To generate a Cas12a-gRNA expression system that is cloneable in most E. coli strains, the E. coli/Methanosarcina shuttle vector pM000 (Supplementary Figure S1B) was constructed from the pWM321 plasmid (Metcalf et al., 1997) by replacing the origin of replication (ori) from plasmid R6K with the ori from plasmid ColE1. To establish markerless editability, the hpt gene that confers sensitivity to the purine analog 8-aza-2,6-diaminopurine (8ADP) for counter selection (Ehlers et al., 2011) was inserted downstream of the pac gene, yielding plasmid pM001 (Supplementary Figure S1C). The LbCas12a gene was amplified from pY016 (Zetsche et al., 2015), fused with the tetracycline-regulated promoter PmcrB (tetO1), and inserted into the multiple cloning site (MCS) of pM001 to yield the plasmid pMCp4. To obtain the final Cas12a-gRNA expression system, the gRNA cassette is introduced into pMCp4 (Figure 1A). It is worth mentioning that tetracycline was not required for the expression of Cas12a, as in the previous studies with Cas9, where no significant difference in the genome editing efficiency was observed in M. acetivorans (Nayak and Metcalf, 2017; Dhamad and Lessner, 2020).

To verify whether the Cas12a protein is not cytotoxic, the shuttle vector pM001 and the Cas12a-expressing plasmid pMCp4 were each transformed into M. acetivorans, and then the transformation efficiencies were compared. When 2 μg of plasmid DNA was used for the transformations, pM001 yielded 40,000 ± 20,000 CFUs of Pur^R transformants and pMCp4 yielded 34,000 ± 19,000 CFUs of Pur^R transformants (p = 0.75 in two-tailed t-test, Supplementary Figure S2), which indicates the expression of Cas12a protein is non-toxic to M. acetivorans cells.

To evaluate the targeting efficiency of the Cas12a-gRNA expression system, five different gRNAs were used to target various locations along the non-essential ssuC gene (which encodes the permease subunit of the sulfonate ABC transporter) for introducing DSBs on the genome. All five gRNAs resulted in less than 15 CFUs (per 2 μg DNA) in the transformation, thus demonstrating the high gene-targeting efficiency of the constructed system (Figure 1B).

To investigate the efficiency of gene deletion, the homologous repair (HR) arms for gene editing were inserted into the Cas12a-gRNA expression system to generate the Cas12a-gRNA gene editing system (Figure 2A). The g1RNA (5’-TTGCGATTCCCTCAGCCATG CCC-3′) was used to target the ssuC locus. For gRNA cassette construction, PCR amplification was used to synthesize the gRNA and direct repeat (DR) sequences that contain promoter (PmtaC1) and terminator (TmtaB1) regions (Supplementary Figure S3A). Deletion lengths varied from 100 bp to 2000 bp (Figure 2B) and a 1,000 bp length of flanking HR sequence was chosen as optimal based on previous work (Nayak and Metcalf, 2017).

When the plasmid pMCp2-g1RNA (Table 1) was transformed into M. acetivorans, the Cas12a-g1RNA complex was expressed and only 10 Pur^R transformants were observed (Figure 2C). Transformation of plasmid pMCp4 (Cas12a-only) yielded 33,000 ± 6,000 CFUs of Pur^R transformants (per 2 μg DNA). Among the various lengths of deletions tested, the 100 bp-deletion from the genome gave the highest repairing efficiency (1,800 ± 600 CFUs). This higher transformation efficiency obtained when only a small fragment of the chromosome is deleted also matches a previous Cas9-mediated gene knockout study, which concluded shorter deletions (i.e., Δ100 bp and Δ500 bp) are more stable and reliable (Nayak and Metcalf, 2017). For the plasmids that led to larger fragment deletions (i.e., Δ1000 bp and Δ2000 bp), similar transformant yields were obtained. To verify the knockout (or editing) efficiency, 10 transformants were randomly selected in each trial and subjected to colony PCR (Figure 2D). The primers involved are presented in Supplementary Figure S4A and Supplementary Table S1. Higher positive rates were observed with the shorter deletions, in which 100% of the 100 bp-deleted transformants were positive. On the other hand, 80% of positive clones were observed in the 2000 bp-deletion experiments. Three isolates from each transformation set were randomly selected for Sanger sequencing and all were edited correctly (Supplementary Figure S4B). The Sanger sequencing results were provided in Supplementary Table S2.

In order to assess the gene insertion performance of our Cas12a-gRNA gene editing system, a heterologous gene cassette for the expression of β-glucuronidase (uidA) gene was placed within the HR arms, given its previous successful use in Methanosarcina (Guss et al., 2008). For uidA cassette construction, the uidA gene was PCR amplified from E. coli BL21 genomic DNA and fused with the mcr promoter (Pmcr) and terminator (Tmcr) from M. barkeri (Supplementary Figure S5A). The insertion efficiency was obtained by comparing the number of Pur^R transformants from the editing plasmid pMCp3-g1-100-uid (8,800 ± 2,400 CFUs) and plasmid pMCp4 (117,000 ± 12,000 CFUs) (Figure 3A). The positive rate for 20 transformants was determined by colony PCR using the primers veri2/veri8 (Figure 3B). To ensure the genome copies within one transformant were all edited (homozygous chromosome), a second set of primers (veri5/veri12) was used to detect the presence of possible mixed transformants (Supplementary Figure S5B). 100% of the transformants were identified with uidA cassette and no heterozygous genotypes were detected. Three isolates were randomly selected for Sanger sequencing to detect the uidA gene cassette and all the edits were positive (Supplementary Figure S5C and Supplementary Table S2), demonstrating the high performance for Cas12a-mediated gene insertion.

