Enrichment and dynamics of novel syntrophs in a methanogenic hexadecane-degrading culture from a Chinese oilfield

Authors


Correspondence: Yahai Lu, College of Resources and Environmental Sciences, China Agricultural University, Beijing 100193, China. Tel.: +86 10 62733617; fax: +86 10 62733617; e-mail: yhlu@cau.edu.cn

Abstract

Methanogenic communities that degrade alkanes have been reported. However, little is known about the key players involved in the process. Methanogenic hexadecane-degrading consortia were enriched from an oilfield (Shengli, China). The microbial dynamics during the transfer incubations were monitored using terminal restriction fragment length polymorphism (T-RFLP) fingerprinting of 16S rRNA genes in combination with cloning and sequencing. The archaeal community shifted from a predominance of aceticlastic Methanosaeta during early cultivation to a substantial increase in hydrogenotrophic Methanoculleus in the highly enriched culture. Bacterial T-RFs 161 and 164 bp were consistently detected during the incubation and became dominant in the highly enriched culture. T-RF 161 bp primarily represented uncultured Waste Water of Evry 1 bacterium, which was possibly associated with Candidatus Cloacamonas acidaminovorans (99.7% sequence similarity). T-RF 164 bp could be assigned to both Thermotogaceae, with the closest relative being Candidatus Mesotoga sulfurreducens (similarity of 97%) and Syntrophaceae, with Smithella propionica as the closest relative (similarity of 96–97%). These bacterial lineages were potentially capable of syntrophic interactions with methanogen partners during hexadecane degradation. Partial assA genes (encoding the α-subunit of alkylsuccinate synthase) were also detected, implying that the mechanism of fumarate addition may function in the hexadecane activation.

Introduction

In the past two decades, it has been shown that the anaerobic degradation of hydrocarbons is widespread under anoxic conditions such as in oil reservoirs and hydrocarbon-contaminated environments (Jones et al., 2008), which has shed light on alternative strategies for in situ bioremediation and enhanced oil recovery (Zengler et al., 1999; Anderson & Lovley, 2000). As one of the main components of crude oil, the anaerobic degradation of aliphatic alkanes has also attracted attention. A number of aliphatic alkane-degrading anaerobes with nitrate, sulphate and iron (III) as electron acceptors were isolated and characterised in the past two decades (Widdel et al., 2010 and cited references).

Under electron acceptor-limited conditions, the methanogenic degradation of alkanes is a common process, and the critical importance of syntrophic interactions between alkane degraders and methane producers has been discussed from the thermodynamic viewpoint (Dolfing et al., 2008). Several methanogenic consortia that degrade long- and/or short-chain alkanes have been obtained (Zengler et al., 1999; Gieg et al., 2008; Jones et al., 2008; Siddique et al., 2011, 2012; Wang et al., 2011; Mbadinga et al., 2012; Zhou et al., 2012). Phylogenetic analysis has revealed that Syntrophus-related members are dominant in methanogenic enrichment cultures that degrade long-chain alkanes (n-C14H30, n-C16H34 and n-C18H38; Zengler et al., 1999; Siddique et al., 2011), short-chain alkanes (n-C6H14, n-C7H16, n-C8H18 and n-C10H22; Siddique et al., 2012), and crude oil alkanes (Jones et al., 2008; Gray et al., 2011). Additionally, several dominant phylotypes related to Firmicutes, Clostridiales, Thermotogales, Chloroflexi, Thermodesulfobiaceae and Bacteroidetes have also been detected in other reports (Gieg et al., 2008, 2010; Wang et al., 2011; Mbadinga et al., 2012; Zhou et al., 2012). Five putative methanogenic pathways have been proposed for the conversion of alkanes into methane (Dolfing et al., 2008). Gieg et al. (2008) revealed that a crude oil alkane-degrading consortium was primarily composed aceticlastic Methanosaeta. Subsequently, they reported that hydrogenotrophic Methanothermobacter spp. dominated in a thermophilic crude oil alkane-degrading consortium (Gieg et al., 2010). Jones et al. (2008) observed that 86–87% of archaeal clones belonged to hydrogenotrophic methanogens and speculated that the CO2 reduction pathway of methanogenesis was linked to syntrophic acetate oxidation. Archaeal phylotypes associated with CO2 reduction and acetate fermentation methanogens were also detected together in previous reports (Zengler et al., 1999; Wang et al., 2011).

