Genome Engineering of Eubacterium limosum Using Expanded

Subscriber access provided by EKU Libraries

Article

Genome engineering of Eubacterium limosum using expanded genetic tools and CRISPR-Cas9 system Jongoh Shin, Seulgi Kang, Yoseb Song, Sangrak Jin, Jin Soo Lee, JungKul Lee, Dong Rip Kim, Sun Chang Kim, Suhyung Cho, and Byung-Kwan Cho ACS Synth. Biol., Just Accepted Manuscript • DOI: 10.1021/acssynbio.9b00150 • Publication Date (Web): 02 Aug 2019 Downloaded from pubs.acs.org on August 3, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Synthetic Biology

Genome engineering of Eubacterium limosum using expanded genetic tools and CRISPR-Cas9 system

Jongoh Shin1,Ŧ, Seulgi Kang1,Ŧ, Yoseb Song1, Sangrak Jin1, Jin Soo Lee1, Jung-Kul Lee2, Dong Rip Kim3, Sun Chang Kim1,4, Suhyung Cho1, and Byung-Kwan Cho1,4,*

1Department

of Biological Sciences and KI for the BioCentury, KAIST, Daejeon 305-701, Republic of Korea

2Department 3Department

of Chemical Engineering, Konkuk University, Seoul 05029, Republic of Korea

of Mechanical Engineering, Hanyang University, Seoul 04763, Republic of Korea

4Intelligent

ŦThese

Synthetic Biology Center, Daejeon 305-701, Republic of Korea

authors contributed equally to this work.

*Correspondence and requests for materials should be addressed to B.-K.C. (email: [email protected])

1 ACS Paragon Plus Environment

ACS Synthetic Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Eubacterium limosum is one of the important bacteria in C1 feedstock utilization as well as in human gut microbiota. Although E. limosum has recently garnered much attention and investigation on a genome-wide scale, a bottleneck for systematic engineering in E. limosum is the lack of available genetic tools and an efficient genome editing platform. To overcome this limitation, we here report expanded genetic tools and the CRISPR-Cas9 system. We have developed an inducible promoter system that enables implementation of the CRISPR-Cas9 system to precisely manipulate target genes of the Wood-Ljungdahl pathway with 100% efficiency. Furthermore, we exploited the effectiveness of CRISPR interference to reduce the expression of target genes, exhibiting substantial repression of several genes in the WoodLjungdahl pathway and fructose-PTS system. These expanded genetic tools and CRISPR-Cas9 system comprise powerful and widely applicable genetic tools to accelerate functional genomic study and genome engineering in E. limosum.

Keywords: Eubacterium limosum; acetogen, inducible promoter; CRISPR-Cas9; CRISPR interference; Wood-Ljungdahl pathway


Page 2 of 39

Page 3 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Eubacterium limosum is a Gram-positive acetogenic bacterium (acetogen) capable of growing autotrophically with C1 feedstocks such as carbon monoxide (CO) or carbon dioxide (CO2) with hydrogen (H2), producing butyrate, caproate, and mainly acetate.1-3 All acetogens utilize the Wood-Ljungdahl (WL) pathway4 for converting CO2 to acetate, which is considered as the most efficient pathway of all carbon-fixation mechanisms.5 Owing to these physiological properties, acetogens play important roles in the carbon cycle by generating large amounts of acetate from the atmosphere,6 and they have been used as biocatalysts for converting industrial synthesis gas to multicarbon chemicals via gas fermentation.7, 8 Especially, E. limosum is, unlike other wellstudied acetogens such as Clostridium autoethanogenum and Clostridium ljungdahlii, capable of converting C1 feedstocks (such as methanol) directly into acetate and butyrate.9, 10 Furthermore, substantial evidence has demonstrated that E. limosum is an important intestinal microbe in the human gut microbiome. E. limosum has been frequently isolated from human feces and the intestinal content of various animals.11 Recent studies have shown that low levels of E. limosum highly correlate with colonic inflammation,12 aging,13 and the effectiveness of anticancer immunotherapies,14 suggesting that E. limosum is represented in healthy gut microbiota. These physiological characteristics make E. limosum an attractive microbe for possible biotechnological applications in industrial synthesis gas fermentation, relevant low-carbon biofuel, and human probiotic therapeutics. Recently, genome-scale analysis has been conducted to better understand E. limosum at the molecular and systems levels. Those studies have included the completion of the whole genome sequence,1-3 global transcription start sites,3 and transcriptomic and translatomic data.16 Although E. limosum has been well investigated at the systems level, a major barrier to biotechnological applications is the lack of available genetic tools and efficient genome editing 3 ACS Paragon Plus Environment


