Type IV-Like Pili Facilitate Transformation in Naturally Competent Archaea

Microbial organisms adapt and evolve by acquiring new genetic material through horizontal gene transfer. One way that this occurs is natural transformation, the direct uptake and genomic incorporation of environmental DNA by competent organisms. Archaea represent up to a third of the biodiversity on Earth, yet little is known about transformation in these organisms. Here, we provide the first characterization of a component of the archaeal DNA uptake machinery. We show that the type IV-like pilus is essential for natural transformation in two archaeal species. This suggests that pili are important for transformation across the tree of life and further expands our understanding of gene flow in archaea.

ABSTRACT

Naturally competent organisms are capable of DNA uptake directly from the environment through the process of transformation. Despite the importance of transformation to microbial evolution, DNA uptake remains poorly characterized outside of the bacterial domain. Here, we identify the pilus as a necessary component of the transformation machinery in archaea. We describe two naturally competent organisms, Methanococcus maripaludis and Methanoculleus thermophilus. In M. maripaludis, replicative vectors were transferred with an average efficiency of 2.4 × 10³ transformants μg⁻¹ DNA. In M. thermophilus, integrative vectors were transferred with an average efficiency of 2.7 × 10³ transformants μg⁻¹ DNA. Additionally, natural transformation of M. thermophilus could be used to introduce chromosomal mutations. To our knowledge, this is the first demonstration of a method to introduce targeted mutations in a member of the order Methanomicrobiales. For both organisms, mutants lacking structural components of the type IV-like pilus filament were defective for DNA uptake, demonstrating the importance of pili for natural transformation. Interestingly, competence could be induced in a noncompetent strain of M. maripaludis by expressing pilin genes from a replicative vector. These results expand the known natural competence pili to include examples from the archaeal domain and highlight the importance of pili for DNA uptake in diverse microbial organisms.

IMPORTANCE Microbial organisms adapt and evolve by acquiring new genetic material through horizontal gene transfer. One way that this occurs is natural transformation, the direct uptake and genomic incorporation of environmental DNA by competent organisms. Archaea represent up to a third of the biodiversity on Earth, yet little is known about transformation in these organisms. Here, we provide the first characterization of a component of the archaeal DNA uptake machinery. We show that the type IV-like pilus is essential for natural transformation in two archaeal species. This suggests that pili are important for transformation across the tree of life and further expands our understanding of gene flow in archaea.

INTRODUCTION

Horizontal gene transfer—the movement of genetic material between organisms—is a major driver of evolution in all domains of life. This can occur through (i) transduction, the acquisition of genetic material through infection by viruses or phages, (ii) conjugation, the transfer of genetic material through direct cell-to-cell contact, or (iii) transformation, the uptake of DNA directly from the environment. Transformation requires that organisms be in a physiological state known as the competent state, and these organisms are often referred to as naturally competent. While mechanisms whereby DNA can be acquired through transduction, cell-to-cell DNA transfer, or self-transmissible plasmids have been described for both bacteria and archaea (1–5), natural transformation machinery has been described only for members of the bacterial domain (6). To date, several members of the archaea have been described as naturally competent. Thermococcus kodakarensis, Methanothermobacter marburgensis, Methanococcus voltae, and a mutant of Pyrococcus furiosus can take up extracellular DNA with varying efficiencies (7–11).

Most naturally competent bacteria rely on pili for DNA uptake (6). Well-studied examples include the competence pilus of Bacillus spp. (12), the type IV DNA uptake pilus of organisms such as Neisseria spp. and Vibrio spp. (13–16), and the Flp pilus of Micrococcus luteus (17). In all cases, there is evidence that pili bind extracellular DNA and transport it across the outer membrane and/or the peptidoglycan layer of the cell envelope. Once DNA is transported to the outer leaflet of the inner membrane, the ComEC DNA transporter transports single-stranded DNA to the cell cytoplasm (6). Single-stranded DNA binding proteins stabilize the DNA, and cellular repair machinery drives integration into the chromosome. An alternative DNA uptake system that is homologous to the bacterial type IV secretion system has been described in Helicobacter pylori (18).

