Development of a CRISPR/Cas9 System for Methylococcus capsulatus In Vivo Gene Editing

In this study, we targeted the development and evaluation of broad-host-range CRISPR/Cas9 gene-editing tools in order to enhance the genetic-engineering capabilities of an industrially relevant methanotrophic biocatalyst. The CRISPR/Cas9 system developed in this study expands the genetic tools available to define molecular mechanisms in methanotrophic bacteria and has the potential to foster advances in the generation of novel biocatalysts to produce biofuels, platform chemicals, and high-value products from natural gas- and biogas-derived methane. Further, due to the broad-host-range applicability, these genetic tools may also enable innovative approaches to overcome the barriers associated with genetically engineering diverse, industrially promising nonmodel microorganisms.

Methanotrophic bacteria are key players in Earth's biogeochemical carbon cycle and are of increasing industrial interest for their capacity to utilize methane as a sole carbon and energy source (13). The model gammaproteobacterial methanotroph Methylococcus capsulatus has been extensively studied for decades and is currently used for the industrial production of single-cell protein. Broad-host-range replicative plasmids containing RP4/RK2, RSF1010, and pBBR1 replicons are functional in M. capsulatus and have enabled the development of promoter-probe vectors and heterologous gene expression in the organism (14,15). Further, chromosomal insertions and unmarked genetic mutations using allelic-exchange vectors and sucrose or p-chlorophenylalanine counterselection have been reported (15)(16)(17)(18)(19). These tools have served as a basis for the recent expansion of the methanotroph genetic toolbox that has enabled several proteobacterial methanotrophs to be engineered to convert methane-rich natural gas and aerobic-digestion-derived biogas into high-value products (20)(21)(22)(23)(24)(25). Notably, these tools also lay the foundation for the development of advanced CRISPR genome-editing systems that facilitate multiplex or high-throughput gene-editing strategies. The development of advanced genome-editing tools offers a means to enable rapid evaluation of fundamental methanotrophic governing mechanisms while expanding metabolic engineering capabilities in these hosts for methane sequestration, bioremediation, and biomanufacturing.
In this study, we developed broad-host-range CRISPR/Cas9 gene-editing tools and evaluated their efficacy in the methanotroph M. capsulatus. Using the CRISPR/Cas9 system, we demonstrated editing of methanotroph-harbored plasmid DNA by introducing in vivo point mutations in a gene encoding green fluorescent protein (GFP) to generate a blue fluorescent protein (BFP) variant. Further, we successfully achieved chromosomal editing by generating a soluble methane monooxygenase (sMMO) mutant strain via the introduction of a premature stop codon in the mmoX open reading frame using the Cas9 D10A nickase. The CRISPR/Cas9 tools developed here will facilitate the development of advanced methanotrophic biocatalysts and have potential utility in an array of nonmodel, industrially promising bacteria.

RESULTS AND DISCUSSION
Development of a broad-host-range CRISPR/Cas9 gene-editing system. We employed the superfolder GFP (26) reporter to evaluate the functionality and strength of heterologous and native M. capsulatus promoters to be used for expression of Cas9 nuclease and gRNA components. We tested whether pCAH01, an RK2-based broad-host-range expression plasmid that contains the inducible tetracycline promoter/operator (P tetA ), previously demonstrated to function in the related gammaproteobacterial methanotroph Methylomicrobium buryatense 5GB1 (21,23), was also functional in M. capsulatus. The P tetA promoter exhibited strong inducible activation in M. capsulatus, as indicated by an ϳ10-fold increase of GFP fluorescence in pCAH01::GFP-harboring cells after exposure to the anhydrotetracycline (aTc) inducer (Fig. 1A). Based on the ability to temporally control gene expression, Cas9-and the Cas9 D10A nickase variant-encoding genes were cloned downstream of P tetA in pCAH01Sp R to generate pCas9 ( Fig. 2A) and pCas9 D10A , respectively (see Fig. S1A in the supplemental material).
