 2019). Here, we leveraged the CRISPRi system to test the phenotypical response of targeting the ald gene (Supplementary Figure 1A). Repression of ald resulted in a decrease by 12% in alanine production under small-scale Duetz plate conditions (Supplementary Figure 1B). CRISPRi results provided preliminary confirmation of the target activities, supporting their potential for further deletion experiments. For gene deletion, we designed sgRNA sequences targeting the same katA and ald genes along with homologous regions of around 1,000 bp for each flank to obtain scarless deletions in B. methanolicus (Figure 1A) (Rütering et al., 2017). The plasmids named pCasPP-katA and pCasPP-ald were constructed (Table 1) and used to transform B. methanolicus by conjugation.

For the confirmation of gene deletions in B. methanolicus, PCR amplification of genomic DNA was performed using primers flanking the homology regions incorporated in the deletion plasmid. The resulting amplicons as visualized by gel electrophoresis confirmed a reduction in fragment size for both the ΔkatA and Δald deletions compared to the wild-type (WT) control, as shown in Figures 1B, C. The deletion efficiency was approximately 87%, with 2–3 negative clones routinely observed among 20 randomly picked colonies, which represents a total of 5.22 × 10⁻⁶ recombinants per cell (data not shown). To further confirm the gene deletions, the genomic DNA from the deletion strains ΔkatA and Δald, and the WT control strain was sequenced and the reads mapped to the reference genome. While reads derived from the WT strain mapped continuously across the katA and ald regions, the deletion strains exhibited clear unmapped gaps in these regions, indicating the successful CRISPR-mediated deletion of the target genes (Figures 1D, E).

In this study, we deleted the catalase-encoding gene katA and assessed its role in H₂O₂ detoxification in B. methanolicus. The specific catalase activity was measured in crude extracts from the WT strain and the catalase deletion strain ΔkatA. Additionally, catalase activity was verified in the complementation strain ΔkatA + katA and its empty vector counterpart, ΔkatA + E (Table 1). The catalase activity in the WT strain was ~1.2 U/mg. In contrast, the katA deletion resulted in a loss of any detectable catalase activity, as observed in both ΔkatA and ΔkatA + E strains. As expected, the complementation strain exhibited significantly elevated catalase activity (~5.8 U/mg), driven by plasmid-based gene expression (Figure 2).

We further evaluated the impact of katA deletion under oxidative stress by assessing cell fitness in response to increasing concentrations of H₂O₂ supplemented into the growth media. The strains were cultivated in a microbioreactor (Biolector Pro system plates). The growth rates of B. methanolicus strains (WT, ΔkatA, ΔkatA + E, and ΔkatA + katA) were measured under H₂O₂ concentrations ranging from 0 to 3 mg/L, which allowed us to investigate the role of catalase in H₂O₂ detoxification. Our results indicate that WT B. methanolicus MGA3 naturally tolerates up to 0.5 mg/L H₂O₂, as the WT strain exhibited sustained growth under these conditions. However, growth of the WT strain was impaired at higher concentrations of H₂O₂ (Figure 3A). The ΔkatA deletion strain showed an ~21% reduction in growth rate at 0.1 mg/L H₂O₂ compared to the no-supplement control. Moreover, approximately 42% growth reduction occurred for the ΔkatA strain in comparison to the WT even without ant H₂O₂ supplementation, indicating the importance of catalase for the cell fitness (Figure 3A). Growth in the ΔkatA strain was completely halted starting from the supplementation of 0.5 mg/L mM H₂O₂. In the ΔkatA + E strain, which carries the empty vector plasmid, a similar growth pattern was observed, with growth maintained only up to 0.1 mg/L H₂O₂. We suggest that a minor residual oxidative stress tolerance in the ΔkatA strain may result from alternative paralogs. A BLAST search identified two thiol-dependent peroxidases (BMMGA3_03510 and BMMGA3_05260) that may play a role in oxidative stress defense in B. methanolicus MGA3. Additionally, four genes annotated with the Gene Ontology term GO:0004601, indicating peroxidase activity, were found. However, the genome of B. methanolicus contains a single gene annotated as a catalase. In contrast, when the katA gene was complemented in the ΔkatA + katA strain, the growth fitness was restored, with growth rates comparable across all H₂O₂ concentrations tested (Figure 3B). Thus, it is confirmed that the deletion of katA was responsible for the low H₂O₂ tolerance phenotype.