Apparently, diverse microbial communities are involved in the methanogenic degradation of alkanes. However, knowledge about the community dynamics during transfer incubation and the core microorganisms involved in alkane degradation is scarce, which is a roadblock to understanding the microbial mechanism of alkane degradation under methanogenic conditions. In this study, we enriched a hexadecane-degrading methanogenic consortium sampled from a Chinese Shengli oilfield. By characterising the archaeal and bacterial dynamics during successive transfer incubations, we aimed to identify the probable microbial members that are responsible for the anaerobic degradation of hexadecane and methane production.

Materials and methods

Enrichment cultivations

The inoculum for methanogenic enrichments was sampled from a disposal field that treats mixtures of crude oil-contaminated soil and oily sludge in the Shengli oilfield of China on 18 February 2009. Samples (c. 4–5 kg) were placed into zip lock plastic bags, of which the inside part (c. 100 g) was incubated on 24 February 2009. Anaerobic freshwater medium was prepared using the Hungate anaerobic technique (Macy et al., 1972), which contained the following (per litre): 0.5 g NaCl, 0.5 g MgCl2.6H2O, 0.1 g CaCl2.2H2O, 0.1 g NH4Cl, 0.2 g KH2PO4, 0.5 g KCl, 1 mg resazurin and 0.5 g L cysteine hydrochloride. Aliquots were distributed into vials sealed with butyl rubber stoppers (Bellco) and aluminium caps under a gas mixture of N2/CO2 (80 : 20, v/v). During transfer incubations, Fuxin glass vials (China) and butyl rubber stoppers (China) were chosen for their higher seal performance. The medium was autoclaved for 30 min at 121 °C and then reduced with Na2S.9H2O (0.03%). An anoxic sterile NaHCO3 (0.5%) solution, vitamin solution (2 mL L−1; Widdel et al., 2006), trace element solution (2 mL L−1; Widdel et al., 2006), vitamin B12 (2 mL L−1; Widdel et al., 2006), vitamin B1 (2 mL L−1; Widdel et al., 2006) and crude oil-contaminated soil extract (Kuhner et al., 1997) were injected into the medium before inoculation, and the pH was adjusted to 7.0–7.2.

Approximately 10 g of a mixture of oil-contaminated soil and oily sludge was distributed into vials containing 50 mL fresh medium under a N2 : CO2 gas mixture (v/v, 4/1) and incubated statically at 35 °C in the dark. The enrichments that showed methane production were transferred into 60 mL fresh medium with 15% inoculum (v/v). The experimental groups were supplemented with 40 μL hexadecane (ca. 137 μmol; Sigma), and the control groups did not receive hexadecane. Successive transfer incubation was carried out with 9–30% inoculum with hexadecane as a sole substrate.

Analytical procedure

Methane was analysed using gas chromatography (Shimadzu GC 2010, Japan) with a Porapak Q column and thermal conductivity detector (Cheng et al., 2011). The temperatures of column, oven and detector were 50, 50 and 70 °C, respectively, and the carrier gas was hydrogen (99.999%), with a flow rate of 50 mL min−1. Gas samples (0.2 mL) were injected into the column using pressure-lock syringes (Vici). The total amount of gas products was calculated based on Avogadro's law after calibration with a gas mixture (N2: CH4: CO2 = 29.96%: 39.99%: 30.05%).

Volatile fatty acids (acetate, propionate and butyrate) were analysed by high-performance liquid chromatography (Agilent HPLC 1200) with a 150 × 4.6 mm C18 column (Agilent Eclipse XDB-C18) and UV detector. The carrier liquid was 15% acetonitrile (Sigma), with a flow rate of 1 mL min−1, and its pH was adjusted to 2.5 by phosphoric acid addition.

