Nitrate-Responsive Suppression of Dimethyl Sulfoxide Respiration in a Facultative Anaerobic Haloarchaeon, Haloferax volcanii

ABSTRACT

Haloferax volcanii is a facultative anaerobic haloarchaeon that can grow using nitrate or dimethyl sulfoxide (DMSO) as a respiratory substrate under anaerobic conditions. Comparative transcriptome analysis of denitrifying and aerobic cells of H. volcanii indicated extensive changes in gene expression involving the activation of denitrification, suppression of DMSO respiration, and conversion of the heme biosynthetic pathway under denitrifying conditions. The anaerobic growth of H. volcanii by DMSO respiration was inhibited at nitrate concentrations of <1 mM, whereas nitrate-responsive growth inhibition was not observed in the ΔnarO mutant. A reporter assay demonstrated that the transcription of the dms operon was suppressed by nitrate. In contrast, the anaerobic growth of the ΔdmsR mutant by denitrification was little affected by the addition of DMSO. NarO has been identified as an activator of denitrification-related genes in response to anaerobic conditions, and here, we found that NarO is also involved in nitrate-responsive suppression of the dms operon. Nitrate-responsive suppression of DMSO respiration is known in several bacteria such as Escherichia coli and photosynthetic Rhodobacter species. This is the first report to show that a regulatory mechanism that suppresses DMSO respiration in response to nitrate exists not only in bacteria but also in haloarchaea.

IMPORTANCE Haloferax volcanii can grow anaerobically by denitrification (nitrate respiration) or DMSO respiration. In facultative anaerobic bacteria that can grow by both nitrate respiration and DMSO respiration, nitrate respiration is preferentially induced when both nitrate and DMSO are available as the respiratory substrates. The results of transcriptome analysis, growth phenotyping, and reporter assays indicated that DMSO respiration is suppressed in response to nitrate in H. volcanii. The haloarchaeon-specific regulator NarO, which activates denitrification under anaerobic conditions, is suggested to be involved in the nitrate-responsive suppression of DMSO respiration.

INTRODUCTION

Haloferax volcanii, an extreme haloarchaeon isolated from the Dead Sea, has been evaluated as a model organism of this type of extremophile (1, 2). H. volcanii is a facultative anaerobic microorganism that can grow by denitrification (nitrate respiration), dimethyl sulfoxide (DMSO) respiration, or fumarate respiration under anaerobic conditions (1–5). Similar to bacterial denitrification, the denitrifying process in archaea proceeds by successive reactions catalyzed by the redox enzymes nitrate reductase, nitrite reductase, NO reductase, and N₂O reductase (6–8). However, because of the lack of the N₂O reductase gene, the end product of denitrification by H. volcanii is not N₂ but N₂O (9). The DMSO reductase genes dmsABCD, which are homologous to the bacterial counterparts present in Escherichia coli, are also found in H. volcanii (5, 10, 11).

Archaea usually possess characteristics similar to those of bacteria in terms of metabolism, whereas their genetic processes, including replication, transcription, and translation, have been considered to be closer to those of eukaryotes than those of bacteria (12, 13). In bacteria, the regulation machinery of anaerobic respiration, especially nitrate respiration and denitrification, has been well investigated (reviewed in reference 14). The transcription of the nar operon encoding the dissimilatory nitrate reductases NarGHI is controlled by the oxygen sensor Fnr in E. coli (15). Fnr possesses an iron-sulfur cluster as the redox sensor, an allosteric site for interaction with RNA polymerase (RNAP), and a helix-turn-helix (HTH)-type DNA binding motif in the molecule (16–18). Under anaerobic conditions, an activated Fnr dimer binds to the Fnr box sequence (consensus, TTGAT-NNNN-ATCAA) in the promoter of the nar operon and then forms a transcription complex with sigma factor and RNAP (19). In combination with Fnr, the two-component regulator NarXL and its functional paralogue NarPQ regulate the nitrate/nitrite-responsive activation of the nar operon (20, 21). NarX is a transmembranous histidine kinase whose sensor motif has been identified in the periplasmic region. Phosphorylated NarL recognizes a heptameric DNA sequence (consensus, TACYYMT) to mediate the transcription activation of the nar operon (21). In the root nodule diazotrophs, the two-component oxygen-responsive regulator FixLJ is the starting point of a regulatory cascade leading to denitrification and nitrogen fixation (22, 23). FixL possesses a protoheme as an oxygen-sensing molecule combined with the PAS motif (24). Under anaerobic conditions, the heme molecule turns into a high-spin ferrous form, causing the self-phosphorylation of FixL and the subsequent activation of FixJ (25, 26).