technology. For example, a genome editing platform based on the clustered regularly interspaced short palindromic repeats (CRISPR) system derived from Streptococcus pyogenes has become a highly versatile genome editing tool in various microorganisms and eukaryotic cells.29 In acetogens, the CRISPR-Cas9 system has been successfully adapted as a genome-engineering tool in C. ljungdahlii30, 31 and C. autoethanogenum.32 In addition, a catalytically deactivated Cas9 (dCas9), which was named CRISPR interference (CRISPRi)34, was used to interfere with the transcription of a specific gene knockdown in C. ljungdahlii.33 To the best of our knowledge, only one heterologous gene expression system developed in our laboratory was published for E. limosum.16 Previous work showed only one replicating shuttle plasmid and selectable marker was tested, limiting available genome engineering platforms for the basic understanding, redesign, and engineering of metabolic pathways. Here, we report expanded genetic tools and CRISPR-Cas9 system for manipulation of the E. limosum genome. We first evaluated several different origins of replication, inducible promoters, and selection markers to demonstrate implementation of an inducible heterologous gene expression system for use in E. limosum. We then used these expanded genetic tools to construct a CRISPRCas9 editing system, demonstrating precise knock-out of the WL pathway via homologous recombination. Furthermore, we exploited capabilities of the CRISPRi technique to repress several genes in the WL pathway and fructose-PTS system. These data, thus, demonstrated that the developed genetic tools and CRISPR system are versatile metabolic engineering tools for E. limosum.

Results and Discussion Determination of suitable replicon and antibiotic concentration 4 ACS Paragon Plus Environment

Page 4 of 39

Page 5 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Several shuttle vectors have been applied to several acetogens such as Acetobacterium woodii,22 C. ljungdahlii,19 and C. autoethanogenum35 for metabolic engineering purposes to produce biochemicals and biofuels from C1 feedstocks. Prior to examination of available plasmids for heterologous expression in E. limosum, we tested antibiotic resistance to determine selectable markers in E. limosum using pJIR750ai and pMTL82254 carrying a chloramphenicol-resistance gene (catP) and an erythromycin-resistance gene (ermB). To select the appropriate concentration of antibiotics, minimal inhibitory concentration (MIC) was determined in a solid RCM agar plate. Electrocompetent cells were prepared using a protocol as previously described with slight modifications.26 For chloramphenicol, wild-type E. limosum cells were phenotypically resistant at > 40 μg/mL concentration (Supplementary Figure S1A) after 6 days, suggesting that the nitro group of chloramphenicol could be reduced by E. limosum, as in the case of clostridia.36 Further, erythromycin was rapidly degraded in acidic aqueous media.37 Alternatively, thiamphenicol (chloramphenicol derivative without the reducible nitro group) and clarithromycin (a pH-stable derivate of erythromycin) can be used for E. limosum. MIC values showed that 15 μg/mL of thiamphenicol and 1 μg/mL of clarithromycin were appropriate to isolate transformed clones (Figure 1A and 1B). The same number of both cells was used for the MIC test, the relatively higher CFU of the wild type strain at 5.0 µg/ml of thiamphenicol and 0.25–0.5 µg/ml of clarithromycin was possibly as a result of cell death induced by the electroporation of each plasmid. PCR amplification targeting catP and the 16s rRNA gene was performed to verify true positive results after each transformation (Supplementary Figure S1B). These results indicated that the chloramphenicol- and erythromycin-resistance genes are effective markers for selection of E. limosum transformants.



To select a suitable replicon, five plasmids carrying different replicons (pMTL82151, pBP1 ori+; pMTL83151, pCB102 ori+; pMTL84151, pCD6 ori+; pMTL85151, pIM13 ori+; pJIR750ai, pIP404 ori+) were selected and introduced into E. limosum by electroporation (Supplementary Table S2). High transformation efficiencies (1.4 × 102 transformants per μg of plasmid DNA) were observed for pJIR750ai; however, extremely low transformation efficiencies (0–1.8 transformants per μg of plasmid DNA) were observed for other plasmids (Figure 1C), suggesting that the pIP404 origin from C. perfringens38 is suitable for optimal plasmid replication in E. limosum. As a result, we selected pJIR750ai containing a pIP404 origin as the E. coli–E. limosum shuttle plasmid.

