ABSTRACT
Methanosaeta spp. are widely distributed in natural environments, and their filamentous cells contribute significantly to sludge granulation and the good performance of anaerobic reactors. A previous study indicated that Methanosaeta harundinacea 6Ac displays a quorum sensing-regulated morphological transition from short to long filaments, and more acetate is channeled into methane production in long filaments, whereas more is channeled into biomass synthesis in short filaments. Here, we performed transcriptomic and physiological analysis to gain insights into active methanogenesis in long filaments of M. harundinacea 6Ac. Both RNA sequencing (RNA-seq) and quantitative reverse transcription-PCR indicated that transcription of the genes involved in aceticlastic methanogenesis and energy metabolism was upregulated 1.2- to 10.3-fold in long filaments, while transcription of the genes for the methyl oxidative shunt was upregulated in short filaments. [2-13C]acetate trace experiments demonstrated that a relatively higher portion of the acetate methyl group was oxidized to CO2 in short filaments than in long filaments. The long filaments exhibited higher catalase activity and oxygen tolerance than the short ones, which is consistent with increased transcription of the oxidant-scavenging genes. Moreover, transcription of genes for cell surface structures was upregulated in the long filaments, and transmission electron microscopy revealed a thicker cell envelope in the filaments. RNA-seq determined a >2-fold upregulation of a variety of antistress genes in short filaments, like those encoding chaperones and DNA repair systems, which implies that the short filaments can be stressed. This study reveals the genetic basis for the prevalence of the long filamentous morphology of M. harundinacea cells in upflow anaerobic sludge blanket granules.
INTRODUCTION
Methanogenic degradation of organic complexes is a widely used approach in wastewater treatment; it promotes the development of upflow anaerobic sludge blanket (UASB) reactors (1). Efficient mineralization of organic complexes to CH4 and CO2 in the UASB reactor is implemented by diverse microbes through a food chain mode, where methanogens implement the final chemical reaction producing CH4 (2, 3). In UASB reactors and other ecosystems, acetate-derived methane (aceticlastic methanogenesis) contributes about 70% of the methane produced by either the generalist aceticlastic methanogens of the species Methanosarcina or the obligate aceticlastic methanogens of the species Methanosaeta (4, 5). Methanosaeta species are believed to be the key components in anaerobic digesters not only because of their ability to use very low concentrations of acetate (threshold, 5 to 20 μM) (6) but also because of their fiber-like cells that serve as a scaffold for the attachment of other organisms to promote the formation of UASB granules (7), an essential self-immobilized organization of the functional microbes. It has been reported that Methanosaeta cells comprise one-third of the anaerobic migrating blanket reactor (AMBR) granule biomass (7).
Methanosaeta harundinacea 6Ac was isolated from a UASB reactor treating beer manufacturing wastewater in Beijing, China (8). When isolated, strain 6Ac cells were long and filamentous, but short filaments grew upon subculturing in the laboratory. Later, we demonstrated that this cell morphology transition is regulated by quorum sensing (9), probably through the FilI/FilR two-component signal transduction system (10). We found that more substrate acetate is channeled into cell biomass synthesis in the short filaments, while more is channeled into methane production in the long filaments; the latter has a lower Ks value and a higher Vmax value in acetate conversion than the short filaments (9). These observations indicate that the long filamentous morphology of strain 6Ac cells is the more active cell form in methanogenesis, yet the genetic basis that enables the robustness of the long filaments remains unclear.
In this study, through RNA sequencing (RNA-seq)-based comparative transcriptomic analysis and assays of relevant physiological characteristics, we attempted to identify the gene profiles related to robust methanogenesis by long filamentous cells, the predominant UASB morphology of Methanosaeta strain 6Ac. Our results indicated that transcription of the genes responsible for aceticlastic methanogenesis, energy metabolism, and ribosome proteins was upregulated in the long filaments, which also exhibited a higher tolerance for oxidative stress and an enhanced reactive oxygen species (ROS)-scavenging ability, while in the short filaments, which appeared to be stressed, both transcriptomic and isotope tracing analysis determined that the methyl oxidative shunt was upregulated.