The structural characteristics of an archaeal RNAP have been reported to be similar to those of a eukaryotic RNAPII (27). In archaea, not an endogenous sigma factor but a TATA binding protein (TBP) and transcription factor B are involved in the recognition of the gene promoter by RNAP (28–30). Because these basic transcription factors are also homologous to eukaryotic TBP and transcription factor IIB, the transcription machinery in archaea is considered to be closer to that in eukaryotes than that in bacteria (12). Archaeal denitrification therefore might be regulated by a genetic mechanism that is different from that in bacteria. In fact, the Fnr family transcription factor is not present in H. volcanii genomes.

It has been demonstrated that the HTH-type regulator NarO (HVO_B0159), which is encoded adjacent to the nar operon (HVO_B0160 to -0166) in the megaplasmid pHV3 of H. volcanii, is involved in oxygen-dependent transcription regulation of the denitrifying genes (4). NarO is widespread in the haloarchaea; narO and its homologous genes are found to be present in a large number of the haloarchaea whose genomic sequences are available, while it does not exist in other organisms at all. Like bacterial Fnr, NarO may act as a regulator that utilizes four conserved cysteine residues as the functional core for oxygen sensing (4, 16).

In this study, a comparative transcriptome analysis of denitrifying and aerobic cells of H. volcanii was performed. The results revealed extensive changes in gene expression involving the activation of denitrifying genes and the suppression of the DMSO reductase gene under denitrifying conditions. Growth phenotyping and a reporter assay demonstrated that DMSO respiration is negatively regulated by NarO in a nitrate-responsive manner. This is the first report to show that nitrate-responsive suppression of DMSO respiration exists not only in bacteria but also in haloarchaea.

RESULTS AND DISCUSSION

Comparative transcriptome analysis.

Transcriptome analyses of aerobic cells and denitrifying cells of H. volcanii were carried out. RNA sequencing (RNA-seq) generated 8,464,043 and 9,772,800 pairs of raw reads under aerobic and denitrifying conditions, respectively. The adapter-trimming and quality-filtering reads (7,752,271 and 8,650,842 pairs under aerobic and denitrifying conditions, respectively) were mapped to the genome sequence of H. volcanii DS2, an ancestral strain of H. volcanii H26. Read counts aligned to the protein-coding genes were calculated, and the resulting alignment rates of the reads under aerobic and denitrifying conditions were 37.7% and 52.7%, respectively. These low alignment rates are due to reads derived from the rRNA and noncoding RNA. As a result of filtering out lowly expressed genes (counts per million [CPM], <2) with little evidence for differential expression, we found the gene expression of 3,870 (96.2%) out of 4,021 protein-coding genes (except for 6 genes on the pHV2 plasmid because strain H26 lacks the pHV2 plasmid), which are listed in Table S1 in the supplemental material along with their annotated functions.

Comparing the gene expression levels of these 3,870 genes between aerobic and denitrifying cells of H. volcanii strain H26, 208 genes were upregulated (false discovery rate [FDR] of <0.05 and log₂ fold change [logFC] of >2) and 97 genes were downregulated (FDR of less than 0.05 and logFC of less than −2) in the denitrifying cells. Eighty-four of the 208 upregulated genes were present in the chromosome, accounting for 2.8% of the 2,969 total genes. On the three megaplasmids pHV1, -3, and -4 in strain H26, 66 of the 375 genes on pHV3 (17.6%), 52 of the 600 genes on pHV4 (8.7%), and 6 of the 77 genes on pHV1 (7.8%) were upregulated in the denitrifying cells. Two-sided Fisher’s exact test confirmed that the upregulated genes in the denitrifying cells were overrepresented in the genes on the megaplasmids, with a P value of <2.2e−16. The up- and downregulated genes are listed in Tables S2 and S3, respectively.

The annotated functions of the 208 upregulated genes and the 97 downregulated genes during denitrification are summarized in Table 1. Thirty-nine genes involved in the degradation or biosynthesis of carbohydrates, amino acids, lipids, and nucleotides and 9 genes encoding putative redox enzymes and proteins for oxidative phosphorylation were upregulated, confirming significant changes in the primary metabolic processes by the transition from aerobic growth to anaerobic growth by denitrification. In addition, 46 transport-related genes, most of which encoded ABC-type transporters, were upregulated in the denitrifying H. volcanii cells. A change in the gene expression of the regulatory DNA binding proteins, including the upregulation of the denitrification regulator NarO, was also observed. Ninety-three of the 208 upregulated genes and 62 of the 97 downregulated genes in the denitrifying cells encoded poorly or uncharacterized proteins or proteins of unknown function.

TABLE 1.

