The role of Rnf in ion gradient formation in Desulfovibrio alaskensis

Growth of cultures

Desulfovibrio alaskensis G20 and rnf mutants were routinely cultured anaerobically (N₂-CO₂, 80:20) in basal medium with lactate-sulfate (50 mM each) as the electron donor and electron acceptor, respectively and containing 0.1% yeast extract (Krumholz et al., 2015). The medium was prepared under a headspace of N₂∕CO₂(80/20, vol/vol) with sodium bicarbonate (3.5 g/L) added as a buffer. Before inoculation, the medium was reduced with 0.025% each cysteine and sulfide. Kanamycin (1050 µg/mL) was added to the mutant cultures. Inoculum consisted of 0.2 ml transferred into 10 ml of media or the same ratio for large volume cultures. Cultures were grown in the incubator at 37 °C and growth of triplicate cultures was measured by optical density at 600 nm. The stock cultures were frozen in 20% glycerol at −20°. Growth of syntrophic cocultures was as previously described (Krumholz et al., 2015).

Growth was tested with a variety of other electron donors and acceptors: lactate (50 mM), pyruvate (25 mM), ethanol (10 mM) and formate (50 mM). Sulfate was used at 50 mM with lactate and 25 mM with other electron donors and sulfite was added at 10 mM.

The protonophore 3,3,4,5-tetrachlorosalicylanilide (TCS, 5 µM or 20 µM), or the sodium-specific ionophore N,N,N′,N′-Tetracyclohexyl-1,2-phenylenedioxydiacetamide (ETH2120, 20 µM) were added to media to test their effects on growth.

Transcriptional analysis of rnf genes

Total RNA was extracted from the parent strain, rnfA, and rnfD mutants using the Qiagen RNeasy kit as previously described (Krumholz et al., 2015). Purity and adequate yield was confirmed with a diode-array spectrophotometer. First-strand cDNA synthesis was performed using the Fermentas Revertaid kit with gene-specific primers covering the entirety of the rnf mutants’ operons. Primers were designed to span the gaps between each gene in the operon (Table 1). PCR was performed with the parent strain and mutants with each primer set followed by agarose gel analysis to determine whether all genes in the operon were expressed and whether the genes were on the same transcript.

Washed cell experiments

Washed cells were used to measure sulfide production under non-growth conditions. Briefly, 100 ml lactate-sulfate culture grown to mid log phase was harvested by centrifugation at 5,300 × g for 15 min, washed twice in buffer containing 50 mM MOPS (pH 7.2), 5 mM MgCl₂ and resuspended in the same buffer. All buffers were flushed with N₂ for 30 min. Assays were carried out in serum tubes containing 2 ml of buffer-cell mixture incubated at 37 °C on a shaker at 100 rpm. The assay mixture contained the washing buffer, 5 mM sodium sulfate and either 50 mM lactate, 50 mM sodium formate (N₂ headspace) or H₂ in the headspace. Between 50 and 200 µg of cell protein was added to each tube. Tubes were sacrificed by addition of 2 ml 10% Zinc Acetate at 60 min intervals for sulfide analysis. Sulfide was determined using the methylene blue assay (Cline, 1969).

Measurement of proton motive force

Triplicate cultures of D. alaskensis parent strain and rnfA mutant were grown in lactate–sulfate basal medium until mid exponential phase (OD₆₀₀ of 0.5) in a 1 liter volume and 30 ml aliquots of cells were dispensed into 100 ml serum bottles under N₂∕CO₂ (80/20, vol/vol). The basal medium contains 5 mM K⁺, mainly added as potassium phosphate salts. Protein concentration was determined using the Bicinchoninic Acid Assay (Pierce BCA Protein Assay Kit).

Transmembrane electrical potential (ΔΨ) was measured as previously described (Tremblay et al., 2013). Briefly, cells were incubated with [³H]tetraphenylphosphonium bromide ([³H]TPP⁺) (ARC, 0.1 µCi/ml) for 15 min at 37 °C (Kashket & Barker, 1977; Shirvan, Schuldiner & Rottem, 1989). [³H]TPP⁺ was dissolved in media and filtered through 0.45 µm filters prior to adding to cells. Controls were incubated with nigericin (ACROS ORGANICS, 20 µM dissolved in ethanol at 200×) and valinomycin (ACROS), 20 µM dissolved in DMSO at 200× for 15 min to eliminate the ΔΨ. The combination of nigericin (proton/K⁺ antiporter) and valinomycin (K⁺ uncoupler) will dissipate the proton gradient (Kessler et al., 1977). Cells were then separated from the medium using silicone oil as described below and [³H]TPP⁺ was determined in the liquid scintillation counter (Packard TriCarb 2100TR). The uptake of [³H]TPP⁺ was corrected for extracellular contamination using the ratio of the intracellular to extracellular volume as described below. Non-specific binding of [³H]TPP⁺ was corrected by subtracting the cell associated [³H]TPP⁺ in the valinomycin/nigericin treatment from that in the untreated cells. The ΔΨ was calculated with the simplified Nernst equation (−2.3[RT∕F] × log[(concentration in)∕(concentration out)]).

