ABSTRACT
The heliobacteria are members of the bacterial order Clostridiales and form the only group of phototrophs in the phylum Firmicutes. Several physiological and metabolic characteristics make them an interesting subject of investigation, including their minimalist photosynthetic system, nitrogen fixation abilities, and ability to reduce toxic metals. While the species Heliobacterium modesticaldum is an excellent candidate as a model system for the family Heliobacteriaceae, since an annotated genome and transcriptomes are available, studies in this organism have been hampered by the lack of genetic tools. We adapted techniques for genetic manipulation of related clostridial species for use with H. modesticaldum. Five heliobacterial DNA methyltransferase genes were expressed in an Escherichia coli strain engineered as a conjugative plasmid donor for broad-host-range plasmids. Premethylation of the shuttle vectors before conjugation into H. modesticaldum is absolutely required for production of transconjugant colonies. The introduced shuttle vectors are maintained stably and can be recovered using a modified minipreparation procedure developed to inhibit endogenous DNase activity. Furthermore, we describe the formulation of various growth media, including a defined medium for metabolic studies and isolation of auxotrophic mutants.
IMPORTANCE Heliobacteria are anoxygenic phototrophic bacteria with the simplest known photosynthetic apparatus. They are unique in using bacteriochlorophyll g as their main pigment and lacking a peripheral antenna system. Until now, research on this organism has been hampered by the lack of a genetic transformation system. Without such a system, gene knockouts, site-directed mutations, and gene expression studies cannot be performed to help us further understand or manipulate the organism. Here we report the genetic transformation of a heliobacterium, which should enable future genetic studies in this unique phototrophic organism.
INTRODUCTION
Photosynthesis is one of the most important biological processes supporting the ecology of the planet, capturing the energy of the sun to power the food webs that encompass almost every living organism. Although oxygenic photosynthesis—in which O2 is released and CO2 is assimilated—is the most familiar type, there are six known bacterial phyla that contain species using a simpler anoxygenic process, in which a single photochemical reaction center is used to drive electron flow to make ATP. The heliobacteria were discovered in 1983 (1) and are still the only known phototrophic group in the phylum Firmicutes. Heliobacteria synthesize a homodimeric photochemical reaction center composed of two subunits of PshA and two of PshX (2). They are the only phototroph that uses bacteriochlorophyll g, an isomer of chlorophyll a, as their pigment (3). The heliobacterial reaction center is the simplest known (2) and is unique among photosynthetic reaction centers in being able to reduce both soluble ferredoxins and menaquinone, depending upon conditions (4, 5).
Heliobacteria are heterotrophs, requiring an organic carbon source such as acetate or pyruvate. They can grow in the dark fermentatively, much like their nonphototrophic cousins, the Clostridiales (3). The heliobacteria are anaerobic soil microbes that fix nitrogen (6) and can reduce heavy metals (7).
Heliobacterium modesticaldum strain Ice1 (herein called H. modesticaldum) was originally isolated from volcanic soil in the vicinity of hot springs in Iceland (8) and is emerging as a model heliobacterial species. It possesses the advantage of being the only known thermophilic member of the family, having an optimal growth temperature of 52°C (8). H. modesticaldum is also the only heliobacterial species whose complete genome has been sequenced and annotated (9, 10) and from which transcriptomes have been reported (11). This has allowed the cloning and expression of heliobacterial genes in other organisms, such as Escherichia coli and Rhodobacter sphaeroides (12, 13). Notably, it is also the strain from which the recent reaction center crystal structure was determined (2).
No native plasmid has been reported for any member of the heliobacteria, nor has any successful genetic transformation been previously reported. In other members of the Clostridiales, restriction-modification systems have been shown to hinder the introduction of plasmids. This has been attributed to restriction endonucleases (REs), which make a double-stranded break in DNA upon recognition of a specific sequence, typically 4 to 8 bp in length. The host organism prevents cleavage of its own chromosome by methylation of the recognition site, which blocks the action of the RE. This allows the organism to simultaneously avoid digesting its own genome and prevent invasion by foreign genetic components. The DNA methyltransferase (DMT) enzymes that catalyze methylation of the recognition sequence are typically coupled to the RE recognizing the same sequence. The restriction barrier can be overcome by premethylation of the plasmid before it is exposed to the RE. For example, to introduce plasmids into the firmicute Moorella thermoacetica, DMTs from the organism were expressed in E. coli and used to modify shuttle vectors prior to their introduction into M. thermoacetica by electroporation (14). In this study, we present methods for overcoming the restriction barrier for transformation by conjugation of premethylated plasmids into H. modesticaldum.