Up- and downregulated genes in denitrifying cells of H. volcanii strain H26

Functional category	No. of genes
	Upregulated	Downregulated	Nonregulated	Totala
Amino acid, carbohydrate, lipid, and nucleotide metabolisms	39	5	504	548
Energy metabolism	9	6	68	83
Cofactor and vitamin biosynthesis	4	1	111	116
Transporter	46	8	288	342
Other metabolic functions	1	6	131	138
Replication and repair, transcription, and translation	3	3	235	241
Transcription regulator and signal transducer	11	6	222	239
Other genetic function (integrase and transposase)	2	0	85	87
Unclassified, uncharacterized, or function unknown	93	62	1,921	2,076
Total	208	97	3,565	3,870

^aLowly expressed genes were excluded.

Denitrification-related genes.

There were 20 upregulated genes whose transcripts per kilobase per million (TPM) values are above 1,000 in the denitrifying cells, 11 of which were concentrated in a region of only 20 kb on the chromosome (Fig. 1A and Table S1). This gene cluster includes genes that are probably involved in denitrification, such as the nitrite reductase gene nirK (HVO_2141), the NO reductase gene norB (HVO_2147), and the copper-containing electron carrier halocyanin genes hcyF (HVO_2145) and hcyG (HVO_2150) (Fig. 1A). The two genes (HVO_2151 and HVO_2152) encoding HTH-type transcription regulators were upregulated under denitrifying conditions and may be involved in the induction of denitrification or anaerobic metabolism. Two upregulated genes, ahbD (HVO_2144) and thiC (HVO_2154), encode radical S-adenosylmethionine (SAM) enzymes that function under anaerobic conditions and are involved in heme and thiamine biosynthesis, respectively. In addition, two genes encoding the multicopper oxidase McoA (HVO_2153) and a zinc finger protein (HVO_2142) of unknown function and four hypothetical genes (HVO_2134, HVO_2137, HVO_2138, and HVO_2143) were also upregulated under denitrification conditions. A palindromic sequence (consensus, CGAA-NNNN-TTCG), which appears to be a recognition motif for the HTH-type regulators, was commonly present in the putative promoter sequences of the most upregulated genes in the cluster (marked by asterisks in Fig. 1A). The functions of these hypothetical or uncharacterized genes should be elucidated to better understand the response of H. volcanii cells to denitrification.

FIG 1.

Relative expression levels of genes involved in denitrification and DMSO respiration in denitrifying H. volcanii cells. Gene arrangements of the denitrifying gene cluster (A), the regulator gene narO and the nitrate reductase operon narBCGHDJ (B), and the regulator gene dmsR and the DMSO reductase operon dmsABCD (C) are indicated on the top of each panel. Functional genes for denitrification and DMSO respiration are shown in pale gray, and the regulator genes are in dark gray. The putative recognition motif (CGAA-NNNN-TTCG) for the HTH-type regulator is marked by asterisks in panels A and B. TPM values (left) in denitrifying cells and logFC values (right) of the genes are indicated by white bars and solid bold lines in the panels, respectively.

The nitrate reductase genes narBCGHDJ, which are carried on the pHV3 megaplasmid, were also strongly upregulated during denitrification, along with the regulator gene narO (HVO_B0159), located upstream of the nar operon (Fig. 1B). It was previously shown that both the consensus sequence and the HVO_B0159 gene encoding a haloarchaeon-specific HTH-type regulator, NarO, are essential for the transcription activation of the nirK gene under denitrifying conditions (4). Unlike H. volcanii, Haloarcula marismortui possesses genes encoding denitrifying enzymes, NarO/nitrate reductase (rrnAC1193/rrnAC1194 to -1208), the nitrite reductase NirK (rrnAC2853), the NO reductase NorB (rrnAC2272), and the N₂O reductase NosZ (rrnAC0389 to -0396), in chromosome I without forming a cluster (31–33). However, a consensus sequence is present in each gene promoter of H. marismortui, suggesting the presence of a regulatory mechanism for denitrification similar to that of H. volcanii (4).

Aerobic and DMSO respiration-related genes.

Cytochrome bd-type ubiquinol oxidase subunit I (HVO_0462) (logFC = −3.65) and subunit II (HVO_0461) (logFC = −2.43) were downregulated under denitrifying conditions (Table S2). In addition, the expression levels of two loci (HVO_0942 to -0946 and HVO_1644 to -1646), both of which encode cytochrome aa₃-type terminal enzymes for aerobic respiration, were also decreased in denitrifying cells of H. volcanii compared to those in aerobic cells. Cytochrome bd functions as a respiratory terminal enzyme under oxygen-limited conditions in various prokaryotic microorganisms, and it has been known to be mostly induced in E. coli cells cultivated when 1% oxygen is used as the gas phase of the culture (34, 35). The downregulation of not only the cytochrome aa₃ gene but also the cytochrome bd gene, which is responsible for oxygen respiration under microaerobic conditions, indicated that a high degree of anaerobicity was achieved in this cultivation experiment.