Construction of an inducible promoter system For efficient genome editing and transcriptional repression with the CRISPR system, it is crucial to control Cas9 or dCas9 expression levels in E. limosum. To this end, we attempted to express Cas9 and dCas9 proteins using a strong promoter isolated from ELM_c2885, which encodes the pyruvate ferredoxin oxidoreductase gene, according to previous omics data.16 However, only frameshift mutants of dCas9 were detected (Supplementary Figure S2), since high expression of dCas9 has a toxic effect on cell growth.39 To reduce expression of Cas9 or dCas9 proteins, an inducible promoter system was used to regulate expression. Recently, lac and tet promoters have been used for Cas9 protein expression in related microbes C. beijerinckii,40 C. autoethanogenum,32 and C. ljungdahlii.33 To choose an optimal inducible system of gene expression, we examined lac promoter,41 Pcm-tetO1, and Pcm-tetO2 promoter42 using the oxygen-independent fluorescent protein CreiLOV as a reporter gene (Figure 2A and Supplementary Figure S3A).43 Recombinant E. limosum cells harboring each inducible 6 ACS Paragon Plus Environment

Page 6 of 39

Page 7 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


promoter and reporter gene were grown in DSM135 medium under inducing (1 mM IPTG or 30 ng/mL anhydrotetracycline) or non-inducing conditions, and relative fluorescence level was measured at mid-exponential growth during heterotrophic growth. When each corresponding repressor was not included in each plasmid, fluorescent protein expressions under control of the Plac, PtetO1, and PtetO2 promoters were 2.5-, 2.4-, and 2.1-fold higher compared to the negative control (P < 0.0001), respectively, indicating each inducible promoter was functional in E. limosum. To confirm that gene expression was regulated by an inducer, each corresponding repressor was integrated into each CreiLOV expression vector (Figure 2B and Supplementary Figure S3B). In the pJIR-lacI-Plac-CreiLOV plasmid containing the lac repressor, very low expression was observed regardless of IPTG induction (Figure 2B), suggesting that IPTG is not effective or cannot be introduced into the cell by permease in E. limosum.44 Although high fluorescence was observed from both PtetO1 and PtetO2, CreiLOV expression was well regulated (1.7-fold higher, P < 0.0001) by the PtetO1 promoter, suggesting that the repression activity of PtetO2 was not functional. In summary, PtetO1 demonstrated much greater inducible activity than Plac and PtetO2. Moreover, the inducer anhydrotetracycline (aTc) had no toxic effect on cells at 30 ng/mL aTc, making it appropriate for heterologous gene expression (Figure 2C). Based on the results, we cloned the Cas9 and dCas9 genes into the tetracycline-inducible PtetO1 plasmid to generate pJIR-Cas9 and PJIR-dCas9 constructs, respectively (Supplementary Figure S3C). The expression of Cas9 and dCas9 proteins was verified by Western blot analysis (Figure 2D). Transformants carrying pJIR-Cas9 and pJIR-dCas9 showed expression of fulllength Cas9 and dCas9 proteins. The quantification of dCas9 expression after 24 h of cultivation showed that 30 ng/mL aTc induced the highest expression of dCas9 without severe growth 7 ACS Paragon Plus Environment


inhibition (Supplementary Figure S4). Furthermore, higher dosage levels of aTc appeared to cause dCas9 degradation as a cellular response to heterologous gene overexpression. Based on the results, we applied 30 ng/mL aTc to E. limosum in order to express Cas9 protein. Taken together, the results demonstrated that PtetO1 is suitable for Cas9 gene expression in E. limosum.

Evaluation of sgRNA availability in E. limosum genome Since the target sequences of CRISPR-Cas9 are defined in the 5′-end leading 20 nucleotides (nt) sequence of crRNA, Cas9 nuclease from S. pyogenes required crRNA and associated tracrRNA to deliver Cas9 to the targeted region in the genome.34 Artificially, crRNA and tracrRNA could be combined into single-guide RNA (sgRNA), and a targeted site could be reprogrammed by modifying 20 nt of the sgRNA.45 Besides sgRNA-DNA base pairing, a protospacer-adjacent motif (PAM) flanking the target region is also essential for Cas9 activity. Thus, selection of spacer sequences relies on their specific locations within PAMs (5′-NGG-3′) in the genome. To evaluate how widely sgRNA can be used for E. limosum genome engineering, we designed a genome-wide collection of sgRNAs for all genes (Figure 3A). Potential spacer sequences (N20– NGG) of sgRNA were predicted from the E. limosum genomic sequences. Among 542,869 sgRNA candidates, 14,328 multi-binding sgRNAs were eliminated. Previous studies demonstrated that Cas9 nuclease could induce DSB at the PAM-including DNA sequences that match spacer sequences imperfectly, resulting in undesired editing at some off-target sites.46, 47 To avoid this off-target editing, we calculated the off-target score using the sgRNA binding metric46 and selected precise sgRNA candidates, as previously reported.48 Finally, 170,691 highfidelity sgRNAs within the gene body were selected as candidate sgRNAs (Figure 3A and Supplementary Table S3). Accordingly, the genome-wide collection of sgRNAs can target the 8 ACS Paragon Plus Environment