Next, we evaluated the activity of M. capsulatus promoters to select a suitable promoter to drive constitutive gRNA expression. Promoters from the gamma subunit of particulate methane monooxygenase 1 (P pmoC1 ), particulate methane monooxygenase 2 (P pmoC2 ), the sMMO hydroxylase component mmoX (P mmoX ), and methanol dehydrogenase mxaF (P mxa ) genes were cloned upstream of a GFP coding sequence in the RSF1010-derived vector pQCH (see Fig. S1B), and promoter activity was assessed via fluorescence. As previously demonstrated via transcriptomic studies (19,27,28), P pmoC1 , P pmoC2 , and P mxa were highly active in M. capsulatus, driving significant GFP transcription under both copper-replete and copper-depleted conditions (Fig. 1B). As expected, we detected limited P mmoX promoter activity under copper-replete growth conditions; however, an ϳ7.5-fold increase in GFP expression was observed when cells were grown under copper-depleted conditions ( Fig. 1B) (19,27,28). Notably, the copper switchregulated P mmoX promoter may be employed for the inducible expression of heterologous genes in M. capsulatus. Both P pmoC1 and P pmoC2 were functional in Escherichia coli, but, interestingly, P mxa and P mmoX promoter activity was not detected in the organism (Fig. 1C), presumably due to the absence of required regulatory factors unique to M. capsulatus (15). Based on these results, we chose the P mxa promoter to express the gRNA due to the undetectable levels of GFP expression from this promoter in E. coli in order to avoid gRNA expression during cloning or conjugation procedures. A schematic of the broad-host-range plasmid pBBR1 containing a P mxa -expressing gRNA and a 1-kb DNA repair template (pgRNA) is depicted in Fig. 2B. Initial experiments evaluating Cas9 or Cas9 D10A expression in M. capsulatus indicated that nuclease expression did not affect bacterial growth in the absence of gRNA expression (Fig. 2D). Next, Cas9 or Cas9 D10A was coexpressed with an mmoX-targeting gRNA to evaluate nuclease targeting and activity using cell viability as a phenotypic readout. Transformants coexpressing Cas9 and pgRNA-mmoX exhibited ϳ99% cell death compared to cells expressing Cas9 without a gRNA (Fig. 2E). The surviving ϳ1% represent the background of the system and may serve as a putative limiting factor for gene-editing efficiency (29). Conversely, DNA digestion by Cas9 D10A did not result in cell death (Fig. 2E) and, assuming Cas9 D10A is functional under these experimental conditions, indicates that dsDNA breaks have a high degree of lethality while ssDNA nicks are more efficiently repaired by native M. capsulatus systems. Fluorescence intensity was measured from cells grown to an OD 600 of 0.5 in LB liquid medium. In panels B and C, the promoterless pQCH::GFP plasmid was used as a negative control. The data represent the fluorescence intensity normalized to OD 600 and are depicted as mean RFU and standard deviations (SD) from 3 independent replicates. ***, P Ͻ 0.001; *, P Ͻ 0.05; ns, not significant.

CRISPR/Cas9 in Methylococcus capsulatus
Applied and Environmental Microbiology In vivo plasmid editing. To evaluate CRISPR/Cas9 gene editing in M. capsulatus, we employed a screening assay with a direct fluorescent readout that shifts the emission and excitation of GFP to that of BFP following targeted gene editing (30). We constructed a vector (pgRNA-GFP BFP ) with a gfp-targeting gRNA and a 1-kb DNA repair template containing the GFP point mutations 194C¡G, 196T¡C, and 201T¡G, which introduces the missense codon substitutions T64S and Y65H to convert GFP to BFP while concurrently abolishing the Cas9 PAM site ( Fig. 3A; see Fig. S2 in the supplemental material). No detectable BFP fluorescence, only GFP fluorescence, was observed in cells expressing pgRNA-GFP BFP and GFP, demonstrating that the native homologousrecombination machinery did not integrate the DNA repair template into the gfp locus in the absence of Cas9 and that BFP is not expressed from the pgRNA-GFP BFP repair template (Fig. 3B). Positive transformants coexpressing Cas9, pgRNA-GFP BFP , and GFP were analyzed for BFP fluorescence to determine editing efficiency. In the absence of Cas9 induction, we observed ϳ5% BFP-positive transformants after selection (Fig. 3B), indicating that leaky Cas9 expression from P tetA is sufficient for gene editing. The induction of Cas9 during the conjugal transfer of pgRNA-GFP BFP significantly increased plasmid DNA-editing efficiency, with 71% of the transformants originally encoding GFP now expressing BFP (Fig. 3B). Fluorescence microscopy of isolated BFP-expressing colonies showed that all the cells uniformly expressed BFP, with no GFP-expressing cells Expression of the gRNA is driven from the M. capsulatus mxaF promoter (P mxa ). (C) Experimental design schematic of the CRISPR/Cas9 gene-editing system. M. capsulatus harboring pCas9 was conjugated with E. coli S17 harboring pgRNA on NMS mating agar supplemented with 500 ng/ml aTc. After 48 h of conjugation, the biomass was spread onto NMS selection agar containing 500 ng/ml aTc, gentamicin, and spectinomycin until colonies appeared. (D) OD 600 of bacterial cultures 24 h postinoculation with (ϩaTc) or without (ϪaTc) Cas9 or Cas9 D10A induction. The cultures were inoculated at an OD 600 of 0.1. (E) CFU of Cas9-or Cas9 D10A -expressing M. capsulatus after conjugation with pgRNA-mmoX. Empty pBBR1 plasmid was used as a negative control. The data are depicted as mean CFU and SD from 3 independent replicates. *, P Ͻ 0.05; ns, not significant. observed in the colony population (Fig. 3C). Sequence analysis of the fluorescentprotein-encoding loci from transformants identified as BFP positive confirmed the incorporation of the p.T64S and p.Y65H GFP-to-BFP mutations (Fig. 3D).