The ΔpH was measured by testing the distribution of [¹⁴C ]benzoate (ARC, 0.4 µCi/ml) across the cell membrane after 15 min incubation at 37 °C with cells (Kashket & Barker, 1977; Shirvan, Schuldiner & Rottem, 1989). Cells were then separated from the supernatant as described below and [¹⁴C ]benzoate was measured. [¹⁴C ]benzoate uptake was corrected for extracellular contamination using the ratio of intracellular to extracellular volume described below. Tetrachlorosalicylanilide (TCS, 20 µM) was added to controls to eliminate the ΔpH and was used to calculate internal pH in the absence of a ΔpH. The external pH was measured, [¹⁴C ]benzoate uptake was quantified and intracellular pH was calculated with the Henderson–Hasselbach equation (Rottenberg, 1979). The proton motive force (PMF) was calculated using the equation: PMF = ΔΨ − zΔpH. At 37 °C , z equals to 61.48.

Measurement of intracellular volume/total volume of cell pellet

³H₂O (1 µCi/ml) was used to measure the total pellet volume while [¹⁴C ]taurine (PerkinElmer, 0.5 µCi/ml) was used to measure extracellular volume and was inferred to penetrate up to the plasma membrane (Kashket & Barker, 1977). The intracellular water volume was estimated from total pellet water volume minus extracellular water volume by measuring the difference between the distribution of ³H₂O and [³H]taurine after incubating cells for 15 min at 37 °C with each isotope.

Separation of cells from the supernatant

Following the 15 min incubation with the isotope, three 10 ml aliquots of the total 30 ml solution were put into three 15 ml Falcon tubes with 3 ml of a mixture of silicone oils (25% Fluid 510, 50 centistokes, and 75% Fluid 550, 115 centistokes; Dow-Corning Corp, Serva. vol/vol). Cells were then centrifuged at 5,300 × g at 4 °C for 10 min. The aqueous layer and the silicone oil was removed with a Pasteur pipette connected to a vacuum line. The bottom of the falcon tube containing the cell pellet was cut off and moved into the scintillation vial and the cell pellet was re-suspended with 200 µl distilled-water. Liquid scintillation cocktail (Ultima Gold; PerkinElmer, Waltham, MA, USA) was added prior to scintillation counting.

Statistical analysis

Statistical analysis used either the t-test or a one-way analysis of variance (ANOVA) with the Tukey’s Range test.

Mutants in the rnf operon

In a previous study, a Tn10 mutant library of D. alaskensis G20 was screened for the ability to grow on butyrate in coculture with Syntrophomonas wolfei (Krumholz et al., 2015). Seventeen mutants were obtained that were deficient in syntrophic growth. Of these, two mutants had the transposon insertions within a putative rnf operon. This operon is composed of 8 genes located on the genome as shown in Table 3 and genes appear to be co-transcribed. They are separated from the nearest upstream transcribed region by 185 bps and from the nearest downstream transcribed region by 342 bps. The two syntrophy mutants had transposon insertions within rnfA and rnfD as determined using Arbitrary PCR (Krumholz et al., 2015). Both genes encode integral membrane proteins likely located in the cytoplasmic membrane. RnfD has been previously shown to bind a flavin (Biegel et al., 2011).

Transcriptional analysis of the rnf operon

The complete transcriptional analysis for strain G20 under lactate-sulfate, H₂-sulfate and under syntrophic growth conditions has been previously published (Krumholz et al., 2015). Here we present a summary of normalized values for transcription of each subunit within the rnf operon (Table 3). Transcription of all genes in the operon including dchA are similar under each condition and are affected similarly by growth condition. The most highly expressed genes are the first two in the operon, the gene for the cytochrome c family protein and rnfC, both of which have 1.5–2 fold higher levels of expression than genes for the other subunits. Expression was also enhanced approximately 2-fold when cultures were grown on H₂, relative to lactate, indicating the importance of this protein for H₂ dependent growth. Syntrophic cultures had enhanced expression of all eight genes compared to lactate/sulfate growth with similar levels of expression in the S. aciditrophicus coculture to H₂-grown cells and higher levels of expression observed in the S. wolfei coculture, indicating the importance of this protein complex under syntrophic conditions. Similar effects on transcription provide evidence that all eight genes are located in one operon.