In addition to denitrification, H. volcanii can grow by anaerobic respiration using DMSO or succinate as an electron acceptor (1, 5). Interestingly, the TPM values of the dmsABCD (HVO_B0363 to -B0366) genes that encode DMSO reductase were significantly reduced in denitrifying cells of H. volcanii compared with those in aerobic cells (Fig. 1C). The expression levels of the dmsABCD genes in aerobic cells (TPM, 25.1 to 146.8) were similar to or higher than the median TPM value of 39.9. Qi et al. (5) reported that H. volcanii cells cultivated aerobically in DMSO-supplemented medium exhibit DMSO-reducing activity albeit weakly. The downregulation of the dms operon under denitrifying conditions suggests the presence of a nitrate-responsive regulatory system that suppresses the induction of DMSO respiration in H. volcanii.

Heme biosynthesis pathway.

Bacteria and archaea are known to perform heme synthesis by two different pathways, called the alternative heme biosynthetic (Ahb) pathway and the coproheme decarboxylase (ChdC) pathway (formerly the HemQ pathway), which are different from that in eukaryotes (36). The ahbD (HVO_2144) gene on the denitrification gene cluster, encoding the SAM-dependent enzyme, was also upregulated during denitrification (Fig. 1A). AhbD is the enzyme that catalyzes the decarboxylation of coproheme III, forming protoheme IX, a final step in the Ahb pathway for anaerobic heme biosynthesis (36, 37). Among the six genes involved in the Ahb pathway for the anaerobic biosynthesis of protoheme IX from uroporphyrin III as a starting compound, the precorrin-2 oxidase/ferrochelatase gene cysG (HVO_2312) and the siroheme decarboxylase gene ahbB (HVO_2313) were upregulated, in addition to the ahbD gene (Table S4). The expression level of the uroporphyrin III C-methyltransferase gene hemX (HVO_0077) also tended to be increased (logFC = +1.08). In contrast, the expression levels of the siroheme decarboxylase gene ahbA (HVO_2227) and the coproheme synthase gene ahbC (HVO_1121) were decreased during denitrification compared with those in aerobic cells. In bacteria, the regulatory mechanism of the Ahb pathway under anaerobic conditions remains unclear. The putative recognition motif for the HTH regulator described above is also present in the ahbD gene promoter, suggesting that transcription of the ahbD gene will probably be controlled by the haloarchaeon-specific regulator NarO. Under aerobic conditions, heme biosynthesis is considered to proceed by the ChdC pathway in H. volcanii (37). The chdC gene (HVO_1871) encoding H₂O₂-dependent coproheme III decarboxylase, catalyzing the final step of the ChdC pathway, was downregulated under denitrifying conditions.

Transporter-related genes.

During the transition from aerobic conditions to denitrification conditions, significant changes in gene expression also occurred in the transporter genes for sugars, amino acids, and metal cations (Table 1). Wanner and Soppa (38) prepared a number of H. volcanii mutants lacking denitrifying activity, and they found that three ABC-type transporter genes are probably essential for anaerobic growth by denitrification based on complementation experiments with these mutants. These are annotated as glucose transporters (HVO_B0314 to -B0316), molybdate-specific transporters (HVO_B0369 and -B0370), and anion-specific transporters (HVO_1886 to -1888). However, the TPM values of the putative glucose transporter and molybdate-specific transporter were extremely low in denitrifying cells, ranging from 0 to 2, and the upregulation of these genes was not observed in the present analysis. The TPM values for the anion-specific transporter genes were relatively high (91 for HVO_1886 and 161 for HVO_1887) in denitrifying cells, while significant upregulation of these genes was also not observed during denitrification.

Effect of nitrate on anaerobic growth by DMSO respiration.

H. volcanii strain H26 was cultivated anaerobically in Hv medium supplemented with 50 mM DMSO and various concentrations of nitrate (0 to 100 mM) to evaluate the effect of nitrate on anaerobic growth by DMSO respiration (Fig. 2A). After 5 days of cultivation, the optical density at 600 nm (OD₆₀₀) of the medium reached 0.40 in the absence of nitrate (Fig. 2C). In medium containing low nitrate concentrations ranging from 0 to 1 mM, the growth rate decreased with increasing nitrate concentrations. The growth rate was minimal in the presence of 1 mM nitrate, and the OD₆₀₀ of the medium reached only a level similar to that in anaerobic medium without the addition of DMSO as the respiratory substrate after 5 days of cultivation. The addition of nitrate, whose concentration ranged from 10 mM to 100 mM, enhanced the growth rate of the microbe with increasing nitrate concentrations. The experimental results indicate that concentrations of nitrate of less than 1 mM inhibited the anaerobic growth of H. volcanii by DMSO respiration. When H. volcanii was cultivated anaerobically in medium supplemented with both 100 mM nitrate and 50 mM DMSO, the OD₆₀₀ value after 5 days of cultivation reached ∼0.6, the same level as in the case of denitrifying growth, suggesting that H. volcanii grew by denitrification rather than by DMSO respiration.