MATERIALS AND METHODS
Methanogen strain and culture conditions.Methanosaeta harundinacea 6AcT (JCM 13211, CGMCC 1.5026, and DSM 17206) was preserved in our laboratory. A previously described prereduced mineral medium containing 100 mM sodium acetate was used for routine cultures, and 150 ml culture was dispensed into a 250-ml anaerobic bottle under an atmosphere of N2-CO2 (4:1) (8, 9, 11). To culture the long filaments with ≥10 cells inside one sheath, 10% of the late-logarithmic-phase short-filament culture grown in 50 mM acetate was used as the inoculum, because of the high levels of acyl homoserine lactones (AHLs) in this phase. The short filaments (<10 cells inside one shell) were cultured by inoculating 10% of the same short-filament culture in the stationary phase, when AHL levels are low. All cultures were incubated at 37°C, unless indicated otherwise.
Total RNA extraction for RNA-seq.Long and short strain 6Ac filaments were harvested in the early-mid-log-phase from a 165-ml culture (Fig. 1C) by centrifugation at 13,400 × g for 15 min at 4°C. The cell pellets were quickly frozen in liquid N2 and then stored at −80°C. Total RNA was extracted using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions with minor modifications, as follows. Cell breakage was carried out with 5 cycles of 1 min each in a mini-bead beater (Biospec Products, Bartlesville, OK, USA) at 4.2 kHz and quickly frozen in liquid nitrogen. Next, the mixture was centrifuged at 13,400 × g at 4°C for 10 min, and the supernatant was moved into RNase-free tubes. Total RNA was then extracted with phenol-chloroform and precipitated with isopropanol as previously described (12). RNA quality and quantity were determined on a 1% agarose gel, and 1 unit of RNase-free DNase I and 50 units of RNase inhibitor (Invitrogen) were then added to 2 μg RNA to digest the remaining chromosomal DNA. This was performed at 37°C for 8 h, and removal of contaminated DNA was then verified by PCR.
M. harundinacea 6Ac cell morphology and methane production. (A and B) Epifluorescence microscopic images of the long (A) and short (B) filaments were taken at 420 nm. Bars = 10 μm. (C and D) Rates of methane production from acetate were determined for growing batch cultures (C) and resting cells (D) of the short (◼) and long (●) filaments. Data are the means and standard deviations from three biological replicates of each culture. Arrow, the sampling point for RNA-seq.
Ion Torrent sequencing and mapping.Five micrograms of total RNA was again treated with DNase I to remove DNA in either RNase-free water or Tris-EDTA buffer. rRNAs were depleted with Ribo-Zero RNAs (Epicentre, Illumina Company, San Diego, CA, USA). A sequencing library was constructed following the standard protocols of the Ion Torrent sequencing technology (Life Technologies).
Two single-end strand-specific libraries were generated for the RNAs extracted from the long and short filaments, which contained 3,320,287 and 2,686,614 high-quality reads, respectively. To test their reproducibility, two additional libraries, L (long filaments) and S (short filaments), were prepared from biological replicates in the early mid-log phase. The biological replicate libraries generated 2,635,191 (L library) and 2,224,565 (S library) high-quality reads. The RNA-seq reads were aligned to the M. harundinacea 6Ac genome using BWA software (13) with four mismatches and the flag is “-l 8 –O 0 -E 0 -o 3.” Reads mapped to rRNA as well as those not mapped or mapped to multiple positions by use of these parameters were discarded. This yielded 2,166,126 and 2,187,758 unique mapped reads for the long- and short-filament RNAs, respectively, and 1,922,658 and 1,488,560 reads from the two biological replicates (L and S libraries, respectively). These reads accounted for over 70× sequence coverage of the entire genome. On the basis of the alignment information, custom Perl scripts were used to calculate transcript coverage depth information.