RESULTS
Development of solid media and a defined medium for H. modesticaldum.Growth of individual cells into colonies on solid medium is an essential tool for classical and molecular genetics. As we initially experienced difficulty growing H. modesticaldum on agar plates at its preferred growth temperature, we turned to the use of Gelrite, a derivative of gellan gum that had been used as a gelling agent for several thermophilic bacteria (15). We were able to mix H. modesticaldum cultures into 2× Gelrite maintained at 65°C, as this material does not solidify until exposed to divalent cations, and colonies grew embedded in the matrix.
We later discovered that H. modesticaldum could be grown on the surface of agar plates, but only if the atmosphere above contained CO2. We do not know the reason for this requirement, although there are several possibilities (see Discussion). As isolation of colonies from the surface of agar is much more convenient than extracting them from within the Gelrite matrix, we standardly perform all manipulations using agar. As far as we can tell, the same general results, in terms of numbers of colonies under different conditions, are obtained with either type of solid medium.
A chemically defined medium is required to perform replicable metabolic studies in H. modesticaldum, as well as make use of nutritional markers. Yeast extract (YE) represents a biologically derived component whose contents are not rigorously known and may vary from batch to batch. In order to determine which YE component is essential for growth, combinations of vitamins were added to medium lacking YE (termed “modified pyruvate mineral salts” [mPMS]) and inspected for growth after three serial cultures at 1/100 dilution (Table 1). It was found that H. modesticaldum has a strict requirement for biotin (vitamin B7) and either folate (vitamin B9) or its direct biosynthetic precursor (16), para-aminobenzoic acid (PABA, also known as vitamin B10). Additionally, when H. modesticaldum cells are grown in medium lacking vitamin B12 (cobalamin), they are initially pink in color and attain a lower late-log-phase density. The pink color is due to the lack of photosynthetic pigments absorbing in the visible and near-infrared range. In subsequent subcultures, the normal brown color returned (indicative of the synthesis of photosynthetic pigments) and the culture was indistinguishable from that grown in vitamin B12-supplemented medium.
Systematic determination of vitamin requirements for H. modesticalduma
MICs of antibiotics.Before successful transformations could begin, we had to determine MICs for antibiotics in H. modesticaldum on solid medium. These are tabulated in Table 2. As previously reported (17), this organism is very sensitive to penicillin derivatives such as ampicillin. Between the two aminocyclitol derivatives tested, tetracycline was more effective than spectinomycin. Thiamphenicol was much more effective at inhibiting growth than the structurally related compound chloramphenicol. Among the aminoglycoside antibiotics tested, apramycin and kanamycin had similar MICs (10 to 15 μg ml−1), but streptomycin’s MIC was so high (>400 μg ml−1) as to render it practically unusable.
MICs for H. modesticaldum grown on mPYE platesa
Cloning of methyltransferase genes.All attempts to transform H. modesticaldum by making use of electroporation, natural transformation, or conjugation were unsuccessful. We hypothesized that the root cause was restriction of the DNA after entry into the heliobacterial cell. Transformation of previously intractable members of the Clostridiales has been achieved by methylation of plasmid DNA either in vitro or in E. coli (14, 18). A search of the REBASE database (19) revealed 7 genes encoding REs and 13 genes encoding DMTs in the H. modesticaldum genome. We narrowed our focus using two criteria: (i) a nonnegligible number (reads per kilobase per million [RPKM] > 3) of transcript reads for the DMT gene must have been detected in the published transcriptome of cells growing in pyruvate-yeast extract (PYE) (11); (ii) the DMT gene must be located on the genome within four open reading frames of a gene encoding a RE and thus be likely to methylate the same recognition sequence. Six genes met these criteria: HM1_0771, HM1_2858, HM1_3004, HM1_3037, HM1_3076, and HM1_3075. The first (HM1_0771) was rejected, because it contains both the DMT and RE domains fused together. HM1_3076 and HM1_3075 code for the specificity and the methylase domains of a single DMT, respectively. Thus, the five remaining genes should together encode four heliobacterial DMTs.