Type IV pilus-like structures play numerous roles in cellular physiology. In addition to DNA uptake, they can drive cellular motility (19), initiate adhesion to surfaces (20), and mediate contacts between cells (1). While no role has been found for either the uptake of environmental DNA or conjugative DNA transfer for archaeal pili, they are known to be important for adherence to surfaces/cells (20). Many archaea also possess a flagellum that is essential for swimming motility and is a homolog of the type IV pilus (19). Despite their diverse functions, type IV-like pili share the same core structure. Among the methanogenic archaea, Methanococcus maripaludis is a model organism for the study of type IV-like pili (21), where genetic and microscopic studies have identified several genes that are required for proper pilus assembly (22–24). A euryarchaeal type IV prepilin peptidase (EppA) processes EppA-dependent (Epd) prepilin proteins. The prepilin protein encodes an N-terminal secretion signal, and a hydrophobic domain serves as scaffold for polymerization of the pilus filament (25). A membrane protein complex composed of EpdJ, EpdK, and an ATPase, EpdL, is essential for the assembly and disassembly of the filament (21, 26). In the case of N-glycosylated pili, there is also an archaeal oligosaccharyltransferase (AglB) that transfers oligosaccharides to the growing pilus (27). The M. maripaludis flagellum is composed of homologous proteins and undergoes a similar construction process with FlaK serving as the peptidase, FlaI as the ATPase, and FlaJ as the membrane anchor (28).

Based on the wide distribution of DNA uptake pili in the bacterial domain, we hypothesized that pili are essential for DNA uptake by naturally competent archaea. To test this, we characterize the role of pili in natural transformation of M. maripaludis and Methanoculleus thermophilus. This is the first demonstration of efficient natural transformation for either of these organisms. Leveraging their transformability to make mutants, we found that pili are essential for transformation. The identification of pili that facilitate DNA uptake in archaea suggests pili are widely conserved components of microbial natural transformation machinery.

RESULTS

M. maripaludis strain JJ is naturally competent.

To determine the efficiency of DNA uptake by natural transformation in M. maripaludis, stationary-phase cultures (optical density at 600 nm [OD₆₀₀] = 1.0 to 1.2) of two strains, JJ and S2, were inoculated with various concentrations of the replicative vector pLW40 (encoding puromycin resistance) or pLW40neo (encoding neomycin resistance) (29). After a 4-h incubation, cells were plated onto selective medium to determine the efficiency of transformation. A 4-h incubation was selected to allow for DNA internalization and expression of the antibiotic resistance gene and to remain consistent with outgrowth times used for the well-established polyethylene glycol (PEG)-based transformation method (30). The number of transformants recovered increased with increasing concentrations of DNA in strain JJ, suggesting that this strain is naturally competent (Fig. 1A). No antibiotic-resistant colonies were observed in strain S2, and there was not a significant difference in the recovery of transformants with pLW40 or pLW40neo. When transformed using the PEG method (30), both strains S2 and JJ were proficient for plasmid uptake (Fig. 1B).

FIG 1.

Natural transformation of M. maripaludis. (A) Natural transformation of M. maripaludis strains JJ and S2 with various quantities (0, 0.1, 0.5, 1, 5, or 10 μg) of either pLW40neo (neo) or pLW40 (pur) added to a 5-ml culture. Number of transformants is presented as the total number of antibiotic-resistant CFU from 5 ml of culture. (B) Comparison of transformation efficiencies with pLW40neo using either the natural transformation (Natural) or the polyethylene glycol (Chemical) method (30). Transformation frequencies for natural and PEG-mediated transformation of strain JJ were 4.4 × 10⁻⁶ and 8.4 × 10⁻³ transformants CFU⁻¹ μg⁻¹ DNA, respectively (average of triplicates). The PEG-mediated transformation frequency of strain S2 was 4.8 × 10⁻⁴ transformants CFU⁻¹ μg⁻¹ DNA (average of triplicates). Data are averages from three independent experiments, and error bars represent one standard deviation around the mean. Transformants were selected on medium containing the appropriate antibiotic. *, P < 0.05.

The natural transformation efficiency of strain JJ with pLW40neo was 2.4 × 10³ transformants μg⁻¹ DNA when DNA was added at a concentration of 1 μg ml⁻¹ (Fig. 1). This concentration was used for all subsequent experiments. We did not find a significant difference in the frequency of transformation when DNA was added to cells during exponential phase (OD₆₀₀, 0.4). Transformation frequencies were 2.2 × 10⁻⁶ ± 1.7 × 10⁻⁶ and 4.4 × 10⁻⁶ ± 7.2 × 10⁻⁷ transformants CFU⁻¹ μg⁻¹ DNA (mean ± standard deviation for triplicate cultures) for exponential or stationary-phase cultures, respectively.