Analysis of differential transcript abundance.Unique mapped reads were used for transcript abundance analysis in both samples and the biological replicates. Transcript abundance for a gene was evaluated by normalization of the read counts to the total mapped unique reads and gene length with the reads per kilobase per million mapped reads (RPKM) function (14). Differences in transcription were identified using the DEGseq package (15) according to the mapped reads count. Those with signatures considered true (P < 0.001) were noted to be differentially transcribed genes, and those with fold changes of >2 or <0.5 were defined to be either up- or downregulated genes, respectively.
Operon definition and gene reannotation.Operons were defined on the basis of the continuous transcript sequence coverage that extended into a codirectional upstream gene (16) and were predicted with our script. On the basis of the transcript coverage data, 12 misannotated genes of the automatic annotation were corrected (see Table S1 in the supplemental material); in these cases, the predicted translation start site occurred upstream of the actual transcription start site.
Quantitative reverse transcription-PCR (qRT-PCR).RNA was quantified by a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Bremen, Germany). Using random primers, the cDNAs were prepared with a GoldScript cDNA synthesis kit (Invitrogen) according to the manufacturer's instructions. Quantitative PCR (qPCR) experiments were carried out in a Mastercycler ep realplex2 S apparatus (Eppendorf, Hamburg, Germany) using Thunderbird SYBR qPCR mix (Toyobo, Osaka, Japan) as a reporter dye. The real-time PCR oligonucleotide primers (Table 1) were designed using Beacon Designer (version 7.0) software (Premier Biosoft, Palo Alto, CA, USA) to gain maximum amplification efficiency and sensitivity. For quantification of the transcript abundance for a gene, a DNA fragment containing the target gene was amplified by PCR and quantified on the NanoDrop ND-1000 spectrophotometer. These fragments were then serially 10-fold diluted (10−2 to 10−9) in triplicate and used to generate a standard curve for quantification. The 16S rRNA gene copy number in the cell was used as a reference. Quantification of a gene transcript was normalized against the 16S rRNA gene copy numbers. The qPCR program was as follows: 1 cycle of 95°C for 30 s and then 40 cycles of 95°C for 10 s, 60°C for 30 s, and 72°C for 30 s (17). Eighteen differentially transcribed genes (see Table S2 in the supplemental material) were tested to verify the RNA-seq data.
Oligonucleotide primers used for qRT-PCR in this study
Determination of resting-cell methane production rates.Two batches each of long and short filaments were harvested during their mid-log growth phases by centrifugation at 5,000 × g for 10 min under anaerobic conditions. Prior to methane production rate determination, the total cell proteins of both types of cells were measured. Cell pellets were washed twice with 0.1 M potassium phosphate buffer (phosphate-buffered saline [PBS]; pH 7.2) and resuspended in 0.5 ml PBS. Cells were lysed by ultrasonication at 260 W in a cycle of 1 s sonication and 2 s pause for 20 min. Total protein was quantified with the Coomassie protein assay reagent (Thermo Fisher Scientific). A standard curve for protein concentration was generated with lysozyme.
Cells at an equivalent of 10.8 mg protein were collected and washed three times with anaerobic washing buffer (pH 7.2, containing KH2PO4, 0.545 g; Na2HPO4, 1.336 g; NaCl, 0.234 g; cysteine-HCl, 0.250 g; and NaHCO3, 3.20 g, per liter deionized water under a gas phase of N2-CO2 [80:20]). Cell pellets were suspended in 2 ml anaerobic washing buffer containing 200 mM sodium acetate in anaerobic tubes and were then used to measure methane production. Methane production from acetate was determined for each of the three long and short-filament biological replicates at 37°C. Methane was determined on a GC-14B gas chromatograph (Shimadzu, Kyoto, Japan) as described by Ma et al. (8). All procedures were performed anaerobically unless indicated otherwise.