The five selected genes were cloned into the multiple-cloning site of plasmid pBAD33 to create plasmid pPB347, a map of which is shown in Fig. 1. Plasmid pBAD33 was chosen as the backbone due to its low copy number, which can minimize toxicity if it expresses foreign genes deleterious in E. coli (20). An artificial operon was constructed in this plasmid, with the five heliobacterial DMT genes placed downstream of the arabinose-inducible araBAD promoter. This allows low-level expression in E. coli using the leakiness of the promoter in cells grown without either glucose or arabinose. Each DMT gene was immediately preceded by an E. coli ribosome binding site.
Map of plasmid pPB347. The backbone is from plasmid pBAD33, on which putative DMT genes HM1_2858, HM1_3076, HM1_3075, HM1_3037, and HM1_3004 were cloned each with separate ribosome-binding sites under the transcriptional control of the arabinose promoter. P15A ori, replication module for low-copy-number maintenance in E. coli; araC, gene coding for arabinose operon regulatory protein; CmR, chloramphenicol resistance gene. The positions of RE sites used for cloning are also shown (see Materials and Methods for details of construction).
Transformation of H. modesticaldum by conjugation with E. coli.There is no evidence for natural transformation of heliobacteria, nor have endogenous plasmids been reported in this family. Thus, we decided to employ conjugation using mobile vectors known to be transferable to members of the Clostridiales. Initial studies employed plasmid pJIR1457 (21), which was initially constructed for transformation of Clostridium perfringens and contains the traJ module necessary for conjugal transfer. E. coli strain S17-1, which has conjugal transfer helper genes integrated into its genome, was selected as the plasmid donor (22).
Potential shuttle vectors were cloned into S17-1 bearing either DMT helper plasmid pPB347 or the control vector pBAD33. The resulting dual-plasmid S17-1 strains were then conjugated with H. modesticaldum, and transformants were selected for resistance to erythromycin at 50°C. Conjugations from S17-1 harboring both pJIR1457 as the shuttle vector and pPB347 as the helper plasmid yielded many colonies growing in the presence of 10 μg/ml erythromycin (Fig. 2). Conjugations from S17-1 containing pJIR1457 alone (Fig. 2), pPB347 alone, or pJIR1457 with the pBAD33 control plasmid gave rise to no colonies. Moreover, the colonies resulting from conjugation with premethylated pJIR1457 were stably erythromycin resistant for many generations. These results are consistent with the idea that premethylation was required for survival of the pJIR1457 plasmid after its introduction into the heliobacterial cells by conjugation, followed by replication of the plasmid and maintenance of its methylation state over the course of generations.
Results of conjugations with and without DMT helper plasmids. Conjugations were performed with no E. coli (mock transformation; top row), S17-1 bearing plasmid pJIR1457 (middle row), or S17-1 bearing plasmid pJIR1457 and DMT helper plasmid pPB347 (bottom row). Pictures shown are of inverted vials containing colonies embedded in Gelrite.
Creation of a more useful shuttle vector.In our hands, plasmid pJIR1457 was difficult to manipulate and sequence. Analysis of the sequence revealed that a region next to the multiple-cloning site is duplicated on the backbone, which was the source of the difficulties. We therefore turned to the pMTL80000 series of modular plasmids, which are now widely used in clostridia (23). The major advantage of these plasmids is their modular nature, easily allowing exchange of the proteobacterial or firmicute replication modules and selectable markers. Unfortunately, none of the plasmids in the basic set provided by Chain Biotech yielded erythromycin-resistant colonies (Table 3). However, none of these plasmids used the same firmicute replication module that is on pJIR1457. We therefore purchased the additional module to construct plasmid pMTL86251, in which the firmicute replication module was replaced with the same pIP404 module as pJIR1457. This plasmid could be conjugated at relatively high efficiency into H. modesticaldum, but only from S17-1 containing the pPB347 helper plasmid (Table 3). For a typical conjugation, in which 2.4 × 107 donor cells are mixed with 4.0 × 107 heliobacterial cells, a conjugation efficiency of 5.8 × 10−5 transconjugants (tcg) (CFU donor)−1 translates to ∼1,400 heliobacterial tcg colonies.
Conjugation efficiencies using different Gram-positive replication modulesa
Optimization of the conjugation procedure.In order to optimize transformation, we systematically altered four of the conjugation parameters, the results of which are shown in Table 4. We first tested the effect of the growth medium in which the donor cells were grown. Since the conjugations are performed on solid modified PYE (mPYE) medium, we attempted growing the donor E. coli in liquid mPYE instead of Luria-Bertani (LB) broth to preadapt it to the new medium. Although E. coli grows to similar densities in mPYE and LB (data not shown), conjugations from donors pregrown in mPYE yielded fewer colonies.