We next determined if M. maripaludis strain JJ could be transformed with nonreplicating plasmids using the natural transformation method. Strain JJ was transformed following the same protocol with the suicide vector pCRuptneo (31) using two independent chromosomal integration loci. When plated onto selective medium, transformation efficiencies of 2 to 4 transformants μg⁻¹ DNA were observed (see Fig. S1 in the supplemental material). This is comparable to previous observations of DNA uptake efficiency in naturally transformed Methanococcus spp., where transformation required recombination of DNA into the chromosome (10, 30). We hypothesize that the lower efficiency of transformation with integrative vectors is due to the requirement for homologous recombination for vector maintenance. Using the PEG method, integrative vectors also had transformation efficiencies several orders of magnitude lower than replicative vectors (Fig. S1).

We additionally tested M. maripaludis strains C5, C6, and C7 for natural transformation (32). While transformants could be generated in each of these organisms, experiment to experiment variability limited our ability to confidently calculate the efficiency of transformation (see Fig. S2 in the supplemental material). For this reason, strains C5, C6, and C7 were omitted from subsequent experiments.

M. thermophilus strain DSM 2373 is naturally competent.

Genetic studies in the methanogenic archaea are currently limited to organisms from the Methanococcales or the Methanosarcinales (33). Organisms from the order Methanomicrobiales are ubiquitous in anaerobic environments and are metabolically distinct from either of these groups (34). In an effort to further study methanogens from the Methanomicrobiales, we have focused efforts toward generating a transformation system for M. thermophilus due to its rapid growth in liquid medium, ability to form colonies on plates in 5 to 7 days, and ability to utilize either hydrogen or formate as electron donors for methanogenesis (35). Additionally, we have found that M. thermophilus is capable of DNA uptake via natural transformation (Fig. 2).

FIG 2.

Natural transformation of M. thermophilus. M. thermophilus was transformed with pJAL1inter (0, 0.1, 0.5, 1, 5, or 10 μg added to a 5-ml culture) and plated on medium containing neomycin. Number of transformants is presented as the total number of antibiotic-resistant CFU from 5 ml of culture. The natural transformation frequency of M. thermophilus was 4.3 × 10⁻⁵ transformants CFU⁻¹μg⁻¹ DNA when 1 μg ml⁻¹ of DNA was used for transformation (average of triplicates). Data are averages from three independent experiments, and error bars represent one standard deviation around the mean.

To date, there have been no plasmids isolated from the Methanomicrobiales, so we used a suicide vector to determine the DNA uptake efficiency of M. thermophilus. pJAL1inter (see Fig. S3 in the supplemental material) was constructed and contains a thermostable variant of kanamycin/neomycin phosphotransferase (36) and a fragment of DNA that is homologous to a noncoding region of the M. thermophilus genome (37). Cells were grown to stationary phase (OD₆₀₀ of 0.20 to 0.25) in medium with H₂-CO₂ as the energy source, and DNA was added directly to the culture tube. Due to the slower growth rate of M. thermophilus, we tested for DNA uptake with extended incubation times before plating on selective medium to grow transformants. Transformation efficiencies increased with longer incubations with 1.1 × 10² ± 2.4 × 10¹, 3.9 × 10² ± 5.7 × 10¹, and 2.7 × 10³ ± 1.0 × 10³ transformants μg⁻¹ DNA observed at 4, 8, and 18 h, respectively (mean ± standard deviation for triplicate cultures). We did not extend incubations beyond 18 h to avoid cell lysis from prolonged incubation. Similar to M. maripaludis, the number of transformants recovered increased with increasing concentrations of DNA (Fig. 2), no antibiotic-resistant colonies were observed when cells were not exposed to plasmid, and we did not find a significant difference when DNA was added to cells during exponential phase (OD₆₀₀, 0.1). Transformation frequencies were 4.2 × 10⁻⁵ ± 3.2 × 10⁻⁵ and 4.3 × 10⁻⁵ ± 2.4 × 10⁻⁵ transformants CFU⁻¹ μg⁻¹ DNA (mean ± standard deviation for triplicate cultures) for cultures in exponential or stationary phase, respectively. Based on these results, DNA was added at a concentration of 1 μg ml⁻¹, and an 18-h incubation was used in all subsequent experiments with M. thermophilus.

Pili are essential for DNA uptake in M. maripaludis.