FIG 2.

Effect of nitrate on anaerobic growth of H. volcanii in DMSO medium. (A and B) H. volcanii strain H26 (A) and strain NO02 (B) were cultivated anaerobically in 50 mM DMSO-containing Hv medium in which KNO₃ was not supplemented (open circles) or supplemented at 0.1 mM (squares), 1 mM (diamonds), 10 mM (triangles), and 100 mM (inverted triangles) in a 96-well microplate as described in Materials and Methods. The two strains were also cultivated in medium without DMSO or nitrate (marked by closed circles). The OD₆₀₀ was measured using a microplate reader every 24 h after the start of cultivation. (C) The OD₆₀₀ values of strain H26 (black bars) and strain NO02 (white bars) after 5 days of cultivation under each condition. The mean values and the standard errors (SE) (error bars) were calculated (n = 8).

A reporter assay was carried out to estimate the transcription levels of the dms and nar operons in H. volcanii strain H26 cultivated anaerobically in medium containing 50 mM DMSO and various concentrations of nitrate (from 0 to 100 mM) according to previous reports (4, 5). The reporter analysis using strain DMP1 indicated that the transcription level of the dms operon was decreased in a nitrate-responsive manner (Fig. 3, white bars). At 100 mM nitrate, the transcription of the dms operon reached the minimum level, similar to that in the control cells incubated anaerobically in DMSO-free Hv medium. In contrast, using the strain NAP01, the transcription of the nar operon was shown to be activated depending on the nitrate concentration (Fig. 3, black bars). It is already known that the nar operon is transcribed in H. volcanii cells incubated anaerobically without the addition of nitrate (4). The nar operon was also transcribed in cells cultivated anaerobically in Hv medium containing 50 mM DMSO, and the transcription level was greatly increased by the addition of nitrate at >10 mM (Fig. 2A and C). The nitrate-responsive changes in the anaerobic growth of H. volcanii in DMSO medium, shown in Fig. 2A and C, may be explained by inhibition of the transcription of the dms operon and activation of that of the nar operon while increasing the nitrate concentration. This result is also consistent with the downregulation of the dms operon under denitrification conditions, as revealed by transcriptome analysis (Fig. 1).

FIG 3.

Effect of nitrate on the transcription activities of the dms operon and the nar operon of H. volcanii cultivated anaerobically in DMSO medium. H. volcanii strain NAP01 (black bars) and strain DMP1 (white bars) were cultivated anaerobically in Hv medium containing 50 mM DMSO and various concentrations of nitrate (0 to 100 mM) in a microplate. After 5 days of cultivation, archaeal cells of both strains were harvested, and the induced β-galactosidase (BgaH) activity in Miller units (m.u.) was measured to estimate the transcription of the nar and dms operons as described in Materials and Methods. The mean values and the SE were calculated from the results of three independent experiments.

A cultivation experiment was also conducted using strain NO02, a ΔnarO mutant of H. volcanii. The growth rate of strain NO02 in anaerobic Hv medium containing 50 mM DMSO was not affected by the addition of nitrate ranging from 0 to 100 mM (Fig. 2B and C). Strain NO01 cannot grow anaerobically by denitrification due to a lack of NarO that participates in transcription activation of the denitrifying genes (4). The experimental results demonstrated that the nitrate-responsive inhibition of anaerobic growth by DMSO respiration is due to the transcription suppression of the dms operon. Furthermore, NarO may also be involved in the nitrate-dependent transcription suppression of the dms operon, in addition to the activation of the nar operon under anaerobic conditions (4).

Effect of DMSO on anaerobic growth by denitrification.