Conjugation efficiencies under various conjugation conditionsa
The second variable tested was the temperature at which the conjugations were performed. The ideal growth temperature for E. coli is 37°C, whereas it is 52°C for H. modesticaldum. We tested the effect of raising the temperature at which the conjugations were performed from 37°C to 42°C. This led to an ∼2-fold increase in the number of tcg colonies obtained.
The third variable tested was performing the conjugation in the light or dark. For the purpose of making mutants that could abolish photosynthesis, investigators may wish to perform bacterial conjugations in which the cells are not exposed to light. Conjugations performed in darkness yielded about one-third the number obtained when cells were exposed to light.
The final variable tested was the time for which conjugations were performed. In order to save time or to remove the possibility of replicate tcg colonies (i.e., nonindependent clones), investigators may wish to incubate the donor and recipient cells together for a shorter period. The number of tcg colonies obtained after a 3-h conjugation was ∼15% that seen after an overnight conjugation.
Note that the last three variables also affect the division of heliobacteria, as they grow faster at 42°C and in the light but will complete fewer cell divisions in 3 h than in 16 h. Thus, we cannot say how much of the observed effects of these variables have to do with the increased number of conjugation events and how much to the increased number of cell divisions of the tcg cells. If one wishes to maximize the number of independent tcg colonies (e.g., transferring a library), then it would be advisable to perform the conjugation for a shorter period of time at 37°C.
Plasmid preparations from H. modesticaldum.Plasmid preparations from heliobacterial pMTL86251 tcg cells using mini-plasmid preparation kits employing lysozyme/alkaline lysis followed by spin column purification yielded only smears of DNA on agarose gels (Fig. 3, lanes 2 and 3), making it difficult to assess whether any intact plasmid had been recovered. However, when these preparations were used to transform chemically competent E. coli cells, transformant colonies were always obtained, and plasmid pMTL86251 could be recovered from them (data not shown). This led us to believe that the plasmid was present in the H. modesticaldum preparations but was being masked by degraded genomic and plasmid DNA. In other members of the Clostridiales, the barrier to recovery of intact plasmid DNA, particularly when using the alkaline lysis method, has been attributed to the presence of cell wall-associated DNase (24–26). This DNase is thought to be released into the suspension buffer upon treatment with cell wall weakening agents such as lysozyme. When cells are lysed and DNA is released into the buffer, it is exposed to the activity of DNase and is degraded.
Agarose gel of plasmid minipreps from H. modesticaldum using modified and unmodified isolation procedures. Lanes 1 and 7, 1-kb ladder (NEB, Ipswich, MA); lanes 2 to 5, HindIII-digested plasmid minipreps from H. modesticaldum tcg cells bearing pMTL86251, using the unmodified (lanes 2 and 3) and modified (lanes 4 and 5) spin column procedure; lane 6, plasmid miniprep (unmodified) of pMTL86251 from E. coli TOP10 cells.
By modifying the plasmid miniprep protocol to remove and/or inactivate cell wall-associated DNase before cell lysis (see Materials and Methods for details), plasmid preparations of markedly improved quality could be obtained. The DNA smears were much reduced, and a band for the plasmid that comigrated with pMTL86251 isolated from E. coli was observed (Fig. 3, lanes 4 to 6).
DISCUSSION
Knowledge of antibiotic sensitivity is necessary for selection of antibiotic-resistant tcg colonies of H. modesticaldum. The unusual sensitivity of H. modesticaldum to ampicillin, which specifically inhibits peptidoglycan synthesis, is consistent with the observation of extreme sensitivity of other members of the genus to penicillins. Specifically, Beer-Romero and coworkers (27) noted for other heliobacteria a sensitivity to penicillin G that was 103-fold or 104-fold greater than that observed for E. coli or Bacillus subtilis, respectively. The authors speculated that this sensitivity is related to the quantity and nature of the peptidoglycan synthesized by these organisms. The greater sensitivity to thiamphenicol than to chloramphenicol was not expected; it might be related to differences in stabilities of the two antibiotics in mPYE medium at 50°C.
At every streptomycin concentration tested, up to 20 large streptomycin-resistant colonies were observed. Streptomycin resistance is often acquired through a single point mutation of the rpsL gene. Thus, we do not recommend the use of streptomycin as a selectable marker in heliobacteria. Erythromycin, kanamycin, and chloramphenicol (with the use of a different antibiotic marker on the DMT plasmid) are good choices as selection agents in H. modesticaldum, because of their commercial availability, moderate MIC, lack of spontaneous resistance, and thermostability at 50°C.