Page 8 of 39

Page 9 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


coding region of 4,152 genes (97.0%) with an average coverage of 41 sgRNAs per gene (Figure 3B and 3C), indicating that most genes in E. limosum can be targeted by the CRISPR-Cas9 system. The remaining 3.0% of non-targeted genes are multi-copy genes, such as tRNAs, rRNAs, and phage-related and uncharacterized proteins (Figure 3B). Therefore, CRISPR-Cas9 can be used as an unbiased genetic engineering platform in E. limosum.

Precise gene deletion using CRISPR-Cas9-mediated homologous recombination We next investigated whether the CRISPR-Cas9 system can be utilized as a genetic tool for Knock-in/Knock-out in E. limosum. In the WL pathway of E. Limosum, the methyl-branch gene cluster (ELIM_c0957–c0961) encodes fhs1 (formyl-tetrahydrofolate synthetase, FTHFS, ELIM_c0957), fchA (formyl-THF cyclohydrolase, MTHFC, ELIM_c0958), folD (methyleneTHF dehydrogenase, MTHFD, ELIM_c0959), and metV and metF (methylene-THF reductase, MTHFR, ELIM_c0960–c0961) genes (Figure 4A). Further, the carbonyl-branch gene cluster (ELIM_c1647–c1655) encodes acsC and acsD genes (MET/CoFeSP complex, ELIM_c1650– c1651) and acsA and acsB genes (carbon monoxide dehydrogenase/acetyl-CoA synthase, ELIM_c1653 and ELIM_c1655). The methyl- and carbonyl-branch enzymes catalyze reduction of the formyl-group to the methyl-group and conversion of acetyl-CoA, respectively.3, 4 To exemplify the CRISPR system, two WL pathway-associated genes (folD and acsC) were targeted (Figure 4A). As a positive control, the ELIM_c0303 gene was also targeted, which is hypothetical and not expressed under heterotrophic and autotrophic growth conditions.16 For CRISPR-Cas9 mediated homologous recombination, we generated a knock-out plasmid (pJET-HA-sgRNA) containing an sgRNA cassette, erythromycin-resistant gene ermB cassette, and target homology arms (HAs) (Supplementary Figure S3D). An approximately 2.7 9 ACS Paragon Plus Environment


kbp template DNA for each target genes (ELM_c0303, folD and acsC) was assembled and cloned into the pJET1.2/blunt plasmid. For sgRNA expression, the BBa_J23119 promoter was selected (Supplementary Figure S3D), since it is a small (35 bp) and constitutive promoter with a well-defined transcription start site, 35 (TTGACA) and 10 (TATAAT) regions.49 From the genome-wide collection of sgRNAs, three sgRNAs for each gene were chosen on the nontemplate strand of the coding sequence. Two oligos, including the promoter, sgRNA 20 nt spacer sequence for each gene, and sgRNA scaffold, were annealed and amplified using PCR (Supplementary Figure S3D). Subsequently, the sgRNA fragments were cloned into pJET-HA (Supplementary Figure S3D) by an infusion reaction, and each pJET-HA-sgRNA plasmid was then individually transformed into cells containing pJIR-Cas9 by electroporation (Supplementary Figure S3E). Transformants were selected on an RCM 1.5% agar plate supplemented with 1 μg/mL clarithromycin. Further, PCR analysis was performed for the target-specific region (TR) and for both the 5′ and 3′ junctional regions (5′J and 3′J) to verify the desired knock-out genotype (Figure 4B). As expected, 1013 bp, 874 bp, and 963 bp of TR PCR products were generated from each wild-type allele, and 2017 bp, 1941 bp, and 1946 bp of TR PCR products were generated from ELIM_c0303, folD, and acsC mutant cells, respectively, implying ermB gene cassette integration at the target region (Figure 4C). Moreover, the 5′- and 3′-junction PCR analysis, covering both ends of the homology arms, confirmed the desired junctions generated by insertion of the ermB gene cassette to each target site (Figure 4C). Furthermore, to identify a pure clonal population of mutants, seven colonies were randomly selected from the screening plate containing pJET-HA-sgRNA-acsC. All colonies were composed of the mutant population, as evidenced by 1946 bp, 1085 bp, and 1882 bp PCR amplicons generated from the target region, 10 ACS Paragon Plus Environment