Due to the involvement of pili in DNA uptake in bacterial systems, we hypothesized that the type IV-like pilus is essential to DNA uptake in members of the archaea. M. maripaludis contains two type IV appendages, flagella and pili (Fig. 3A). To determine if one or both are essential for DNA uptake, we generated the following in-frame deletion mutants for the prepilin and preflagellin peptidases of M. maripaludis strain JJ: the ΔeppA mutant to target pili and the ΔflaK mutant to target flagella (20). A defect in DNA uptake was observed only in the ΔeppA strain (Fig. 3B). We additionally generated a mutant lacking the membrane anchor and ATPase of the type IV-like pilus (the ΔepdJKL strain) to ensure that observed phenotypes were not a result of the processing of a secondary peptidase target by EppA. The transformation efficiency of the ΔepdJKL strain was 2 orders of magnitude lower than that of the wild type (Fig. 3B). These data suggest that DNA uptake is pilus dependent, and the type IV pilus-like flagellar filament does not play a role in DNA uptake. The transformation defects of the ΔeppA and ΔepdJKL strains were rescued when these genes were expressed in trans on the replicative vector pLW40neo (see Fig. S4 in the supplemental material).

FIG 3.

The type IV-like pilus is important for natural transformation of M. maripaludis. (A) Model of M. maripaludis type IV-like filaments. Flagella and pili are both type IV-like structures with similar composition. First, a prepilin or preflagellin is processed by a peptidase (EppA for pili and FlaK for flagella) and loaded onto an anchor (EpdJK for pili and FlaJ for flagella). The arrow indicates peptidase-dependent cleavage of prepilins or preflagellins and incorporation of monomers into the filament. The energy for the polymerization/depolymerization is driven by an ATPase (EpdL for pili and FlaI for flagella) (cartoon adapted from reference 53 with permission of the publisher). (B) Natural transformation of M. maripaludis strain JJ and mutants lacking eppA, epdJKL, or flaK. Data are averages from three independent experiments, and error bars represent one standard deviation around the mean. *, P < 0.05.

Heterologous expression of pilin genes enables transformation of M. maripaludis S2.

We initially hypothesized that a specific pilin was responsible for DNA uptake in M. maripaludis strain JJ and the lack of transformation in strain S2. The genomes of strains JJ and S2 share >95% average nucleotide identity (2-way ANI) with similar gene contents (38–41). To identify candidate pilins, we searched for genes encoding predicted class III signal peptide-containing proteins (a signature of EppA substrates [25]) that are unique to strain JJ. Of those identified on the JJ genome, MMJJ_RS05685 (NCBI accession number WP_104838031) was found to be missing in strain S2. In addition to the class III signal peptide motif, this gene encodes a predicted N-terminal secretion signal and a hydrophobic domain that is similar to the hydrophobic domain of the M. maripaludis major pilin subunit, EpdE (23) (see Fig. S5 in the supplemental material). The C-terminal domain of MMJJ_RS05685 differs from EpdE; however, this is common for different type IV and type IV-like pilins. While additional experiments are necessary for verification, these features suggest that MMJJ_RS05685 encodes a pilin.

MMJJ_RS05685 is present in a putative operon with epdE and two hypothetical proteins, MMJJ_RS05690 and MMJJ_RS05700 (41) (Fig. 4A). To test their roles in transformation, we generated in-frame deletion mutants in strain JJ for each gene in the putative operon. Both the MMJJ_RS05685 and ΔepdE mutants were defective for transformation (Fig. 4B). Complementation of these deletions rescued these mutants (Fig. S4). Deletion of MMJJ_RS05690 or MMJJ_RS05700 did not have an effect on the transformation efficiency of strain JJ (Fig. 4B). Other studies have shown that deletion of any one of the pilins encoded on the M. maripaludis genome is sufficient to abolish pili on the cell surface (22, 23); therefore, we cannot deduce whether DNA uptake defects in pilin mutants are due to a specific role in DNA binding or a general defect in pilus assembly.

FIG 4.