CE preparation.Long- and short-filament cell extracts (CEs) were prepared as described previously, with modifications (18, 19). Briefly, cultures were harvested from mid-log-phase cells by centrifugation at 5,000 × g at 4°C for 15 min, and the cell pellets were washed twice with chilled 0.1 M PBS (pH 7.2). Cell pellets were moved to sterile, precooled 1.5-ml Eppendorf tubes, and 0.5 ml PBS containing 0.2 mM phenylmethylsulfonyl fluoride was then added. Cell pellets were lysed by ultrasonication at 240 W in a cycle of 1 s sonication and 2 s pause for 15 min. The cell lysate was centrifuged at 13,400 × g at 4°C for 15 min to precipitate the cell debris, and the supernatant (CE) was collected in a precooled Eppendorf tube and kept on ice. Total protein was quantified as described above.
Incorporation of [2-13C]acetate.[2-13C]acetate was added to the routinely cultured long and short filaments of M. harundinacea 6Ac at a final concentration of 2% (wt/wt) total acetate. Cultures without isotope-labeled acetate were included as controls. Both cultures were prepared in triplicate. Acetate was determined with a GC-14B gas chromatograph (Shimadzu) with previously reported parameters (11). The stable isotope composition was determined with a Trace GC Ultra gas chromatograph (Thermo Fisher Scientific, Milan, Italy) which was connected to a Delta V Advantage isotope ratio mass spectrometer with a GC Isolink combustion reactor interface (Thermo Fisher Scientific, Bremen, Germany) (20).
Fluorescence microscopy method.Fifty microliters of the strain 6Ac culture was spread on a glass slide and dried at room temperature. The fluorescence at a 420-nm wavelength was examined under a Leica DMI 3000 B microscope (Leica Microsystems GmbH, Wetzlar, Germany).
TEM.Long and short filaments were harvested in the late-log growth phase, and samples for transmission electron microscopy (TEM) were prepared as described previously, with slight modifications (21). Cell pellets were fixed with 2.5% glutaraldehyde overnight, followed by postfixation in 1% osmium tetroxide for 2 h; after dehydration and infiltration, samples were embedded in Spi-pon 812 resin (Spi Supplies, West Chester, PA, USA), polymerized, and then sectioned with a Leica EM UC6 ultramicrotome (Leica Microsystems GmbH) (22). The ultrathin sections were approximately 70 nm thick. They were stained with uranyl acetate and lead citrate and observed under an FEI Tecnai Spirit transmission electron microscope (FEI, Hillsboro, OR, USA).
Catalase activity assay.Bubble generation was first used to test hydrogen peroxide scavenging (23) with 25 μl 9.0-mg/ml CEs from both long and short filaments. Catalase activity was determined by the hydrogen peroxide decomposition method (24) with modifications: 65 μl 0.1 mM hydrogen peroxide in PBS and 25 μl 9.0 mg/ml CEs were mixed in an Eppendorf tube for 1 min at room temperature. The reaction was stopped by adding 565 μl stop solution containing 2.5 mM 4-amino-antipyrine (4-amino-2,3-dimethyl-1-phenyl-3-pyrazolin-5-one; Sigma, San Francisco, CA, USA) and 0.17 M phenol. After 4 min at room temperature, horseradish peroxidase (Sigma) was added at a final concentration of 50 mU/ml. After another 4 min at room temperature, the optical density at 510 nm was measured with a Unico 2100 visible-light spectrophotometer (Shanghai, China). A standard curve was generated with chemical H2O2.
Oxygen tolerance assay.The oxygen tolerance of both types of cells was detected by adding various volumes of air to sterile tubes with 80% N2 and 20% CO2, which generated final oxygen concentrations of 0%, 0.37%, 0.74%, and 1.48%; each oxygen concentration was prepared in triplicate. Next, 5-ml cultures in the mid-log phase were added to the air-containing tubes, and methane production was monitored.