The ability to genetically transform H. modesticaldum opens the door to a much greater understanding of this family of organisms by the use of modern molecular genetic techniques. Investigators may now introduce genes using a shuttle vector and reisolate the plasmid from transformed heliobacteria afterwards. We have also shown that heterologous gene expression is possible, since the erythromycin gene expressed from these plasmids conferred antibiotic resistance to H. modesticaldum.
To isolate intact plasmid DNA from H. modesticaldum, steps must be taken to inactivate DNase enzymes, some of which may be present in the cell wall, prior to cell lysis. We have accomplished this by adding lysozyme, which hydrolyzes the glycosidic bonds of peptidoglycans, to degrade the cell wall while simultaneously using diethyl pyrocarbonate (DEPC) and EDTA to inactivate the DNase(s). This treatment resulted in plasmid DNA of much greater quality and quantity, consistent with the hypothesis of degradation by heliobacterial DNase(s), which may have been in the cell wall. It has been shown that some organisms in aquatic environments can utilize DNA as a primary source of phosphorus (28). The activity of the postulated cell wall-associated DNase may aid H. modesticaldum in utilizing DNA as a nutrient source in its environment in the soil.
Our results also indicate a strict requirement for either PABA or folate. The genome of H. modesticaldum lacks the genes pabA, pabB, and pabC, which encode the enzymes necessary for PABA biosynthesis (29). Since folate can replace PABA in the defined mPMS medium, the only use for PABA in H. modesticaldum must be to make folate, which is used in a variety of cellular contexts for 1-C metabolism. The strict requirement for biotin was previously reported (30). The reason for this is not entirely clear, as the organism appears to contain most of the genes of the biosynthesis pathway, except for the one encoding 8-amino-7-oxononanoate synthase (9).
There is no strict requirement for vitamin B12, although it has often been included in the PYE medium used for heliobacteria. Since the enzyme BchE, which is absolutely necessary for the anaerobic biosynthesis of (bacterio)chlorophyll, contains a cobalamin prosthetic group derived from vitamin B12, an initial inability to natively biosynthesize vitamin B12 is the likely reason for the initial lack of (bacterio)chlorophyll in H. modesticaldum cells subcultured into medium lacking vitamin B12. Apparently, native biosynthesis of B12, and then bacteriochlorophyll g, recovers after a period of time, as evidenced by the eventual return of the phototrophic phenotype.
The requirement for CO2 when growing H. modesticaldum on solid medium was unexpected and was discovered fortuitously. The failure of heliobacterial cells to form colonies on agar plates without CO2 in the atmosphere indicates that CO2 may be required for growth of heliobacteria. It is possible that this requirement was not noted previously, as 5 to 20% CO2 is a standard component in the fill gas of many microbiology glove boxes. Also, in the sealed serum vials in which liquid cultures are grown or when embedded in vials of Gelrite, CO2 generated by pyruvate fermentation may be trapped in the medium and allow cells to grow. In contrast, when individual cells are growing on a solid surface, CO2 produced by the cells would diffuse into the atmosphere. Consistent with this, we found that the addition of bicarbonate to agar plates would allow heliobacteria to form colonies when grown under an atmosphere lacking CO2. We thus recommend that CO2 be included standardly in the anaerobic chamber atmosphere when working with H. modesticaldum.
Based on its genome (9) and previous carbon pathway analysis (31, 32), H. modesticaldum is unable to assimilate CO2 through a true autotrophic CO2 fixation pathway, as it lacks key enzymes in all known pathways. However, it has been shown that this organism does take up CO2 via anaplerotic reactions. The enzyme phosphoenolpyruvate carboxykinase (PEPCK) was identified in H. modesticaldum, allowing for anaplerotic assimilation of CO2 to generate oxaloacetate directly from phosphoenolpyruvate, which can be generated from pyruvate by pyruvate phosphate dikinase (32). Even with PEPCK present, net carbon flow is thought to lead to the net production of CO2, fueled mostly by pyruvate oxidation and oxidative production of α-ketoglutarate (3). Thus, the reaction catalyzed by PEPCK (PEP + ADP + CO2 ⇆ OAA + ATP) is likely one of the key reactions for central carbon metabolism in heliobacteria that requires CO2 and may explain the need for CO2 during growth.