Page 10 of 39

Page 11 of 39 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


5′-junction, and 3′-junction, respectively (Supplementary Figure S5A). To rule out any insertion and deletion, 5′- and 3′-junction PCR products were additionally purified and sequenced (Supplementary Figure 5B–5D). The desired genotypes without any mutations were found at both 5′- and 3′-junctions, indicating all targets were precisely engineered. Taken together, these data demonstrated that the CRISPR-Cas9 system can be used to seamlessly edit any target gene in E. limosum.

FolD and AcsA are essential genes for autotrophic growth in E. limosum To assess the roles of folD and acsC in autotrophic acetogenesis of E. limosum, three knock-out and wild-type strains were grown on CO2 and H2 (80/20, v/v) as a carbon source and reducing agent, respectively. Under autotroph growth conditions, the folD and acsC strains were unable to grow within 96 h, whereas the ELIM_c0303 and wild-type strains reached an optical density (OD600) of ~0.3 after 96 h (Figure 4D). As a main product of acetogenesis, acetate production of the ELIM_c0303 and wild-type strains was dependent on autotrophic cell growth. However, the folD and acsC strains displayed no acetate production on CO2-H2 growth condition, indicating interruption of the acetogenesis process. This is in agreement with the essential role of methylene-THF dehydrogenase and MET/CoFeSP complex in the WL pathway.4 The phenotypic effect of inactivating folD and acsC is consistent with the acsA strain of C. autoethanogenum, with acsA deletion by ClosTron mutagenesis.23 This result demonstrates that folD and acsC genes of the WL pathway are essential for autotrophic growth conditions.

Effect of CRISPR interference system on target gene repression Next, we tested the ability to knock-down gene expression by CRISPR-dCas9. Specifically, five 11 ACS Paragon Plus Environment


target genes were selected that have important roles in the WL pathway and PTS fructosespecific system: fhs1, folD, acsC, acsD, and ptsF (Figure 5A). From the genome-wide collection of sgRNAs, five sgRNAs for each gene were selected on the non-template strand near the experimentally confirmed transcription start site.3 Each sgRNA expression cassette was amplified and cloned into pJIR-dCas9 as described above (Supplementary Figure S3F). Each pJIR-dCas9-sgRNA plasmid was transformed into E. limosum, and transformants were selected on an RCM 1.5% agar plate supplemented with thiamphenicol. The CRISPRi-mediated knockdown was induced at low cell density (OD600 < 0.05), and we confirmed the repression of target genes under heterotrophic condition using quantitative RT-PCR (Figure 5B). The level of folD mRNA was most repressed (>84% repressed, P < 0.02) relative to the pJIR-dCas9 control. For the other four targeted genes, the transcript levels were significantly reduced (>30% repressed, P < 0.02) in the presence of both dCas9 and the specific sgRNA. To examine the effects of CRISPRi-based gene knock-down on cell growth, we measured cell growth under both fructose and CO2-H2 conditions for 64 h. Although the growth rates of all transformants containing pJIR or pJIR-dCas9 were slightly decreased, the final OD was the same as wild-type under heterotrophic conditions (Supplementary Figure 6A). Heterotrophic growth analysis showed the repression of ptsF-induced growth retardation (P < 0.03) (Supplementary Figure 6B), indicating that the ptsF contribute to fructose metabolism. Interestingly, repression of the acsC gene also arrested cell growth significantly (P < 0.01), suggesting that the gene has an important role under heterotroph conditions, similar to acsA of C. autoethanogenum.23 However, comparison of the effects of CRISPRi-based gene repression was difficult in the autotrophic growth assay (Supplementary Figure 6D). All transformants containing pJIR or pJIR-dCas9 showed growth defects (OD600

Genome Engineering of Eubacterium limosum Using Expanded

Recommend Documents