RESULTS
Genome-wide differentially transcribed genes in long versus short filaments of M. harundinacea.By inoculating short filaments of strain 6Ac in their late log or stationary phase, long filaments (Fig. 1A) and short filaments (Fig. 1B) were cultured, respectively. When consuming the same quantity of acetate, the long filaments had a higher methane yield (12.81 ± 0.01 mmol) than the short filaments (11.86 ± 0.01 mmol), while the short filaments exhibited higher methanogenic rates (0.135 ± 0.023 mmol/day) than the long ones (0.105 ± 0.047 mmol/day) in the batch culture (Fig. 1C). To differentiate the methane production rates caused by differences in cell morphology from those caused by differences in biomass during growth, the long and short filamentous resting cells (10.8 mg cell protein) were used to determine methane production from acetate (200 mmol/liter). The results revealed that the long filaments exhibited higher methanogenic rates (0.025 ± 0.000 mmol/day) than the short ones (0.017 ± 0.001 mmol/day) (Fig. 1D). This indicates that the higher methanogenic rate observed in the short filaments during the growth phase is a result of a higher biomass.
To examine the gene profiles with differential transcription in long versus short filaments, high-resolution RNA-seq (Ion Torrent) was performed. The comparison indicated that 683 genes showed transcript abundances that differed by more than 2-fold between long and short filaments; of these, 243 and 440 genes were upregulated in the long and short filaments, respectively (see Table S3 in the supplemental material).
Furthermore, using qRT-PCR, we quantified the transcript abundance for 18 genes that differed in their levels of transcription in the transcriptomic analysis (see Table S2 in the supplemental material). qRT-PCR and RNA-seq data were well correlated (correlation coefficient [r] = 0.810) (see Fig. S1 in the supplemental material), indicating the reliability of the RNA-seq analysis.
Higher levels of transcription of the genes for methanogenesis and energy metabolism genes in long filaments.Among the genes with differential transcription, those involved in the aceticlastic methanogenesis pathway, including acs, cdh, mtr, and mcr, were upregulated 1.5- to 10.3-fold in the long filaments (Fig. 2). In particular, two acetyl coenzyme A synthetase genes (Mhar_0749, Mhar_0751), which function in the limiting initial step, increased their transcript levels 10.1- and 3.8-fold, respectively, and the transcript abundance of the methyl coenzyme M (methyl-CoM) reductase operon (Mhar_0495 to Mhar_0498) increased 2.2- to 10.3-fold in the long filaments. Moreover, the levels of transcription of the F420 H2 dehydrogenase operon (Mhar_1410 to Mhar_1420), which is involved in oxidative phosphorylation, and the vacuolar type H+-ATP synthase operon (Mhar_2253 to Mhar_2261) were also increased in the long filaments compared to their levels of transcription in the short filaments (Table 2). The enhanced transcription of genes involved in energy metabolism in the long filaments provides genetic evidence for the robustness of long filamentous cell morphology.
Differential expression of genes involved in aceticlastic methanogenesis and the methyl oxidative shunt in M. harundinacea 6Ac. Bar graphs flanking the methanogenic pathways show the related transcript abundances detected in the transcriptomes of short and long filaments. Red and green, genes upregulated in long and short filaments, respectively; black, genes for which there was no differential transcription. *, transcription was verified by qRT-PCR. MFR, methanefuran; THMPT, tetrahydromethanopterin; MPT, methanopterin; HS-CoB and CoB, coenzyme B; CoB-S-S-CoM, heterodisulfide of coenzyme M and coenzyme B; HS-CoM, coenzyme M; Acetyl-S-CoA, acetyl-coenzyme A; acs, acetyl coenzyme A synthetase; [CO], CO-carbon monoxide dehydrogenase complex; cdh, CO dehydrogenase; Fd, ferredoxin; fmd, molybdenum-formyl-MFR dehydrogenase (Mhar_1283 to Mhar_1288); fwd, tungsten-containing formyl-MFR dehydrogenase (Mhar_0014, Mhar_0373 to Mhar_0376, and Mhar_2308 to Mhar_2310); ftr, formyl-MFR:THMPT formyltransferase (Mhar_2214); mch, N5N10-methenyl-THMPT cyclohydrolase (Mhar_2174); mtd, F420-dependent N5N10-methylene-THMPT dehydrogenase (Mhar_1470); mer, F420-dependent N5N10-methylene-THMPT reductase (Mhar_0856); mtr, N5-methyl-THMPT methyltransferase (Mhar_2090 to Mhar_2097); mcr, methyl-CoM reductase (Mhar_0495 to Mhar_0498 and Mhar_0529); hdr, heterodisulfide reductase (Mhar_0604 to Mhar_0606 and Mhar_0792 to Mhar_0793).