For analyzing the effect of DMSO on denitrifying growth, H. volcanii strain H26 was cultivated anaerobically in Hv medium containing 50 mM nitrate and various concentrations of DMSO. Supplementation of DMSO at concentrations ranging from 0 to 10 mM to the medium did not affect the growth rate, and the OD₆₀₀ of each culture had reached ∼0.6 4 days after the start of anaerobic cultivation (Fig. 4A). The growth rate was gradually decreased when DMSO was added at >10 mM. A similar result was obtained in the cultivation experiment using strain DR1, a ΔdmsR mutant that could not grow anaerobically by DMSO respiration due to the lack of DmsR, which controls the expression of the DMSO reductase (5) (Fig. 4B and C). Ansel et al. (39) reported that 10% DMSO strongly inhibited the growth of E. coli, Pseudomonas aeruginosa, and Bacillus megaterium and induced cytological changes in E. coli cells. However, it has been shown that DMSO at a low concentration of ∼100 mM (equivalent to 0.78%), which showed growth inhibition in H. volcanii cells, did not affect the growth of these bacteria (39). DMSO is known to form pores in the cell membrane and promote membrane permeability of chemicals (40). The reason why the growth rate of H. volcanii was decreased by low concentrations of DMSO is not clear, but it may be due to the excessive permeability of the cell membrane caused by DMSO.

FIG 4.

Effect of DMSO on denitrifying growth of H. volcanii. (A and B) H. volcanii strain H26 (A) and strain DR1 (B) were cultivated anaerobically in 50 mM nitrate-containing Hv medium in which DMSO was not supplemented (open circles) or supplemented at 0.1 mM (squares), 1 mM (diamonds), 10 mM (triangles), and 100 mM (inverted triangles), as described in Materials and Methods. The two strains were also cultivated in medium without DMSO or nitrate (marked by closed circles). The OD₆₀₀ was measured using a microplate reader every 24 h after the start of cultivation. (C) The OD₆₀₀ values of strain H26 (black bars) and strain DR1 (white bars) after 5 days of cultivation under each condition. The mean values and the SE were calculated (n = 8).

Nitrate-responsive suppression of DMSO respiration.

These experimental results indicated that the induction of DMSO respiration in H. volcanii is suppressed by nitrate-dependent transcription inhibition of the dms operon, whereas the induction of denitrification is not affected by DMSO. E. coli is a facultative anaerobic bacterium and, similar to H. volcanii, can grow anaerobically by using nitrate or DMSO as a respiratory substrate. Nitrate-responsive inhibition of the induction of DMSO respiration has been observed in E. coli (41, 42). Transcription of the E. coli dmsABC genes, which encode a membrane-bound DMSO reductase, is not regulated by DMSO itself but is dually activated by the oxygen sensor/transducer Fnr and the molybdenum binding regulator ModE (43). On the other hand, induction of nitrate respiration in E. coli is controlled by the nitrate-sensing two-component system NarXL (21). The NarL regulator is activated by phosphorylation catalyzed by the nitrate-responsive histidine kinase NarX. Phosphorylated NarL activates the transcription of the nitrate reductase genes narGHJI but also simultaneously suppresses the transcription of the dmsABC genes in E. coli (21). The phototrophic alphaproteobacterium Rhodobacter capsulatus can also grow anaerobically using nitrate or DMSO as the respiratory substrate, and anaerobic growth by DMSO respiration is strongly inhibited by nitrate (44). For Rhodobacter sphaeroides 2.4.1^T, it has been reported that the expression of the periplasmic DMSO reductase DorA is suppressed by nitrate (45). The transcription of the dorA gene is controlled by a cascade-type regulatory system involving oxygen-sensing Fnr and the DMSO-responsive two-component transducer DorSR, while the mechanism for nitrate-responsive inhibition of DMSO respiration remains unknown (46).

In this study, we found that H. volcanii preferentially utilizes nitrate as a respiratory substrate rather than DMSO when both compounds are present under anaerobic conditions. Transcriptome, growth phenotype, and reporter analyses demonstrated that the transcription of the dms operon, encoding a terminal enzyme for DMSO respiration, was suppressed in response to nitrate in anaerobic cells. The results showed that the nitrate-responsive suppression of DMSO respiration is a common process that is observed not only in bacteria but also in archaea. The transcription regulator NarO was previously reported to activate the transcription of the denitrification-related genes in response to anaerobic conditions (4). Here, it was revealed that NarO is also involved in the nitrate-dependent suppression of the dms operon. The redox potential of nitrate (nitrate/nitrite, E₀′ = +433 mV) is higher than that of DMSO (DMSO/dimethyl sulfide [DMS], E₀′ = +160 mV), and ATP synthesis coupled with nitrate respiration is therefore more efficient than that with DMSO respiration (47). The common observation of nitrate-responsive transcriptional suppression of the dms operon in both bacteria and archaea suggests that the acquisition of the genetic mechanism is a beneficial microbial survival strategy.