Similar to the experimental design with Cas9 described above, we tested whether the Cas9 D10A nickase could also be utilized for targeted DNA editing. In the absence of Cas9 D10A induction, no BFP fluorescence was detected in pgRNA-GFP BFP -expressing transformants (Fig. 4A). However, when Cas9 D10A and pgRNA-GFP BFP were coexpressed, we found that 71% of the transformants exhibited BFP fluorescence at varying intensities (Fig. 4A). Intriguingly, many BFP-expressing transformants were also positive for GFP fluorescence, with the degree of BFP fluorescence intensity positively correlated with a decrease in GFP fluorescence intensity (Fig. 4B). Fluorescence microscopy showed that transformant colonies consisted of both GFP-and BFP-expressing cells (Fig. 4C), presumably due to the expansion of plasmid copies that either were not cleaved by Cas9 D10A or were repaired after cleavage without incorporation of the homologous DNA repair template.  In vivo genome editing. We next evaluated Cas9-and Cas9 D10A -mediated chromosomal editing by targeting the mmoX gene encoding the sMMO hydroxylase component. We constructed a vector that harbored an mmoX-targeting gRNA and a DNA repair template containing an HpaI endonuclease restriction site, GTTAAC, which concurrently introduces a nonsense mmoX C151X (mmoX TAA ) mutation (pgRNA-mmoX TAA ) ( Fig. 5A; see Fig. S3 in the supplemental material). Transformants coexpressing pgRNA-mmoX TAA and Cas9 or Cas9 D10A were screened by colony PCR and subsequent digestion with HpaI endonuclease. We were unable to isolate an mmoX TAA transformant in Cas9-and pgRNA-mmoX TAA -expressing tranformants. In contrast, we identified the targeted mmoX TAA edit in 2% (5/250) of the Cas9 D10A nickase-and pgRNA-mmoX TAA -expressing transformants, as verified by HpaI digestion (Fig. 5B). Incorporation of the mmoX TAA nonsense mutation into the chromosome was also verified by sequence analysis (Fig. 5C). Further, the colorimetric sMMO activity assay demonstrated that the mmoX TAA strain was unable to convert naphthalene to naphthol after copper-depleted growth, confirming disruption of sMMO functionality (Fig. 5D).
Inducing single-nick-assisted HDR by the Cas9 D10A nickase has been employed in other microbes when wild-type Cas9-based editing has been unsuccessful (7-9). Previous studies have demonstrated that single-nick-assisted HDR undergoes repair via an independent mechanism with higher fidelity than dsDNA break-induced repair (31). Our data indicate that the Cas9 D10A -mediated nick induces higher chromosome recombination efficiency than a wild-type Cas9 dsDNA break, leading to increased numbers of transformants and enhanced mmoX locus-editing efficiency ( Fig. 2D and 5). Intriguingly, contrary to the differential observed for chromosomal editing, plasmid-editing efficiencies were identical whether Cas9 or Cas9 D10A was utilized. Future studies are thus needed to understand the differences in plasmid-and chromosome-editing efficiencies observed during the course of these investigations. Notably, successful incorporation of exogenous DNA into plasmid or genomic DNA by M. capsulatus was achieved via native recombination machinery. Evaluating the efficacy of heterologous recombinases, such as those employed in lambda red recombineering, may offer a potential means to enhance Cas9-mediated editing while also decreasing repair template size requirements (32, 33).

Conclusions.
Rational metabolic-engineering pursuits in methanotrophic bacteria require several metabolic mutations in order to enhance metabolic flux, end products, stress tolerances, and substrate utilization for the improved production of bio-based products (20). The tools developed here may enable CRISPR-based multiplex gene knockout strategies that can accelerate this time-and labor-intensive process. Further, CRISPR-based gene editing, together with high-throughput oligonucleotide synthesis, presents a path to genetic-library construction targeting rate-limiting metabolic enzymes to isolate strains with enhanced or altered characteristics (10,12). The ability to utilize Cas9 for DNA targeting makes it possible to also leverage the suite of available Cas9 variants (e.g., dCas9) (34-41) for transcriptional control, development of novel regulatory circuits, or optimization of CRISPR/Cas editing efficiency to enable advanced synthetic biology applications in methanotrophic bacteria.