Expression analysis of rnf operons in mutants

To determine whether the transposon insertion eliminated downstream expression of genes in the rnf operon, gap analysis was performed where all intergene regions were amplified except the region between rnfB and rnfA using cDNA prepared using mRNA. All of the intergene regions were amplified for the parent and the two mutants (Fig. S1) confirming that all 8 genes form an operon and indicating that the mutants likely have intact transcripts of the interrupted operon. In this operon, the insertions are likely only affecting the one interrupted gene.

We carried out RT-PCR analysis of the terminal gene in the operon, Dde_ 587 (rnfF) as well as the two genes that followed that operon, Dde_ 589 (uspA) and Dde_ 590 (cls) to be certain that downstream expression was not affected by the insertions. Expression of the terminal gene in the operon was impacted in the mutants with a 4 fold decrease in expression in the rnfA mutant and a twofold increase in the rnfD mutant (Table 4). The following two genes were minimally impacted indicating that the insertions present in the mutants likely do not affect downstream expression.

Growth and sulfate reduction by rnf mutants

The two mutants exhibited similar growth rates to that of the parent strain when lactate was the electron donor with sulfate (Fig. 1A) or sulfite as electron acceptor (Fig. 1B). When pyruvate was the donor, growth yields were reduced for the mutants (Fig. 1C). We previously reported that rnf mutants grew poorly (rnfD) or not at all (rnfA) on H₂ (Krumholz et al., 2015) and a recent study has reported that rnf mutants grow poorly or not at all with sulfate as the electron acceptor on malate, fumarate, ethanol, hydrogen, and formate, but growth is not affected in lactate and pyruvate (Price et al., 2014). Here, it is shown that the rnfD mutant has a long lag phase with formate as the donor, but eventually grew to a similar OD to the parent strain (Fig. 1D). The rnfA mutant did not grow on formate or H₂ clearly indicating the involvement of Rnf in formate/H₂-dependent growth. We also confirmed that the rnfA mutant will not grow on ethanol (Fig. S3).

Sulfide production by washed cells

Washed cells were used to determine if the role for rnf was linked to biosynthesis, as mutants were able to grow better on more complex carbon compounds. Incubations of washed cells of the parent strain and the rnfA mutant produced 0.57 and 0.17 µmol sulfide.hr⁻¹.mg⁻¹ protein, respectively, with lactate as electron donor. Addition of either 5 µM TCS to the washed cells of the parent strain prevented 95% of the sulfide production and addition of 20 µM to washed cells inhibited sulfide production by 100%, suggesting TCS directly disrupts the proton gradient needed for respiration. With H₂ as the electron donor, parent strain cells produced 0.30 µmol sulfide.hr⁻¹.mg⁻¹ protein whereas the rnfA mutant did not have any detectable activity. To further test whether the lack of growth on H₂ was due to the mutant’s inability to biosynthesize carbon intermediates, we attempted to grow the mutant and the parent strain with H₂ as the electron donor with 0.1% Casamino acids added to the media. The mutant would still not grow (Fig. S4), providing further evidence that the role for RNF was not directly linked to biosynthesis.

Growth curves with ionophores

The sodium ion ionophore ETH 2120, tested at 20 µM, did not have a significant influence on the growth of the parent strain or the rnf mutants with lactate as the electron donor whether sulfate or sulfite were present as electron acceptors (Figs. S5 and S6). Interestingly, the protonophore, TCS , at 5 µM differentially inhibited cultures (Fig. 2). TCS partially inhibited the growth of the parent strain and completely inhibited the growth of rnfmutants on lactate-sulfate (Fig. 2A and Fig. S5). Growth on lactate sulfite was then tested and 5 µM TCS was shown to have a smaller but still significant inhibitory effect on growth in the parent strain (Fig. 2B) and again completely inhibited the rnfA mutant (Fig. 2B and Fig. S6). A higher level of TCS was then tested and it was shown that 20 µM TCS almost completely inhibited the growth of the parent strain on lactate-sulfite (Fig. 2B) and again completely inhibited the mutants (Fig. S7).

Measurement of ΔΨ, ΔpH and PMF

To determine whether the Rnf complex is involved in the formation of a proton gradient, the transmembrane potential (ΔΨ), ΔpH and proton motive force (PMF) were measured in cells of the G20 parent strain and the rnfA mutant growing on lactate-sulfate. The ΔΨ and ΔpH were both reduced in the mutant (Table 5), which led to a significantly lower PMF in the rnfA mutant relative to the G20 parent strain. The lower Δ pH and PMF in the rnfA mutant shows that the Rnf complex clearly had an impact on the energy conservation system by affecting the proton gradient.