The concentration of dissolved oxygen in Dead Sea water saturated by 1 atm air (45 μM at 25°C) is quite low compared with that in freshwater (258 μM) and that in seawater (212 μM) (48). Shatkay et al. (49) reported that nearly saturated concentrations of oxygen are dissolved in Dead Sea water up to a depth of 300 m. On the other hand, <1 mM nitrate is present at all depths in Dead Sea water (50). The concentration of DMSO in Dead Sea water has not been reported, but the presence of up to 180 nM DMSO has been indicated in a hypersaline lake in Canada (51). It seems that, at least in a planktonic lifestyle, facultative anaerobic haloarchaea, including H. volcanii, cannot grow actively by denitrification or DMSO respiration in such a moderate aerobic environment. H. volcanii was not isolated from the water column of the Dead Sea but was isolated from the sediment (1). The growth of haloarchaea by anaerobic respiration may occur only in the sediment of the Dead Sea, where the oxygen concentration can be expected to be close to zero. Metatranscriptome analysis of the microbial communities in the Dead Sea and investigation of nitrate and DMSO concentrations in the sediments may provide information on the ecological significance of anaerobic respiration in the haloarchaea and the mechanism of its regulatory system.

MATERIALS AND METHODS

Strains used in this study.

H. volcanii strain H26, an orotate:phosphoribosyl transferase (pyrE2) mutant of strain DS70, was kindly supplied by T. Allers (Institute of Genetics, Nottingham University, UK) and used for cultivation experiments and transcriptome analyses in this study (52). Strains NO02 and DR1, which cannot grow anaerobically by denitrification and DMSO respiration, respectively, were also used for cultivation experiments (4, 5). Two transformants, strains NAP01 and DMP1, derived from strain H26 were used for the transcription activity assay of the nar and dms operons, respectively (4, 5). Genotypes of the H. volcanii strains used in this study are summarized in Table 2.

TABLE 2.

Genotypes of the H. volcanii strains used in this study

Strain	Genotypea	Description and/or reference
H26	ΔpyrE2	52
NO02	ΔpyrE2 ΔnarO	4
DR1	ΔpyrE2 ΔdmsR	5
NAP01	ΔpyrE2{pnarA::bgaH + Novr}	The 5′-flanking region of the HVO_B0160 (narA) gene was used as a promoter of the nar operon for construction of the reporter plasmid (4)
DMP1	ΔpyrE2{pdmsE::bgaH + Novr}	The 5′-flanking region of the HVO_B0362 (dmsE) gene was used as a promoter of the dms operon (5)

^aEpisomal plasmids are indicated by curly brackets.

Microplate cultivation.

Cultivation experiments with the mutants, reporter strains, and their parental strain H26 were carried out using Hv medium that was prepared as follows. A medium that contained 5.0 g · liter⁻¹ Bacto yeast extract (Becton, Dickinson and Company, Sparks, MD), 2.0 g · liter⁻¹ tryptone (Oxoid Ltd., Basingstoke, Hampshire, UK), 2.0 g · liter⁻¹ KCl, 176.0 g · liter⁻¹ NaCl, 20.0 g · liter⁻¹ MgCl₂·6H₂O, and 0.1 g · liter⁻¹ CaCl₂·2H₂O was adjusted to pH 7.4 before autoclaving. Solutions of chelated iron (100 mg · liter⁻¹ FeSO₄·7H₂O and 100 mg · liter⁻¹ EDTA) and trace elements (100 mg · liter⁻¹ Na₂MoO₄·2H₂O, 200 mg · liter⁻¹ MnCl₂·6H₂O, 2 mg · liter⁻¹ CoCl₂·6H₂O, 100 mg · liter⁻¹ ZnSO₄·7H₂O, and 100 mg · liter⁻¹ CuSO₄·5H₂O), which were also prepared and autoclaved separately, were mixed with the medium at a volume of 1/1,000 each. The resulting Hv medium was used for the cultivation of the strain (4). Hv medium that was supplemented with an appropriate amount of KNO₃ and/or DMSO was used for the anaerobic cultivation of the mutant strains.

Aerobically cultivated H. volcanii cells at a late exponential growth phase (OD₆₀₀ = ∼0.7) were harvested and then diluted with fresh Hv medium containing respiratory substrates to make the OD₆₀₀ 0.05. The OD₆₀₀ of the medium was measured using an MPS-2000 spectrophotometer (Shimadzu Co., Kyoto, Japan). For efficient evaluation of anaerobic growth, cultivation of H. volcanii was carried out on a microplate as described previously by Jantzer et al. (53). The diluted medium was dispensed at 200 μl into each well of the 96-well assay plate (flat bottom, with a diameter of 6.4 mm by 10.8 mm; Asahi Glass Co. Ltd., Tokyo, Japan), and 100 μl of liquid paraffin (Wako Pure Chemical Industries Ltd., Osaka, Japan) was then layered on the medium to prevent volatilization. The assay plate was sealed in an airtight container with an AnaeroPack oxygen absorber (Mitsubishi Gas Chemical Co. Inc., Tokyo, Japan). Anaerobic growth of the haloarchaea was very slow in the medium without the addition of nitrate (see Fig. S1 in the supplemental material), confirming that sufficient anaerobicity had been maintained under this experimental condition. Every 24 h after the start of static cultivation at 37°C under dark conditions, the assay plate was taken out of the container, and the OD₆₀₀ was measured using a microplate reader (Infinite M200 Pro; Tecan, Männedorf, Switzerland). The plate was immediately sealed again in the container with a new oxygen absorber.