In this study, we developed dual-plasmid, broad-host-range CRISPR/Cas9 tools and demonstrated targeted plasmid and chromosomal DNA editing with Cas9 and Cas9 D10A in the methanotroph M. capsulatus. This genetic system represents an advance in methanotroph molecular microbiology via expansion of the genetic toolbox. These advanced genetic tools may facilitate innovative strain engineering strategies that enable the development of methanotrophic biocatalysts for the production of biofuels, platform chemicals, and high-value products from methane. Further, novel molecular mechanisms underlying methanotroph biology can be probed with the addition of CRISPR/Cas9 to the methanotroph genetic toolbox. The replicons utilized in the CRISPR/ Cas9 system developed here are recognized by phylogenetically diverse bacteria; thus, they have the potential to facilitate facile genetic querying and innovative strainengineering strategies for the development of industrial biocatalysts in an array of nonmodel microbes.

MATERIALS AND METHODS
Bacterial strains and cultivation conditions. The strains used in this study are described in Table  1. Methylococcus capsulatus (Bath) was cultured on modified nitrate mineral salts (NMS) agar supplemented with 5 M CuSO 4 (the formulation is shown in Table S1 in the supplemental material), unless otherwise indicated, at 37°C inside stainless-steel gas chambers (Schuett-biotec GmbH) containing 20% (vol/vol) methane in air (42). The NMS agar was supplemented with 100 g/ml kanamycin, 100 g/ml spectinomycin, and/or 30 g/ml gentamicin for selection and cultivation of the respective M. capsulatus pQCH with promoterless superfolder GFP This study pQCH::P mmoX -GFP P mmoX cloned upstream of GFP in pQCH::GFP This study pQCH::P mxa -GFP P mxa cloned upstream of GFP in pQCH::GFP This study pQCH::P pmoC1 -GFP P pmoC1 cloned upstream of GFP in pQCH::GFP This study pQCH::P pmoC2 -GFP P pmoC2 cloned upstream of GFP in pQCH::GFP This study pQCH::P pmoC1 -BFP P pmoC1 -BFP cloned from P pmoC1 -GFP This study pBBR1MCS-5 pBBR oriT aacC1 45 pBBR1-GFP P mxa -gRNA-GFP cloned into pBBR1 This study pBBR1-GFP BFP P mxa -gRNA-GFP BFP cloned into pBBR1 This study pBBR1-mmoX P mxa -gRNA-mmoX cloned into pBBR1 This study pBBR1-mmoX TAA P mxa -gRNA-mmoX TAA cloned into pBBR1 This study strains. E. coli strains were cultured on lysogeny broth (LB) agar or in LB liquid medium at 37°C at 200 rpm. LB liquid medium was supplemented with 50 g/ml kanamycin, 50 g/ml spectinomycin, 10 g/ml gentamicin, and/or 100 g/ml ampicillin for selection and cultivation of the respective E. coli strains. Broad-host-range plasmids were transferred to M. capsulatus via biparental mating using E. coli S17-1 cells on NMS mating agar (the formulation is shown in Table S1), as described previously (25). Prior to conjugation, M. capsulatus biomass harboring pCas9 or pCas9 D10A was spread on NMS mating medium supplemented with 500 ng/ml aTc and incubated at 37°C inside stainless-steel gas chambers (Schuettbiotec GmbH) containing 20% (vol/vol) methane (99.97% purity) in air for 24 h. A schematic of the experimental design for using the CRISPR/Cas9 system in M. capsulatus is shown in Fig. 2C.
To evaluate promoter activity, M. capsulatus harboring GFP reporter plasmids was spread onto NMS agar supplemented with 0 M or 5 M CuSO 4 . GFP expression from the tetracycline promoter/operator (P tetA ) in pCAH01 was induced by plating on NMS agar supplemented with 500 ng/ml aTc. Strains were incubated at 37°C inside stainless-steel gas chambers (Schuett-biotec GmbH) in 20% (vol/vol) methane in air for 72 h, and the GFP fluorescence intensity was quantified as described below. Promoter activity was determined in E. coli subcultured 1/100 in LB liquid medium and incubated for ϳ3 h to an optical density at 600 nm (OD 600 ) of 0.5 at 37°C at 200 rpm. A 200-l volume of E. coli cell suspension was transferred to a 96-well plate for quantification of the GFP fluorescence intensity as described below.