Oxidative phosphorylation and biomass synthesis gene transcript abundance in long versus short filaments of M. harundinacea 6Aca
Coenzyme M-coenzyme B heterodisulfide reductase (Hdr) is key to the cellular supply of coenzyme M, a compound essential for the reduction of methyl to CH4. Two types of hdr operons are present in the M. harundinacea 6Ac genome, and hdrED (Mhar_0792, 0793) transcription was upregulated in long filaments. The level of expression of the membrane-bound hdrD gene was approximately 9.3-fold higher than that of hdrABC (Fig. 2) at the transcription level. This suggests that HdrED plays a major role in aceticlastic methanogenesis in M. harundinacea 6Ac. Additionally, most of the ribosome-encoding genes were upregulated in long filaments (see Table S4 in the supplemental material).
However, expression of some genes in the methyl oxidation shunt, such as the tungsten formylmethanofuran dehydrogenase operon (fwd), was higher in short filaments than in long filaments at the mRNA level (Fig. 2). qRT-PCR also determined a 2-fold upregulation of a gene for the F420 redox reaction (Mhar_2358) involved in the methyl oxidative shunt in short filaments (see Table S2 in the supplemental material), implying that more reducing equivalents are gained in short filaments than in long filaments. Moreover, genes involved in gluconeogenesis were upregulated at the mRNA levels in short filaments (see Table S5 in the supplemental material). This indicates that biomass synthesis is more active in short filaments than in long filaments, which is consistent with previous findings (9).
The methyl oxidative pathway is active in short filaments.To compare methyl oxidative shunt activity in both types of cells, 2-13C-labeled acetate was applied to determine 13CO2 production from the methyl group of acetate, which is believed to be reduced to CH4. Upon acetate depletion by incubating strain 6Ac at 37°C for 55 days, the 13CO2 yield was higher in short filaments (δ13CCO2, −1.87 ± 0.05) than in long filaments (δ13CCO2, −14.31 ± 0.30) (P < 0.01) (Table 3), supporting the idea that the methyl oxidative shunt is more active in short filaments.
Gas chromatography/isotope ratio mass spectrometry-determined δ13CCO2 and δ13CCH4 in the culture gas phases and total methane and carbon dioxide yield from the acetate consumed by long and short filaments of M. harundinacea 6Ac
By subtracting the CH4 yield from the consumed acetate, we calculated that, under the culture conditions used in this study, 0.5% and 3.6% of the acetate methyl carbon-formed CO2 was produced via the methyl oxidative shunt in the long and short filaments, respectively (Table 3). It is believed that the methyl oxidative shunt is used to generate reducing equivalents that can be used for ATP synthesis or cell biomass, because the mch mutants of Methanosarcina barkeri and Methanosarcina acetivorans have lost the ability to grow on acetate (25). Therefore, the isotope tracing experiment supported the different ratios of acetate channeled to CH4 production and biomass formation in cells of the two morphologies.
Long filaments efficiently scavenge ROS.ROS are lethal to anaerobes because they possess insufficient ROS-scavenging proteins; these include superoxide dismutase (SOD) and catalase in aerobes and superoxide reductase (SOR) in anaerobes (26, 27). M. harundinacea 6Ac possesses genes encoding catalase, SOD, and other peroxidases but not genes encoding SOR. Some oxidant-scavenging genes were upregulated in long filaments (Table 4) at the transcription level. This included catalase (Mhar_0135), flavoprotein (Mhar_0340), thioredoxin (Mhar_1159), and alkylhydroperoxidase-like protein (Mhar_1400). However, the genes for rubredoxin-type Fe(Cys)4 protein (Mhar_1374) and ferritin (Mhar_1375) were upregulated in short filaments at the transcription level (Table 4). This implies that cells of the two morphologies may exhibit varied potentials in antioxidative stress.