In B. subtilis, alanine dehydrogenase catalyzes the reversible reductive amination of pyruvate to alanine (Supplementary Figure 2A). The B. methanolicus strains WT, Δald, Δald + E, and Δald + ald (Table 1) were tested for Ald activity. Enzyme activity was measured in crude cell extracts to evaluate whether deletion of the ald gene in B. methanolicus impairs the oxidative deamination of alanine and to examine potential directional preferences. Hence, the consumption or accumulation of the reaction cofactor NADH was monitored. The specific enzyme activities are presented in Table 2. The deletion strains Δald and Δald + E displayed comparable levels of Ald activity for the reductive amination of pyruvate (2.8 ± 0.5 mU/mg and 4.3 ± 1.8 mU/mg), which was five-fold lower compared to the control strain WT. Whereas the over-production of Ald in the complemented deletion strain (Δald + ald) lead to a more than 200-fold increase in activity at 4,709 ± 200 mU/mg. Interestingly, in the oxidative deamination of alanine, only Δald + ald displayed activity, which was substantially reduced in comparison to the activity displayed in the same strain for the reverse reaction (Table 2).

Furthermore, to assess the effects of ald deletion in B. methanolicus, the accumulation of alanine was measured in the strains WT, Δald, Δald + E, Δald + ald grown in minimal media in 10-well Duetz plates system. The results from CRISPRi-mediated repression of the ald gene led to significant changes in alanine titer in B. methanolicus strain iald, with a reduction in alanine production in comparison to the non-targeting control (Supplementary Figure 1B). However, the deletion of ald did not replicate the effects observed with gene repression. The WT strain secreted approximately 168 mg/L of alanine, while the Δald and Δald + E strains produced around 204 mg/L and 211 mg/L of alanine, respectively. The complementation strain produced about 142 mg/L of alanine. Although slight trends were observed between the strains, no significant differences (p < 0.01) were detected in the Scott-Knott test (Supplementary Figure 2B). These findings suggest that while ald deletion may slightly affect alanine production, its impact on biosynthesis is not substantial enough to reach statistical significance under the conditions tested. Further studies may be needed to fully elucidate the role of ald in alanine metabolism.

To analyse putative genetic variants derived from the CRISPR-Cas9 system, the genome sequences of the deletion strains ΔkatA and Δald, as well as the WT were used to conduct the variant calling analysis in comparison to the B. methanolicus MGA3 reference genome sequence (GenBank accession numbers CP007739, CP007741, and CP007740 for chromosome and plasmids pBM19 and pBM69, respectively) (Irla et al., 2014). Variants that were shared between the WT and deletion strains were not included in Table 3, as they are inherent to the WT genome (Supplementary Table 2). Consequently, Table 3 presents only the deletion-associated mutations. Here, we identified 13 variants in the ΔkatA strain and 14 variants in the Δald strain (Table 3). These variants represent deviations from the reference genome that do not appear in the WT sequence. These findings suggest that both deletions resulted in similar genetic plasticity, regardless the distinct locations and putative functions of the targeted genes. Notably, both deletion strains harbored variants in genes encoding enzymes involved in insertional mutagenesis and DNA repair processes. Alterations in DNA repair mechanisms have been shown to increase mutation rates in B. subtilis (Dervyn et al., 2023). Here, the ΔkatA strain exhibited indel mutations in genes encoding SAM-dependent methyltransferases and DNA gyrase gyrA. The ald deletion strain presented a mutation in the gene encoding methionine-tRNA ligase metG (Table 3).

Regarding the mutations potentially associated with the loss-of-function effects of gene deletions, the absence of catalase activity in the katA deletion strain resulted in mutations in enzymes and metabolic pathways that rely on water as a cofactor. The accumulation of H₂O₂ could interfere with these pathways, as it may modify enzyme active sites or other critical components, thereby disrupting their function. Hence, such mutated genes identified in the katA deletion strain included asparagine synthetase [glutamine-hydrolyzing] 3 (asnO), NADP-dependent glyceraldehyde-3-phosphate dehydrogenase (gapN), and 2,3-bisphosphoglycerate-independent phosphoglycerate mutase (Table 3). Surprisingly, the ald deletion strain displayed mutations in the dapG gene, which encodes aspartokinase 1, the enzyme catalyzing the phosphorylation of aspartate in lysine biosynthesis. The synthesis of lysine precursor, oxaloacetate, through carboxylation of pyruvate in an anaplerotic node is a competitive pathway to the alanine dehydrogenase reaction (Nærdal et al., 2017). However, the biological relationship between the ald deletion and the mutation in dapG remains unclear and warrants further investigation.