Differential gene expression analysis of H. volcanii H26 under aerobic and denitrifying conditions.

Total RNA of H. volcanii strain H26 cultivated under aerobic and denitrifying conditions was extracted using a PureLink RNA minikit (Thermo Fisher Scientific, Waltham, MA). rRNA molecules were removed from the total RNA using the Ribo-Zero rRNA removal kit for bacteria (Illumina Inc., San Diego, CA), and the rRNA-depleted RNA was purified using the RNeasy MinElute cleanup kit (Kapa Biosystems Inc., Woburn, MA). Libraries for strand-specific RNA sequencing were constructed using the Kapa stranded mRNA-seq kit (Kapa Biosystems) according to the manufacturer’s instructions, with 14 cycles of library amplification, except for skipping an mRNA capture step. The libraries were sequenced using a HiSeq 4000 platform (Illumina) to generate 2- by 101-bp paired-end sequence reads at Macrogen Inc. (Seoul, South Korea).

The raw reads were cleaned up using Trimmomatic version 0.36 by trimming adapter sequences, the last 101 bases, and low-quality ends with a quality score of less than 15 and dropping out reads of less than 75 bp (54). The resultant high-quality reads were aligned to the genome sequence of H. volcanii DS2 (GenBank accession numbers CP001953.1 to CP001957.1), an ancestral strain of H. volcanii H26, using HISAT2 version 2.2.0 with the options –dta, –rna-strandness RF, and –no-spliced-alignment (55). Read counts were calculated from the BAM files using featureCounts version 2.0.0 with a strand-specific read-counting option (56), and TPM values were calculated to normalize gene lengths and total read counts. Differential gene expression analysis between the transcriptomes of H. volcanii H26 cultivated under aerobic and denitrifying conditions was performed using the edgeR package version 3.16.5 (57). Read counts for genes were filtered by removing lowly expressed genes with a count per million (CPM) value of less than 2 and normalized with scaling factors calculated for the library sizes using the trimmed mean of M values (TMM) method (58). Differentially expressed genes (DEGs) were identified by the likelihood-ratio test implemented in the edgeR package, which is capable of the detection of DEGs with no replicates. Typical values for the common biological coefficients of variation (BCVs) are 0.1 for data on genetically identical model organisms (57). To conservatively detect DEGs, we set a dispersion value of 0.1, which is larger than that calculated from this BVC value. Genes with a logFC value of >|2| and an FDR of <0.05 were defined as DEGs.

Growth phenotype analysis.

Hv medium containing 50 mM DMSO and 0, 0.1, 1.0, 10, or 100 mM KNO₃ was used for analysis of the effects of nitrate on anaerobic growth by DMSO respiration. Anaerobic cultivation of the strains H26, NO02, and DR1 was carried out as described above without measuring the OD₆₀₀ every 24 h. The OD₆₀₀ value of the medium 5 days after the start of cultivation was regarded as the growth activity. To evaluate the effect of DMSO on denitrifying growth, Hv medium containing 50 mM KNO₃ and 0, 0.1, 1.0, 10, or 100 mM DMSO was prepared and used for a similar cultivation experiment.

Reporter assay.

The transcription activity of the nar promoter was determined by measuring the induced β-galactosidase (BgaH) activity using H. volcanii strain NAP01, which had been generated by Hattori et al. (4). Strain NAP01 was anaerobically cultivated in 0.5-mg · liter⁻¹ novobiocin-supplemented Hv medium containing 50 mM DMSO and 0, 0.1, 1.0, 10, or 100 mM KNO₃ in an anaerobic container. After 5 days of cultivation, the medium (200 μl × 96 wells = 19.2 ml) was collected, the OD₆₀₀ value was measured using an MPS-2000 spectrophotometer, and the culture was then harvested by centrifugation at 10,000 × g for 10 min. BgaH activity was measured according to methods described previously by Hattori et al. (4). The transcription activity of the dms promoter was also analyzed by a similar procedure using strain DMP1, which had been prepared by Qi et al. (5).

Data availability.

Raw reads for RNA-seq analyzed in this study have been deposited in the DDBJ Sequence Read Archive (DRA) under accession numbers DRR213973 and DRR213974 for the transcriptome of H. volcanii H26 cultivated under aerobic and denitrifying conditions, respectively.