Pilins are essential for DNA uptake in M. maripaludis. (A) Cartoon representation aligning the genomic regions surrounding epdE in strains JJ and S2. (B) Transformation efficiencies with pLW40neo for deletion mutants in genes from the putative epdE operon in strain JJ. (C) Transformation efficiencies with pLW40 for S2 heterologously expressing MMJJ_RS05685, epdE_JJ, or epdE_S2 from pLW40neo. Data are averages from three independent experiments, and error bars represent one standard deviation around the mean. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Next, we expressed epdE from strain S2 (epdE_S2) or strain JJ (epdE_JJ) and tested for natural transformation in strain S2. We also heterologously expressed MMJJ_RS05685. Successful expression of these genes was deduced from the observed phenotypes. Strain S2 expressing epdE_S2 from pLW40neo had an average transformation efficiency of 1.7 × 10³ transformants μg⁻¹ DNA (Fig. 4C), and heterologous expression of epdE_JJ resulted in an average efficiency of 62 transformants μg⁻¹ DNA. Heterologous expression of MMJJ_RS05685 resulted in an average transformation efficiency of 3.2 × 10² transformants μg⁻¹ DNA (Fig. 4C). While the transformation efficiency varied depending on the specific gene that was expressed, it is clear that expression of any one of them was sufficient to induce strain S2 into a competent state.

Pili are essential for DNA uptake in M. thermophilus.

We hypothesized that the use of pili for DNA uptake is conserved in naturally competent archaea. To test this, we adopted the M. maripaludis pCRuptneo-based selection/counterselection system to generate M. thermophilus mutants lacking flagella and type IV-like pili. M. thermophilus possesses a single prepilin peptidase (flaK) that likely processes both prepilins and preflagellins (37). Attempts to mutagenize flaK were unsuccessful; therefore, we targeted membrane components to make mutants specifically lacking genes for flagella (ΔflaIJ mutant) or pili (ΔepdJKL mutant) (see Fig. S6 in the supplemental material).

Both the ΔepdJKL and ΔflaIJ mutants were transformed with the pJAL1inter vector. Consistent with what was seen in M. maripaludis, the M. thermophilus ΔepdJKL mutant was unable to take up extracellular DNA, and the transformation efficiency of the ΔflaIJ mutant was comparable to that of the wild type (Fig. 5). Because the ΔepdJKL mutant could not be transformed, we were unable to attempt complementation of this strain. These data verify that pili are important to DNA uptake in phylogenetically diverse archaea. Additionally, our ability to recover in-frame deletion mutants verifies that transformation of M. thermophilus with pCRuptneo can be used to generate chromosomal deletions.

FIG 5.

The type IV-like pilus is necessary for transformation of M. thermophilus. Natural transformation of M. thermophilus strain DSM 2373 and mutants lacking epdJKL or flaIJ. Strains were transformed with pJAL1inter and selected on medium containing neomycin. Data are averages from three independent experiments, and error bars represent one standard deviation around the mean. **, P < 0.01.

DISCUSSION

Transformation is an important driver of evolution in microbial systems; however, the mechanistic basis of DNA uptake by members of the archaeal domain has remained uncharacterized. We have shown that, similar to several known DNA uptake systems in bacteria (6), pili are essential for transformation of M. maripaludis and M. thermophilus. The ability to induce competence in M. maripaludis strain S2 by heterologous expression of epdE_JJ further highlights the importance of pili to natural transformation. It is particularly interesting that expressing epdE_S2 from a replicative vector induced strain S2 into a competent state. This suggests that strain S2 possesses all of the required cellular machinery for DNA uptake but that, under our laboratory growth conditions, native expression of pili is insufficient to support natural transformation. Despite the sequence similarity between EpdE_JJ, EpdE_S2, and the N-terminal domain of MMJJ_RS05685 (see Fig. S5 in the supplemental material), there was a difference in the transformation efficiency when the genes encoding these proteins were expressed in strain S2. We hypothesize that this is due to differences in the amino acid sequences of these proteins or posttranslational modifications of the pilus filament. EpdE_S2 is a glycoprotein with multiple sites occupied by a glycan (23). EpdE_JJ lacks one of these glycosylation sites (Fig. S5). This difference may account for the low transformation efficiency observed when epdE_JJ was expressed in strain S2. In any case, we find it significant that strains expressing pilin genes or MMJJ_RS05685 became competent. We could not identify a homolog of EpdE in M. thermophilus or other naturally competent archaea, suggesting that M. thermophilus uses a different pilin as the major structural subunit of the pilus filament.