Cloning and genetic manipulation. The plasmids used in this study are described in Table 1. The primers and synthetic DNA fragments used in this study were synthesized by Integrated DNA Technologies, Inc. (IDT), and are described in Table 2 and Table S2 in the supplemental material, respectively. Plasmids and DNA inserts were amplified using Q5 High-Fidelity 2ϫ Master Mix (NEB), assembled using Gibson NEBuilder HiFi DNA assembly (New England Biolabs), and transformed into Mix and Go competent E. coli strain Zymo 10B (Zymo Research), according to the manufacturers' instructions. Genetic constructs were verified by Sanger sequencing (Genewiz). The Cas9 open reading frame was amplified, using primers TT16 and CAH537 (Table 2), from Addgene plasmid no. 42876 (29) and cloned into pCAH01Sp R via Gibson assembly. The Cas9 D10A nickase variant was generated by site-directed mutagenesis with primers TT143 and TT144 (Table 2) using the QuikChange primer design program and protocol (Agilent). Single gRNAs containing a 20-mer adjacent to the PAM site on the target DNA and editing cassettes were synthesized by IDT (see Table S2) and cloned into the pBBR1MCS-5 vector via Gibson assembly. pQCH was constructed by replacing a 3,312-bp region of RSF1010 containing the sulfR, smrA, and smrB genes with the kan2 locus from pAWP78 (25). To evaluate promoter activity, native M. capsulatus promoters were cloned upstream of the superfolder green fluorescent protein-encoding gene (26) into pQCH.
GFP and BFP expression quantification. To evaluate Cas9-and Cas9 D10A -mediated plasmid editing, fluorescence intensity was measured in a Fluostar Omega microplate reader (BMG Labtech) at an excitation wavelength ( ex ) of 485 nm and an emission wavelength ( em ) of 520 nm (GFP) or a ex of 355 nm and a em of 460 nm (BFP). For GFP-to-BFP gene-editing experiments, the data represent relative fluorescence units (RFU) of the measured BFP intensity relative to pQCH::P pmoC1 -GFP control intensity normalized to the cell density.
Verification of mutations by colony PCR and HpaI digestion. To evaluate Cas9-and Cas9 D10Amediated genomic editing of the mmoX locus, colony PCR was performed with primers TT290 and TT291 ( Table 2) using Taq 2ϫ Master Mix (NEB) according to the manufacturer's instructions. The PCR mixture was used directly as the template for HpaI endonuclease (NEB) digestion, and the edited strains were identified by positive DNA digestion visualized by DNA electrophoresis. Targeted editing of the mmoX locus in positive transformants was verified by sequence analysis.
Colorimetric sMMO assay. M. capsulatus sMMO activity was tested with a colorimetric assay as previously described (43). Briefly, ϳ1e6 cells were spotted onto NMS agar with or without 5 M CuSO 4 and cultured at 37°C. After 96 h of growth, ϳ300 to 400 mg naphthalene (Sigma-Aldrich) crystals was placed into the petri dish lid and incubated with the bacteria for 1 h at 37°C to allow conversion of naphthalene to naphthol. After incubation, 20 l of freshly prepared 5-mg/ml o-dianisidine (Sigma-Aldrich), which turns purple in the presence of naphthol, was added directly to the M. capsulatus biomass. Color development was allowed to occur for 15 min at 37°C. Statistical analysis. Statistical analysis of data was performed and graphical representations were created using GraphPad Prism 6.0 software. Determination of statistical significance between two comparisons was achieved using an unpaired t test. Determination of statistical significance between multiple comparisons was achieved using a one-way analysis of variance (ANOVA) followed by Dunnett's test with the appropriate controls. Normal distribution and equal variance between test groups were assumed prior to performing statistical tests using Prism software.

SUPPLEMENTAL MATERIAL
Supplemental material for this article may be found at https://doi.org/10.1128/AEM .00340-19.

ACKNOWLEDGMENTS
We thank Bryon Donahoe for providing fluorescence micrographs and Ellsbeth Webb for technical assistance.
This work was conducted at the National Renewable Energy Laboratory (NREL), operated by the Alliance for Sustainable Energy, LLC, for the U.S. Department of Energy (DOE) under contract no. DE-AC36-08GO28308.
The work was supported in part by the Laboratory Directed Research and Development (LDRD) and Director's Fellowship Programs at NREL.
The views expressed in the article do not necessarily represent the views of the DOE or the U.S. Government.