The knowledge of how to transform H. modesticadum, to express heterologous genes, to select transformants on the surface of solid medium, and to grow it in a defined medium will greatly aid the study of this family of organisms. In subsequent work, we have used these techniques to perform gene editing in H. modesticaldum, for which it was necessary to analyze and utilize the endogenous CRISPR system of the organism as a negative selection tool. Analysis of the CRISPR-Cas loci in this organism, their use for gene editing, and characterization of the mutant thus produced are the subject of a second paper.
MATERIALS AND METHODS
Strains and growth conditions.All experiments were performed in a vinyl anaerobic chamber (Coy Laboratory Products, Grass Lake, MI). Initially, the atmosphere in the chamber contained 2 to 3% H2, with the balance nitrogen. In later experiments with agar plates, the atmosphere included 20% CO2. Cultures were illuminated at 790 nm using light-emitting diodes (LEDs) (Marubeni America, New York, NY) at a flux of ∼30 μmol photons m−2 s−1 at 50°C in prewarmed modified pyruvate-yeast extract (mPYE) or modified pyruvate minimal salts (mPMS) medium. Growth of H. modesticaldum was monitored by the optical density at 735 or 900 nm to avoid wavelengths at which photosynthetic pigments absorb light.
PYE medium (8) was routinely used for growth with the following modifications: chelated FeSO4 was used instead of chelated FeCl3 at the same final concentration (20 μM), and Na2SeO3 was added to a final concentration of 12.5 nM. For initial experiments with solid medium, 100 μl of cell suspension was added to 5 ml of prewarmed 2× mPYE medium supplemented with 0.05% MgCl2 (wt/vol) in a clear, glass, screw-cap vial. An equal volume of 1.3% (wt/vol) Gelrite (Research Products International G35020) at 65°C was added and immediately mixed before the mixture was allowed to harden, which typically took only 20 min. Afterwards, the vial was inverted and placed inside an incubator at 50°C under illumination. To make plates, molten 1.5% (wt/vol) agar in mPYE medium was poured on 100-mm-diameter petri dishes and allowed to harden in the anaerobic chamber, where they were stored at room temperature until use. After heliobacterial cells were spread on the surface of agar plates, the plates were placed in clear sealable bags and incubated at 50°C under illumination.
PMS medium (17) with the additions of 12.5 nM Na2SeO3, 20 μM FeSO4, and 1 mM Na2S2O3 (mPMS) was used for testing the vitamin requirements for H. modesticaldum. Initially, biotin and vitamin B12 were omitted. This medium was supplemented with vitamins as follows: either Wolfe’s vitamin mixture (ATCC MD-VS, 10 ml) (33) or individual components as described in Table 1 (folate, Sigma, F7876, 20 μg/liter; pyridoxine, Sigma, P9755, 100 μg/liter; riboflavin, Sigma, R9504, 50 μg/liter; biotin, Acros Organics, 230095000, 15 μg/liter; thiamine, Acros Organics, 14899-0100, 50 μg/liter; nicotinate, Sigma, N4126, 50 μg/liter; pantothenate, Sigma, P2250, 50 μg/liter; vitamin B12, Sigma, V2876, 20 μg/liter; p-aminobenzoate, Sigma, A9878, 50 μg/liter; lipoate, Sigma, T5625, 50 μg/liter). Growth was inspected visually after three subcultures with 100-fold dilution, followed by 48 h of growth.
Plasmids and E. coli strains and their sources are listed in Table 5. E. coli cultures were grown in mPYE or Luria-Bertani (LB) broth (BD Difco 244610) supplemented with 100 μg/ml erythromycin and/or 15 μg/ml chloramphenicol, as appropriate, or on LB plates (1.5% agar) supplemented with 300 μg/ml erythromycin and/or 30 μg/ml chloramphenicol, as appropriate.