The mechanism of DNA transport across the cell membrane remains uncharacterized. DNA transfer systems that are dependent on direct cell-to-cell contact have been described for both Crenarchaeota and Euryarchaeota. In Sulfolobus spp., transfer of DNA between cells is mediated by a membrane-bound complex called the crenarchaeal system for exchange of DNA (Ced) (1). The genes encoding Ced machinery have not been identified in Methanococcus spp. or Methanoculleus spp., indicating that these proteins are likely not required for natural transformation. Conjugal DNA transfer between bacteria and archaea has been demonstrated in laboratory cultures (42), but in this case, DNA transfer was catalyzed by machinery present in the bacterial donor. In bacteria, the best characterized DNA transporters are ComEC proteins found in organisms, such as Bacillus spp., Vibrio spp., H. pylori, and others (13, 43, 44). Homologs of ComEC have not been found in archaea. It was recently proposed that Escherichia coli possesses a pathway for DNA internalization dependent on an ATP-binding cassette (ABC)-type transporter (45). M. maripaludis possesses several ABC transporter systems, many of which are essential for growth (46); it is possible that one of these participates in DNA uptake. In any case, the DNA transporter in members of the archaea remains to be identified and is the subject of ongoing work.

The observation that cells can be transformed with circular plasmids is consistent with what has been seen in other naturally competent archaea (11). It is unknown whether DNA is linearized or maintained as intact circular DNA during transport across the cell membrane. If DNA is linearized before transport, there must be a mechanism to regenerate the circularized vector. Nonhomologous end joining (NHEJ) could catalyze this reaction; however, this activity has not been observed in archaea, and genes encoding putative homologs of the Ku proteins required for NHEJ are rare in this group (47). Regardless of the uptake mechanism, internalized plasmids can either replicate autonomously or recombine into the chromosome through homologous recombination (47). Archaeal histones may also play a role in this stage of transformation (48).

Finally, we have described the first system to generate targeted mutations in a methanogen from the order Methanomicrobiales. These organisms are widely distributed in anaerobic environments, including aquatic and terrestrial habitats, anaerobic and wastewater digesters, and the rumen where they catalyze the terminal step in the degradation of organic matter (49). Genetic manipulation will enable further characterization of this physiologically and metabolically interesting group. Transformation of M. thermophilus expands the known naturally competent archaea to include species from the following four classes: Methanobacteria, Methanococci, Methanomicrobia, and Thermococci (7–11). Thus, competence is a widely distributed feature in the phylum Euryarchaeota. The data presented here suggest that pili are crucial for transformation across the tree of life and play an important role in horizontal gene transfer in natural archaeal populations.

MATERIALS AND METHODS

Strains, medium, and growth conditions.

Strains used in this study are listed in Table S1 in the supplemental material. M. maripaludis was acquired from William Whitman and M. thermophilus DSM 2373 was purchased from DSMZ (Leibniz Institut, Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany).

M. thermophilus DSM 2373 and its mutants were routinely grown anaerobically at 55°C with agitation in McTry medium. Liquid McTry medium was composed of NaHCO₃ (5 g liter⁻¹), NaCl (14 g liter⁻¹), NH₄Cl (0.5 g liter⁻¹), KCl (0.34 g liter⁻¹), MgCl₂·6H₂O (2.8 g liter⁻¹), MgSO₄·6H₂O (3.5 g liter⁻¹), CaCl₂·2H₂O (0.14 g liter⁻¹), K₂HPO₄ (70 mg liter⁻¹), FeSO₄ (9.5 mg liter⁻¹), NaCH₃COO (4.8 g liter⁻¹), Casamino Acids (2 g liter⁻¹), peptone (4 g liter⁻¹), tryptone (4 g liter⁻¹), and resazurin (1 mg liter⁻¹). Trace mineral and vitamin solutions were also added (32). All chemicals were purchased from Fisher Scientific. Medium was brought to a boil, and headspace was degassed over a stream of N₂-CO₂ (80:20) before dithiothreitol (0.5 g liter⁻¹) was added as a reductant. Medium was aliquoted into Balch tubes in a Coy anaerobic chamber (2 to 3% H₂, 10% CO₂, balance nitrogen in the atmosphere) before the headspace was vacuum exchanged to H₂-CO₂ (80:20). Tubes were autoclaved before use. (NH₄)₂S was added to tubes immediately before inoculating with culture material (0.1 ml of a 0.5% solution added to a 5-ml culture tube). For solid medium, 1.5% noble agar was included and NaHCO₃ was reduced to 2.5 g liter⁻¹. Plates were poured in the anaerobic chamber. When necessary, the following antibiotics were added after autoclaving medium at the noted concentrations: neomycin (0.25 mg ml⁻¹), 8-azahypoxanthine (0.2 mg ml⁻¹), or 6-azauracil (0.2 mg ml⁻¹). These concentrations were experimentally determined as the minimum concentrations necessary to prevent growth. M. thermophilus DSM 2373 was grown in liquid medium under 280 kPa H₂-CO₂ (80:20). For growth on solid medium, plates were added to an anaerobic incubation vessel containing 10 to 20 ml of 25% Na₂S as a sulfur source, pressurized to 140 kPa H₂-CO₂ (80:20), and incubated at 55°C.