DNA extraction, cloning and sequence analysis

Approximately 1.5–3 mL of liquid enrichment cultures sampled from the exponential growth phase of the methanogenic hexadecane-degrading consortium during transfer incubations was regularly collected and stored at −20 °C for genomic DNA extraction according to Peng et al. (2008). DNA products were purified with the Promega wizard DNA clean up system (Promega) and stored at −25 °C.

The archaeal and bacterial 16S rRNA gene fragments were amplified using primer sets A109F/A934R and B27F/B907R, respectively (Peng et al., 2008; Rui et al., 2009). Reaction mixtures of 50 μL consisted of 5–50 ng template DNA, 5 μL 10 × PCR buffer, 4 μL dNTP (each 2.5 mM), 3 μL MgCl2 (25 mM), 1 μL each forward and reverse primer (each 10 μM) and 0.25 μL Takara Taq (5 U μL−1; TakaRa, Japan). Amplification with archaeal primers was carried out as follows: 94 °C for 4 min, followed by 30 cycles of 94 °C for 1 min, 52 °C for 1 min, and 72 °C for 2 min, with a final extension at 72 °C for 7 min. The polymerase chain reaction conditions for bacterial 16S rRNA gene application were as follows: an initial denaturation step of 3 min at 94 °C; 29 amplification cycles of 30 s at 94 °C, 45 s at 53 °C, and 80 s at 72 °C; and 10 min at 72 °C. The putative assA gene fragments were amplified using primers assF/assR, 1432F/assR and 1294F/1936R, respectively (Callaghan et al., 2010), with some modifications. The reaction was performed in 25 μL reaction mixtures containing 5–50 ng template DNA, 2.5 μL 10 × PCR buffer, 2 μL dNTP (each 2.5 mM), 1.5 μL MgCl2 (25 mM), 0.5 μL each forward and reverse primers (each 50 μM) and 0.25 μL Takara Taq (5 U μL−1). The PCR of assA genes was performed according to Callaghan et al. (2010).

The PCR products were purified using the LangGang general DNA agarose gel recovery kit (LangGang, China). The cloning and sequencing of gene fragments were conducted according to Rui et al. (2009). The 16S rRNA gene sequences were evaluated with the ‘Chimera check with Bellerophon’ program of greengene database (DeSantis et al., 2006) and aligned with clustal x software (Larkin et al., 2007). Using the furthest neighbour method in the dotur program with a 97% threshold (Schloss & Handelsman, 2005), sequences from three 16S rRNA gene clone libraries were grouped into operational taxonomic units (OTUs). Representative sequences from each OTU were selected for phylogenetic analysis. The phylogenetic dendrogram of archaeal and bacterial 16S rRNA genes was constructed using the arb program package (Ludwig et al., 2004), and the phylogenetic tree of deduced amino acids sequences of assA genes was constructed with mega 5 (Tamura et al., 2011).

The nonparametric richness estimators Chao1 (Chao, 1984) and ACE (Chao & Lee, 1992; Chao et al., 1993) were used to evaluate the species richness of the samples based on the number of rare OTUs using the dotur program (Schloss & Handelsman, 2005). The diversity coverage of the constructed clone library was calculated using Good's formula (Good, 1953) as C = [1 − (n1/N)] × 100, where n1 is the number of unique OTUs and N is the total number of clones in the library.

Nucleotide sequence accession number

The aligned sequences were deposited in the GenBank database with the following accession numbers: HQ132939HQ133109 (archaeal and bacterial 16S rRNA genes from the first subculture), HQ689140HQ689322 (archaeal and bacterial 16S rRNA genes from the second subculture), HQ704416HQ704438 (assA gene from the first subculture), JN181728JN181766 (archaeal 16S rRNA genes from the fourth subculture) and JN202626JN202718 (bacterial 16S rRNA and assA genes from the fourth subculture).