Strains and plasmids used in this study
Determination of MICs of antibiotics.MICs of antibiotics were determined by preparing plates in a range of concentrations of each antibiotic, as follows (all in μg/ml): ampicillin, 10, 1, 0.1, 0.01, and 0.001 (Fisher BP1760); tetracycline, 0.3, 0.2, 0.15, 0.1, and 0.075 (Sigma 87128); erythromycin, 5, 4, 3, 2, 1, 0.4, 0.3, 0.2, and 0.1 (Sigma E6376); thiamphenicol, 16, 12, 8, 4, and 2 (Alfa Aesar J63675); apramycin, 32, 24, 16, 10, 8, and 4 (Goldbio A-600); spectinomycin, 15, 10, 5, 4, 3, 2, and 1 (Sigma S4014); kanamycin, 30, 25, 20, 15, 10, and 5 (Goldbio K-120); chloramphenicol, 64, 48, 32, 16, 13, 10, 8, 5, and 2 (Acros 227920250); streptomycin, 400, 350, 300, 250, 200, and 100 (Research Products International S62000). The MIC was determined on the basis of colony formation after spreading 100 μl of a 2-day liquid culture of H. modesticaldum on plates in triplicate, followed by incubation for 5 days at 50°C under illumination. The lowest concentration that prevented formation of any colonies is reported as the MIC.
Construction of the DMT helper vector pPB347.Putative DMT genes were chosen on the basis of REBASE (19) analysis (see Results). Genomic DNA was isolated from H. modesticaldum using a modification (see below) of the Wizard Genomic DNA kit (Promega, Madison, WI) and used as a template in PCR using Q5 DNA polymerase (New England Biolab, Ipswich, MA). PCR primers (IDT, Coralville, IA) were designed to amplify the genes from start to stop codon and including a Shine-Dalgarno sequence for E. coli upstream of the start codon and a unique restriction enzyme site on each end of the PCR product (Table 6). The PCR products were digested and ligated into the pBAD33 vector (20) by using restriction digestion and ligation with enzymes from NEB. All plasmids were verified by Sanger sequencing. For one of the genes, HM1_3004, PCR amplification was unsuccessful, even after designing multiple sets of primers. HM1_3004 was therefore synthesized with flanking restriction sites and ordered as a G-Block through IDT with the following sequence (uppercase sequence is HM1_3004, which is identical to the genome sequence, underlined sequences indicate restriction enzymes, and double-underlined sequences indicate the ribosome-binding site for E. coli): ggtacccatatgtaacaggaggaattaaccATGGAGTTCTCAAAAGTTCAAAAGGAAATATTCAATCCGATCGCTGAAAATGTAAAAAAGCTGGCCCAACTCTTTCCTGCTGCCGTGAAAGACGGGCAAGTGGACTTGGAAGCCTTGAAAGCGGAGCTGGGACAATTTGAATCGGTCAGCGAAAAAATGTCGGAGCGCTATGAACTCGGTTGGGTTGGCAAGGAAGAAGCGAAGAAGCTTGCCAACCAGGACGTGGTCGGCCGAACCTTGAAATATGTGCCAGAGGAAAGCAAACACCCTGAAACGACAGAGAACCTGTATATTGAAGGGGACAACTTAGAGGTTTTGAAGCTACTCCGGAACAGTTATTACAATAAGGTGAAGATGATTTACATTGATCCTCCGTACAACACGGGCAATGATTTTATCTACAAAGACAACTTTGCGATGAATCAGCGTGAAAACAGTGCATTGGAAGGCGAGATAGATGAAATGGGAGAGCGTCTGATTGTTAATCAAAAGAGTAACGGGCGGTATCATTCCAACTGGTTGTCGATGATGTATCCTCGGTTGAAAGTTGCCAAGGATTTGCTAAAAGAAGATGGTGTGATCTTTATTAGCATTGATGATAATGAACACTCAAACTTAAAATTATTATGTGACGAGGTATTTGGTTCTAATAGCTTTATTGGAGACATTGTTTGGCGATCAAGTGACAATAGTAACAATAATGCGTTGACGTTTTCGGAGGATCATAATTATATTTTGGTATACGCAAAAAGTCCGGATTGGAAACCTAATTTCCTTAATAATGATAGCAAAAGGCAACACTTTAAAAACCCAGATAATGATCCTCGAGGTCCATGGTTTGATGGAAATCCCGTAAATAATCCAGGATTAAGACCAAATCTCCAATTTGATATAATAACGCCAAGTGGGAAACTTATTAAACATCCACCAAACGGATGGAGATGGTCCAAAGAAACTATTGAAGAAAAATTAAAGACAGGTGAACTGCGCTTCTCTGAAGATGAAACACGCCTTATAAGAAGGACTTATTTATATGAAATGAAAGGGTTACCGCCTTCAAGCTTGTGTATTGATTTGGAGATAACTGGGCACACAAGAAGAGCGAAATATGAATTAAAAAAGTTGTTTCCAGAAGTACCTGTAACAAGTTTATTTAGCACTCCAAAGCCAACGTTATTATTGAAATATATACTAACAATAGCTTCGGATAATAATGCTATTGTTTTGGATTTTTTTTCAGGTTCAGCAACATCAGCGGATGCTGTTATGCAGCTAAATGCTGAAGATGGGGGTAAAAGAAAGTTTATTATGGTTCAATTACCTGAAGTTTGTGAGAGTGGAGATTATGGAAATGCCGTTAAGTTAAAGAACATCTGCGAAATCGGCAAAGAACGCATCCGCAGAGCCGGAGACAAGATCATCGAAGAAAACAAAGACAAGAAGGGCATCGAAAACCTCGACATCGGCTTCAAGGTCTTCAAAGTGGCCGACACCAATATCCGTTGGTTCAGTGAAGCGATGAAATCCGATCAAATGACACTTGATGAAAGCGCCATGACCGATAAAGACAGGTTGGACTTCACCCCCGGCTTTACCGATCTCGATGTGGTCTATGAGATCCTGCTCCGTCACCGGGATATCCCCTTATCGTCCAAGGTGGAGAAAGTAGCTTCTATTGGAGAGAGAACCTATATCTTCGCCGACACCGTTCTCGTTTGCTTGGAAGAAAACGTAACGGAAAGCATGATTGATAAAATCGCCGCCATCGAGCCCATGCCCACCAAGATCATCTTCCGTGACAGCGCCTTCGGCGCCGACATCAGCTTGAAAACGAACAGCATGCTCCGATTGGAAGCGCAGATGAAGAAAAACAGCGGCCTGAAGAAGAAAGCCTATCGCGTCGAGTTCATATAActgcaga.
Oligonucleotide primers used in this study
H. modesticaldum transformation via conjugation.Plasmids were introduced into H. modesticaldum by conjugation with E. coli strain S17-1. Chemically competent E. coli S17-1 was transformed with the DMT helper plasmid, followed by the shuttle vector, or cotransformed with both plasmids. S17-1 transformant colonies were grown overnight at 37°C in LB (or mPYE) liquid medium supplemented with appropriate combinations of erythromycin (100 μg/ml) and chloramphenicol (15 μg/ml) for plasmid maintenance. One milliliter of cells was harvested by centrifugation at room temperature at 5,000 × g for 5 min. The supernatant was removed, and cells were brought into the anaerobic chamber, washed in 1 ml of phosphate-buffered saline (PBS), and centrifuged as described above. Cell pellets were resuspended in 200 μl of a 48-h H. modesticaldum culture (optical density at 900 nm [OD900], ∼0.8) that had been cooled to 37°C. The mixture of cells was centrifuged again, and 190 μl of supernatant was removed. The remaining supernatant was spotted and dried onto a 2% (wt/vol) agar-mPYE plug (5 ml) occupying a well (2.26-cm diameter) of a 12-well plate. The 12-well plate was inverted and incubated on top of a dry bath set to 37°C (or 42°C) overnight (or for 3 h) under 790-nm LED lighting. After the allotted time, cells were scraped from the agar plug and resuspended in 200 μl of PBS before being added to Gelrite vials or spread on selective mPYE-agar plates (supplemented with 10 μg/ml erythromycin) using 5-mm glass beads. The plates and beads were prewarmed to 50°C prior to plating in order to prevent growth of the E. coli cells.
Plasmid preparation from H. modesticaldum.Plasmids were isolated from H. modesticaldum by a modification of the Monarch plasmid miniprep kit (New England Biolabs, Ipswich, MA). A volume of 1.5 ml of overnight cell culture was harvested by centrifugation at 13,000 × g for 1 min. The supernatant was removed, and the cell pellet was resuspended in 480 μl of 50 mM EDTA (pH 8). Lysozyme and diethyl pyrocarbonate (DEPC) were added to the suspension at 0.6 mg/ml and 0.4%, respectively. The suspension was incubated at 37°C for 10 to 30 min, after which the cells were pelleted again, and the supernatant was removed. After this preliminary treatment, the cell pellet was removed from the anaerobic chamber and the plasmid preparation was performed as specified in the kit.
ACKNOWLEDGMENT
This work was funded by the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, U.S. Department of Energy through grant DE-SC0010575 to K.E.R.
FOOTNOTES
- Received 6 June 2019.
- Accepted 21 July 2019.
- Accepted manuscript posted online 2 August 2019.
- Copyright © 2019 American Society for Microbiology.