To expand the functionality of the CRISPR-Cas9 system in B. methanolicus, we employed it for gene deactivation by leveraging the bacterium's native DNA repair pathways, enabling error-prone end-joining in the absence of a repair template. The B. methanolicus chromosome encodes three main mechanisms for repairing DNA double-strand breaks (DSB). The first pathway is mediated by RecA (BMMGA3_06490), a multifunctional protein that mediates homologous recombination. Moreover, the genome search suggests that B. methanolicus may possess an additional template-dependent repair pathway involving the addA and addB gene cluster (BMMGA3_04250–04255), which encodes a RecBCD-like helicase/nuclease complex (Chayot et al., 2010). The third pathway, putatively active when no repair template is available, is the classical non-homologous end joining (NHEJ). It involves the genes ykoV (BMMGA3_09830) encoding DNA-end-binding protein Ku and ykoU (BMMGA3_09875) encoding DNA repair polymerase/ligase LigD (Hernández-Tamayo et al., 2022). We did not identify annotated genes encoding a complete microhomology-mediated end-joining pathway (Sfeir and Symington, 2015). To investigate the functionality of an error-prone end-joining mechanism in B. methanolicus, we chose the katA gene encoding a catalase (BMMGA3_04865) as a target and a pCasPP-based plasmid carrying only the katA targeting sgRNA without providing homologous flanks for HDR. Approximately 600 colonies were obtained, showing a much higher number than expected for typical escapers (Li et al., 2023). From those, 7 clones picked and were tested for catalase activity, and a function loss was observed for all of them in comparison to the WT control (Figure 4A). Sequencing of the katA gene PCR-amplicons from these clones revealed that the activity loss was possibly caused by mutations located in 9 to 18 bp downstream the targeted PAM region (Figure 4B). By that approach, the functionality of an error-prone DNA repair system in B. methanolicus was confirmed. All 7 colonies analyzed contained mutations at the Cas9 cleavage site, highlighting the potential of this system for rapid gene inactivation and future applications in multiplex genome editing.

The katA gene was also employed as a proof-of-concept target to demonstrate the applicability of the CRISPR-Cas system for gene insertions in B. methanolicus. In this experiment, the katA gene was replaced with the reporter gene mcherry. For this purpose, the construct pCasPP-ΔkatA::mcherry contained the same katA homology arms and the sgRNA sequence used for katA deletion in the pCasPP-ΔkatA plasmid, but with the homology arms flanking the Pmdh-mcherry cassette (Table 1; Figure 5A). Successful gene insertion was verified by PCR. PCR primers bind to the chromosomal regions flanking the katA locus, outside the homology arms (Supplementary Table 1). The ~0.3 kb difference between WT (~3.5 kb) and mcherry insertion clones (~3.8 kb) corresponds to the replacement of the katA coding sequence with the mcherry ORF plus the mdh promoter, consistent with the expected fragment size (Figure 5B). The insertion frequency was comparable to that observed for deletion experiments, with approximately 3 non-recombinant clones among 20 colonies screened. The resulting strain, ΔkatA::mcherry, exhibited approximately threefold higher mCherry fluorescence compared to the WT (Figure 5C). The background mCherry fluorescence observed in the WT had also been demonstrated previously (Irla et al., 2016). Hence, these results confirm the successful application of the CRISPR-Cas system for targeted gene insertion in B. methanolicus.

Next, we extended the CRISPR system to an additional genetic target. We targeted the gene encoding the major regulator Spo0A (BMMGA3_11365; WP_004435551.1), which is known to directly or indirectly influence more than 500 genes (Molle et al., 2003) and can be used in various microbial strains to decrease sporulation (Hou et al., 2016; Huang et al., 2004). Firstly, we have evaluated the efficiency of the CRISPR-Cas9 for the gene deletions (Supplementary Figure 3A). The plasmid pCasPP-spo0A was transferred to B. methanolicus via conjugation, and the empty pCasPP plasmid (no targeting sgRNA and no homologous flanks) was used as a control. Colony PCR analysis revealed 100% conjugation efficiency for the non-targeting pCasPP empty control (8/8 tested clones positive from 288 in total) as well as 100% editing efficiency for the targeting pCasPP-spo0A construct (10/10 tested clones positive from 28 in total). To evaluate the efficiency of the CRISPR-Cas9 system for targeted gene replacements, as previously described for P. polymyxa (Meliawati et al., 2022b), the pCasPP-spo0A construct was modified by inserting an mRFP1 encoding gene in between the homology flanks, thereby yielding the insertion plasmid pCasPP-spo0A::mrfp1. This plasmid carried the same sgRNA as utilized for the deletion of the spo0A gene but was designed to replace the spo0A gene with the mRFP1 encoding gene at its native locus. For that approach, a slightly lower editing efficiency of 95.2% was observed (20/21 positive clones from 2,784 in total). The positive integration of the mrfp1 gene was confirmed via colony PCR and subsequent sequencing of the resulting amplicons (Supplementary Figure 3B). By that, the editing efficiencies were proven to be close to 100% for single gene deletions as well as for gene replacements in B. methanolicus.