M. maripaludis was grown on McCas medium at 37°C with agitation as described previously (50), except that sodium sulfide was replaced by ammonium sulfide for liquid medium. McCas is a modification of McC medium (32) with yeast extract replaced with Casamino Acids. When necessary, the following antibiotics were added to medium at the noted concentrations: neomycin (1 mg ml⁻¹ for liquid medium and 0.5 mg ml⁻¹ for plates), puromycin (2.5 μg ml⁻¹), or 6-azauracil (0.2 mg ml⁻¹).

Plasmid construction.

Plasmids and primers are listed in Table S2 in the supplemental material. To generate in-frame genomic deletion mutant constructs for both M. maripaludis and M. thermophilus, PCR products for the genomic regions flanking the gene of interest were fused and assembled into pCRuptneo (31). These flanking genomic regions each covered 500 bp of sequence homology and were fused through Gibson assembly (51) or restriction cloning using an AscI restriction site. Deletion constructs were assembled into the XbaI and NotI sites of pCRuptneo; this plasmid is described in reference 31 and has features for propagation in E. coli (origin of replication and ampicillin resistance gene) and for selection (neomycin selection) and counterselection (uracil phosphoribosyltransferase) in methanogens. For expression on pLW40neo, genes were PCR amplified and placed onto the vector under the control of the Methanococcus voltae histone promoter. Constructs were cloned into E. coli ig 5-alpha (Intact Genomics) by chemical transformation or by electroporation of E. coli DH5α. Transformed E. coli were selected on lysogeny broth agar medium containing ampicillin (50 μg ml⁻¹) before plasmids were extracted and transferred to methanogens following the described protocols. All constructs were sequence verified by Sanger sequencing at the University of Minnesota Genomics Center.

The pJAL1 plasmid was kindly provided as a gift from Thomas Lie, University of Washington. A map of this vector is provided in the supplemental material. Briefly, it contains features for propagation in E. coli (origin of replication and ampicillin resistance gene) and for selection (neomycin resistance gene) in methanogens. For integration in the M. thermophilus genome, pJAL1inter was constructed. pJAL1inter additionally contains an ∼500-bp genomic integration locus upstream of the gene WP_066955586 (NCBI locus tag) as a template for homologous recombination. This ∼500-bp region was placed into the vector using Gibson assembly (51).

Transformation of M. maripaludis.

All mutants were generated in an M. maripaludis strain lacking the gene for uracil phosphoribosyltransferase (Δupt mutant). For S2, this was strain MM901 (31). For JJ, we constructed a strain lacking upt to regenerate a new version of strain J901 (previously described in reference 52).

Cultures for natural transformation were inoculated with a 2% (vol/vol) inoculum in McCas medium with a 280 kPa H₂-CO₂ (80:20) headspace. After overnight (∼18 h) growth to stationary phase (OD₆₀₀, ∼1), cultures were moved into a room temperature Coy anaerobic chamber (2 to 3% H₂, 10% CO₂ atmosphere). Plasmid DNA from a PureLink quick plasmid miniprep kit (Invitrogen) was mixed with 0.5 ml fresh McCas medium and added to the culture using a syringe. Plasmid DNA was allowed to equilibrate in the anaerobic chamber for 1 h before transformation. The culture/plasmid mixture was pressurized to 280 kPa H₂-CO₂ (80:20) and incubated at 37°C with agitation for 4 h (∼2 doublings) to ensure plasmid uptake and expression of the antibiotic resistance gene. After transformation and outgrowth, cultures were placed onto antibiotic medium (neomycin or puromycin) to select for transformants. For integrative vectors, this generated a merodiploid that was allowed to resolve by overnight growth without selection. Mutants were isolated by plating onto 6-azauracil medium and screening colonies by PCR. For strains that were not transformed by the natural transformation method, polyethylene glycol (PEG) transformations were performed as previously described (30) with PEG 8000 purchased from Millipore Sigma (catalog number 89510-250G-F). For complementation studies, genes were placed on the pLW40neo vector under the control of the Methanococcus voltae histone promoter. Cultures containing complementation constructs were grown to stationary phase in medium containing neomycin before pLW40 was introduced. After the 4-h outgrowth, cells were plated onto agar medium containing puromycin. Neomycin was not present in the agar medium.