Transcript abundance of genes encoding stress-related proteins in long versus short filaments of M. harundinacea 6Ac
To compare ROS-scavenging activity in both types of cells, catalase activity was detected by adding 0.093 mmol hydrogen peroxide. More bubbles were produced in the long-filament culture than in the short-filament culture. Furthermore, 0.047 ± 0.002 mmol H2O2 was degraded within 1 min by 0.225 mg CE from the long filaments, while only 0.023 ± 0.001 mmol H2O2 was removed by the same amount of CE from the short filaments. This indicates higher oxidant-scavenging activity in the long filaments.
To further determine the oxygen tolerance of both types of M. harundinacea 6Ac cells, various volumes of air were added to the gas phase of the cultures. A growth assay showed that the short filaments ceased growth when 0.74% oxygen was added to the headspace, while the long filaments stopped growing with 1.48% oxygen (Fig. 3). Both the enzymatic assay and O2 tolerance further confirmed the higher antioxidative stress capacity of the long filaments.
Oxygen inhibition on long (A) and short (B) filaments of M. harundinacea 6Ac. Methane production was measured upon the addition of different volumes of air, from which the final oxygen concentration was calculated. ○ and ●, no oxygen; □ and ◼, 0.37% oxygen; △ and ▲, 0.74% oxygen; ▽, 1.48% oxygen. Data are the means and standard deviations from three biological replicates of each culture.
Moreover, we did not observe oxygen-induced expression of the genes encoding ROS-scavenging proteins, indicating that they are not regulated by oxygen, but the quorum sensing-regulated morphology alteration is related to oxidant clearance in M. harundinacea 6Ac.
Short filaments appear to be stressed.As in bacteria and other archaea, a gene cluster (Mhar_2125 to Mhar_2127) for chaperones (dnaJ, dnaK, and grpE) was present in the M. harundinacea 6Ac genome. Transcriptomic analysis indicated that these genes were upregulated in the short filaments (Table 4). Additionally, two of the three thermosome subunit genes (Mhar_0878 and Mhar_2128) that function as protein-folding chaperones and a gene encoding the heat shock protein Hsp20 (Mhar_1971) were also upregulated at the mRNA level in short filaments. Moreover, many genes related to DNA repair were upregulated in short filaments at the transcription level (see Table S6 in the supplemental material). These data combined suggest that M. harundinacea 6Ac short filaments, the morphology induced under laboratory conditions, were stressed.
Cell envelope synthesis and cell division are enhanced in long and short filaments, respectively.Epifluorescence microscopic images revealed that long filaments of M. harundinacea 6Ac consisted of multiple rod cells wrapped in a single sheath (Fig. 1A). RNA-seq determined that several genes for cell surface proteins, including an S-layer protein gene (Mhar_0562) and a cell surface protein gene (Mhar_1933), had increased levels of transcription in the long filaments (see Table S7 in the supplemental material). Additionally, genes encoding ABC transporters for tungstate (Mhar_1363 and Mhar_1364) and molybdate (Mhar_2341 to Mhar_2343) also had increased transcript abundance.
Ultrathin sections of long and short filaments were examined by electronic microscopy to identify the cell structures relevant to the differential transcription of the cell surface protein genes. Thicker cell walls and a large intracellular cleat-like structure were observed in the long filaments (see Fig. S2 in the supplemental material). The cross-sectional profiles also revealed that 90% of long filamentous cells were wrapped in an irregular envelope (see Fig. S2B in the supplemental material); this probably resulted from the presence of abundant cell surface proteins. Such a structure was infrequently observed in the short filaments (see Fig. S2D in the supplemental material).
RNA-seq revealed that DNA replication and cell division genes were upregulated in the short filaments of M. harundinacea 6Ac (see Table S8 in the supplemental material). This could be related to fast growth in the short filaments.