First, plasmid pMCp2-g1g9RNA was constructed to simultaneously target the ssuC and frhA (encoding the coenzyme F420 hydrogenase alpha subunit) genes (Figure 4A), where the gRNA cassette contains two tandem gRNA sequences (g1RNA and g9RNA) equipped with promoter PmtaC1 and terminator TmtaB1 (Supplementary Figure S6). Next, to assess the effect of double-site deletions, two sets of HR sequences that generate 100-bp deletions in each of the genes (see above) were designed and assembled into pMCp2-g1g9RNA, producing the plasmid pMCp3-g1g9–100. Among the obtained Pur^R transformants (107 ± 23 CFUs per 2 μg DNA), eight were randomly selected for colony PCR. Primers veri1/veri2 and veri11/Cp54 were used to target the genome regions flanked the ssuC-and frhA-breakages, separately (Figure 4B). All eight transformants were positive and the desired gene deletions were detected in both ssuC and frhA regions. Three isolates were randomly picked for Sanger sequencing and all clones were correctly edited (Supplementary Figure S7A).

To evaluate the versatility of our CRISPR/Cas12a toolbox for multiplex editing, simultaneous gene deletion and insertion at two independent sites were attempted with the plasmid pMCp3-g1-uid-g9-100. This construct facilitates the insertion of the uidA cassette into the ssuC locus as well as makes a 100-bp deletion in the frhA gene in one round of transformation (Supplementary Figure S7B). Out of the 280 CFUs of Pur^R transformants, six were randomly selected for colony PCR verification using primers veri2/veri8 and veri9/veri13 (Supplementary Figure S7C). 5/6 were edited correctly in both sites (i.e., contain the uidA cassette within the ssuC gene and exhibit a 100 bp-deletion within the frhA gene), which indicates the potential of using CRISPR/Cas12a system in multi-tasks genome editing in metabolic engineering applications.

To demonstrate iterative editing with the CRISPR/Cas12a system, the plasmid curing efficiency was calculated for the positively engineered M73-uidA isolate. For plasmid removal, the isolate was cultivated in HS-methanol medium without puromycin for five serial transfers (i.e., 1% inoculum of cells in early stationary phase was utilized in each transfer). Subsequently, cells were streaked out on HS solid medium containing 20 μg/mL 8ADP for counterselection. Seven 8ADP^R transformants were randomly selected to detect the curing efficiency and the consistency of gene editing. As a result, 7/7 of the isolates exhibited outright plasmid removal, while 5/7 had correctly inserted uidA cassettes (Supplementary Figure S8). To explore a faster procedure, plasmids removal efficiency after only one round of transfer (same conditions as above) was evaluated. Same plasmid curing efficiency was obtained with the randomly selected 8ADP^R transformants, demonstrating the reliable application of the CRISPR/Cas12a system for performing multiple rounds of gene editing in M. acetivorans.

High-salt (HS) medium containing 125 mM methanol was used for cultivating M. acetivorans in single-cell form at 37°C (Sowers et al., 1993). M. acetivorans WWM73 was used as the host strain in this study and relevant derivatives are listed in Table 2. HS solid medium containing 1.4% agar (Bacto™ Dehydrated, Fisher Scientific) and 2 μg/mL puromycin (InvivoGen, Inc.) was used for screening M. acetivorans transformants. HS solid medium containing 1.4% agar and 20 μg/mL 8ADP (Sigma-Aldrich) was used for screening M. acetivorans transformants in the plasmid curing experiment. E. coli DH10B (Fisher Scientific) and E. coli XL10-Gold (Agilent Technologies) were used as the cloning hosts for constructing CRISPR plasmids. E. coli transformants were selected using lysogeny broth (LB) solid medium supplemented with 100 μg/mL ampicillin.

All plasmids used in this study are listed in Table 1. All primers are listed in Supplementary Table S1. PrimeSTAR® Max DNA Polymerase (Takara Bio) was used for amplifying gene fragments by PCR. SapphireAmp Fast PCR Master Mix (Takara Bio) was used for verifying E. coli and M. acetivorans transformants by colony PCR. HiFi DNA Assembly Master Mix (New England BioLabs) was used for Gibson assembly and T4 DNA Ligase (Promega) was used for ligation. crRNA sequences for multiplex genome targeting were synthesized and combined using the Gibson assembly method. Donor DNA with upstream and downstream of HR arms was inserted immediately downstream of the gRNA cassette using either Gibson assembly or ligation to construct the CRISPR/Cas12a genome editing system, known as plasmid pMCp3-X in Table 1.

Polyethylene glycol (PEG) 4,000-mediated transformation of M. acetivorans was performed with 2 μg DNA transformed according to the method described previously (Oelgeschläger and Rother, 2009). M. acetivorans transformants were grown on plates of HS solid medium with required antibiotics and incubated at 37°C for 10–15 days in an anaerobic jar with a controlled headspace of N₂/CO₂/1% H₂S (75/20/5). Chemically competent E. coli cells were used for the transformation of assembled and ligated plasmids.

The genomic DNA was extracted with Wizard® Genomic DNA Purification Kit (Promega). The Microbial WGS was performed using Illumina NovaSeq 6,000 by Novogene (UK) Company. Samples were sequenced in a paired-end 150 bp sequencing strategy and aligned with the M. acetivorans (C2A) reference genome (NC_003552.1) through BWA (Li and Durbin, 2009) software (parameters: mem-t 4-k 32-M). Structural variants (large deletions, insertions and translocations) were predicted using BreakDancer (Chen et al., 2009). Relevant data has been uploaded in National Center for Biotechnology Information (NCBI) with accession no. PRJNA1026875.