For transformation efficiency calculations, experiments were performed with 5 μg plasmid in a 5 ml culture (1 μg ml⁻¹ final concentration) unless otherwise specified. After outgrowth, cells were 10-fold serially diluted, plated onto antibiotic selective medium, and allowed to grow for 4 to 5 days in anaerobic incubation vessels pressurized to 140 kPa with H₂-CO₂ (80:20) at 37°C.

Transformation of M. thermophilus.

All genetic manipulations of M. thermophilus were performed in the wild-type background using pCRuptneo. M. thermophilus is naturally resistant to 6-azauracil (see Fig. S7 in the supplemental material), and the merodiploid with pCRuptneo integrated into the genome acquires 6-azauracil sensitivity due to the presence of M. maripaludis upt on the vector. Tests for transformation efficiency were performed with the pJAL1inter vector.

Cultures were inoculated with a 2% (vol/vol) inoculum into 5 ml fresh McTry medium under an atmosphere of 280 kPa H₂-CO₂ (80:20) and allowed to grow at 55°C for approximately 48 h to stationary phase (OD₆₀₀, 0.2 to 0.25). Cultures were moved into a room temperature Coy anaerobic chamber (2 to 3% H₂, 10% CO₂ atmosphere) for transformation. Plasmid DNA (5 μg) from a PureLink quick plasmid miniprep kit (Invitrogen) was mixed with 0.5 ml fresh McTry medium and added to the culture using a syringe. Plasmid DNA was allowed to equilibrate in the anaerobic chamber for 1 h before transformation. The culture/plasmid mixture was pressurized to 280 kPa H₂-CO₂ (80:20) and incubated at 55°C with agitation overnight (∼18 h) to allow for DNA uptake, integration into the genome, and expression of the antibiotic resistance cassette. These steps were followed for every experiment unless otherwise indicated in the text. After outgrowth, cultures were brought back into the anaerobic chamber, 10-fold serially diluted in sterile McTry medium, and spread onto agar medium with neomycin. Plates were incubated for 7 days at 55°C in anaerobic incubation vessels as described in the section on growth.

To generate deletion mutants, cultures were transformed with pCRuptneo containing deletion constructs. Following transformation and selection on neomycin, 1.3 × 10³ and 6.7 × 10³ transformants μg⁻¹ DNA were recovered for the ΔflaIJ and ΔepdJKL plasmid constructs, respectively. Transformants were grown in liquid medium containing neomycin before a single drop of culture was transferred using a syringe to 5 ml fresh McTry medium without antibiotic and grown to stationary phase. This allowed for recombination and loss of the vector from a subset of the population. Cultures were brought into the anaerobic chamber, 10-fold serially diluted in McTry liquid medium, and plated onto McTry agar medium containing 6-azauracil. After 7 days, multiple colonies were streak purified before screening by PCR for the mutation of interest. The 6-azauracil-resistant colonies contain a single allele, either wild type or mutant; therefore, we anticipated that 50% of recovered colonies should be mutants. For the ΔflaIJ construct, 5 out of 10 colonies screened were mutants. For the ΔepdJKL construct, 3 out of 8 colonies screened were mutants. Mutants were grown in McTry liquid medium to an OD₆₀₀ of 0.2 for subsequent experiments. Additionally, we verified that mutants were sensitive to neomycin, indicating loss of the pCRuptneo vector DNA.

Strains were stored as anaerobic freezer stocks. Stocks were prepared by anaerobically adding 3 ml of culture material to 2 ml of 50% (vol/vol) glycerol under an N₂ atmosphere in 5-ml serum bottles. These freezer stocks were stored under an N₂ atmosphere and remained viable (assayed by growth of cultures derived from the stock) for at least a year, the maximum length of time tested to date.

Statistical analysis.

Data are presented as means ± standard deviation unless otherwise indicated. One-way analysis of variance (ANOVA) was used to compare the means of groups. This was followed by post hoc pairwise t tests with Bonferroni correction to test for differences between experiments or strains. Differences were considered significant if adjusted P values were <0.05.