DISCUSSION
Methanosaeta strains are characterized by their ability to produce methane even from very low concentrations of acetate, which may be the reason why they are so widely distributed, e.g., in anaerobic waste digesters, rice paddies, and natural wetlands (28, 29), and are the predominant methane producers in many environments (30). The fiber-like Methanosaeta cells are particularly advantageous in the formation of granules, hence their important role in wastewater treatment (31, 32). Previously, we have found that M. harundinacea 6Ac undergoes a quorum sensing-regulated cell morphology transit from short to long filaments, and the former channels more acetate to cell biomass synthesis, while the latter directs more into methane production (9). This work, through analyzing the differentially transcribed gene profiles, isotope tracing, and biochemical experiments, reveals the molecular basis of the different physiological features between the two types of cells.
CO2 reductive methanogenesis is the only pathway for methane production in the hydrogenotrophic methanogens, while it acts as the methyl oxidative shunt to generate reducing equivalents in the methylotrophic methanogens. Smith and Mah reported that small quantities of acetate are oxidized to produce reductive equivalents for biosynthesis in Methanosarcina sp. strain 227 (33). The function of this pathway for generating reductive biosynthesis has been confirmed in Methanosarcina species growing on acetate through assays of enzymatic activity (5, 34). It has been further demonstrated that Methanosarcina mutants with a deletion of mch, a gene encoding N5N10-methenyl-tetrahydromethanopterin cyclohydrolase in the methyl oxidative shunt, fail to grow in acetate (25). Though implementing obligate aceticlastic methanogenesis, Methanosaeta spp. carry a suite of genes for the CO2-reductive methanogenesis pathway in their genomes. Through molecular and enzymatic studies, as well as 2-13C-labeled acetate tracing assays, we have determined that this pathway also acts as the methyl oxidative shunt in M. harundinacea 6Ac (11). This work demonstrates that the methyl oxidative shunt is active only in short filaments, which is consistent with the previous finding that the short filaments produce more biomass but less methane. Data from RNA-seq and the physiological experiment show that short filaments of M. harundinacea 6Ac exhibit a lower methanogenic rate and lower oxidative tolerance than long filaments but higher levels of transcription of the chaperone and DNA repair genes than long filaments. Hence, the laboratory-induced short filaments are under stress, while the long ones are healthier. In Escherichia coli, the filamentous phenotype can overcome host innate immunity, for instance, during urinary tract infection, as the phenotype can protect cells from the lethal environment (35). Similarly, long filaments can benefit Methanosaeta species survival in diverse environments, such as those in the oxidative stressed state.
To test whether quorum sensing-regulated long filaments of M. harundinacea 6Ac benefit UASB granulation, we carried out experiments by dosing either M. harundinacea alone or M. harundinacea combined with its spent culture containing the quorum-sensing signal molecules into UASB reactors treating synthetic wastewater. After 3 months, we found that quorum sensing-mediated long filaments of M. harundinacea 6Ac did promote granulation and constant chemical oxygen demand (COD) removal efficiency. Moreover, 16S rRNA homolog analysis detected the presence of Methanosaeta spp. in the UASB granules and found that species diversity was increased in the long filaments of M. harundinacea 6Ac-dominant UASB granules including not only methanogen species but also, presumably, functional bacterial species, such as syntrophic bacteria (data not shown).
A recent study found that the obligate aceticlastic methanogen M. harundinacea 6Ac even carries out CO2-reductive methanogenesis when it gains electrons from Geobacter metallireducens through nanowires (36). This indicates that dormant pathways in microbes can be activated when the right triggers are present, thus maximizing their metabolic potential.
ACKNOWLEDGMENTS
We thank Tong Sun in the State Key Laboratory of Microbial Resources for valuable suggestions on mass spectrum experiments and Yanxia Jia and Lei Sun in the Center for Biological Imaging, Institute of Biophysics, Chinese Academy of Sciences, for TEM work.
This work was supported by a grant from the National Natural Science Foundation of China (31100035).
FOOTNOTES
- Received 20 September 2014.
- Accepted 11 November 2014.
- Accepted manuscript posted online 14 November 2014.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.03092-14.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.