T-RFLP analysis

The PCR amplification of bacterial and archaeal 16S rRNA genes used the same primers as described above, but one primer of each set was 5′-end-labelled with 6-carboxyfluorescein (FAM; Peng et al., 2008; Rui et al., 2009). The FAM-labelled PCR products were purified with the TIAN Quick Midi Purification Kit (TIANGEN, China) and digested at 37 °C for bacterial DNA by MspI (TakaRa) and at 65 °C for archaeal DNA by TaqI (TakaRa), as described previously (Peng et al., 2008; Rui et al., 2009). The digestion products were further purified using the ethanol precipitation method. Dried DNA samples were resuspended in 10 μL ddH2O, and a portion was mixed with deionised formamide containing 2% (v/v) of the internal standard ROX 30–1000 (Bioventure), denatured at 95 °C for 3 min and chilled on ice for 10 min. The DNA fragments were separated by capillary electrophoresis on a Genetic Analyzer 3130xl (ABI), and the relative T-RF abundances of representative phylotypes were analysed with genemapper software 4.0 (ABI) according to Peng et al. (2008) and Rui et al. (2009). The T-RF values of the phylotypes represented in the 16S rRNA gene clone libraries were calculated according to the determined sequences by mega 5 (Tamura et al., 2011), verified and adjusted to the calculated T-RF values using the corresponding clones as templates. The relative abundance of T-RFs was calculated as the percentage of an individual T-RF of the sum of all concerned peak heights in a given T-RFLP profile, with T-RFs accounting for < 1% in a given T-RFLP profile and having a detection frequency < 20% in all samples being grouped together.

Results

Methane production in transfer incubations

In the first transfer incubation, significant methane production was detected after 90 days of inoculation in the hexadecane group, and a total of 1.15 mmol of methane was generated at day 180. In contrast, the control group without hexadecane addition accumulated 0.26 mmol of methane after 180 days of incubation. Furthermore, only trace amounts of acetate ranging from 0.023 to 0.071 mM were detected in both groups (Fig. 1b).

Figure 1.

Methane (a) and acetate (b) production in the first enrichment transfers. ■: with C16H34; □: control without C16H34. Error bar: two duplicates with the exception of sampling points at day 167 and 180.

Archaeal dynamics in transfer incubations

The composition of T-RFLP profiles in the first subculture was dominated by T-RF 284 bp and did not fluctuate much (72.6 ± 9.9%) between 91 and 180 days when CH4 was actively produced (Fig. 2). Throughout the transfers, T-RF 284 bp systematically decreased, whereas T-RFs 186 and 495 bp markedly increased. The latter two T-RFs combined became dominant and accounted for 71.6% of the total fluorescence frequencies from the fourth to fifth transfers (Fig. 2).

Figure 2.

The relative T-RF abundance of archaeal 16S rRNA genes retrieved from five enrichment transfers. A representative T-RFLP profile from each transfer is presented. Data from the first and third subcultures are presented as the means ± standard deviation (n = 9–12).

Three clone libraries of archaeal 16S rRNA genes were constructed from the first, second and fourth hexadecane-degrading enrichment cultures. All of the samples reached almost complete coverage (> 90%) based on the Good's coverage estimator, indicating that < 10 additional phylotypes would be expected for each group (Supporting Information, Table S1). All of the archaeal sequences (141 clones) belonged to the phylum Euryarchaeota, and most of them (94.3%) were assigned to three OTUs (Fig. 3). One OTU (type clone HA_1, T-RF 186 bp) was closely related to hydrogenotrophic Methanoculleus (Cheng et al., 2008). These Methanoculleus-like sequences increased from 2.2% in the first transfer to 76.9% in the fourth transfer. The other two OTUs belonged to obligate acticlastic methanogens (Patel, 1984; Ma et al., 2006; Fig. 3). One OTU (type clone HA_22, T-RF 284 bp), which was closed related to Methanosaeta concilii (99% sequence similarity; Patel, 1984), predominated (90.9%) in the first transfer and decreased dramatically in the fourth transfer. The other OUT was represented by T-RF 495 bp, which was almost identical to aceticlastic Methanosaeta harundinacea (99% sequence similarity; Ma et al., 2006) and increased markedly in the second transfer (11/58 clones).

Figure 3.

Phylogenetic tree of archaeal 16S rRNA gene sequences. The clone sequences from this study were generated from the first, second and fourth transfers of hexadecane-degrading methanogenic consortia. Numbers in the first set of parentheses indicate the length of in silico T-RFs; numbers in the second parentheses denote the clone number from different transfers (distinguished by the colour: green for the first, blue for the second and red for the fourth transfer). The scale bar represents 10% sequence divergence. Thermococcus mexicalis (Z75218) was used as the outgroup sequence.

Bacterial dynamics in transfer incubations

The T-RFLP profile of bacterial 16S rRNA genes remained constant in the hexadecane-free groups during the first transfer incubation, which was dominated by the T-RF 212-bp (Fig. 4a). However, T-RFLP profiles in the hexadecane-degrading enrichment cultures changed, with a marked decrease in T-RF 212 bp and the gradual increases in T-RFs 161 and 164 bp over time. T-RFs of 188, 189 and 300 bp also increased during hexadecane incubation (Fig. 4a). Furthermore, T-RFLP analysis was performed only for hexadecane cultures during exponential phase, and a representative profile is shown for comparison (Fig. 4b). In the second and third transfers, the relative abundance of 161 bp increased continuously. T-RF 164 bp stayed high, and T-RF 212 bp was present in some cultures but absent in others (Fig. 4b). In the fourth and fifth transfers, the bacterial diversity decreased substantially, with only five T-RFs in the T-RFLP profiles (Fig. 4b). The 161 bp accounted for over half of the total fluorescence frequency, the 164 bp remained the second most dominant T-RF, and the 209, 212 and 261 bp showed low abundances. Some T-RFs (154, 189, 209, 301, 488, 505, 517, 518 and 526 bp) that were occasionally detected in the early transfers were absent in these highly enriched cultures (data not shown).

Figure 4.

The relative T-RF abundance of bacterial 16S rRNA genes retrieved from enrichment incubations. (a), comparison of T-RFLP profiles between control (C16H34−) and hexadecane (C16H34+) incubation in the first transfer; (b), comparison of T-RFLP profiles among five enrichment transfers. A representative T-RFLP profile from each transfer is presented. Data from the first and third subcultures are presented as the means ± standard deviation (n = 9–12).

Three bacterial clone libraries were also generated from the same time points as archaea (Table S1). The diversity index analysis indicated that over 90% of predicted phylotypes were detected in the second and fourth transfers, but only 80% was detected in the first transfer (Table S1). The species richness of the bacterial community decreased from the first to fourth transfers based on the ACE and Chao1 estimator analysis at the species level (Table S1). The phylogenetic analysis of 331 clones revealed that uncultured members related to uncultured JS1 (T-RFs 154 and 207 bp), WWE1 phylotypes (T-RF 161 bp), Spirochaetales (T-RF 202 bp), Thermotogaceae and Syntrophaceae (both T-RF 164) were consistently present in the three enrichment cultures (Fig. 5). A large proportion of sequences that belonged to Chloroflexi (mostly Anaerolineaceae like), Clostridiales and the uncultured JS1 bacterium that were abundant in the libraries of the first and second transfers decreased in the fourth transfer. In contrast, the community in the later highly enriched cultures was dominated by members related to the uncultured WWE1 bacterium (161 bp), Syntrophaceae (164 bp), Thermotogaceae (164 and 261 bp) and Spirochaetaceae (212 bp). Although Bacteroidetes (91 bp) also increased in the clone library of the fourth transfer, this trend was not evident in T-RFLP profiles (Fig. 5).

Figure 5.

Phylogenetic tree of archaeal 16S rRNA gene sequences. The clone sequences from this study were generated from the first, second and fourth transfers of hexadecane-degrading methanogenic consortia. Numbers in the first set of parentheses indicate the length of in silico T-RFs, and numbers in the second set of parentheses denote the clone number from different transfers (distinguished by the colour: green for the first, blue for the second and red for the fourth transfer). The symbol ‘*’ denotes the dominant T-RF. The scale bar represents 10% sequence divergence. Aquifex pyrophilus (M83548) was used as the outgroup sequence.

Composition of assA genes

We constructed two assA gene clone libraries from the first and fourth transfers. A total of 37 sequences (23 from the first transfer and 14 from the fourth transfer) were analysed. The results revealed that all sequences belonged to a single OTU (> 99% identity; Fig. 6). The deduced amino acid sequences showed 94% sequence similarity to an environmental clone (GenBank No. GU453662) retrieved from a paraffin-degrading methanogenic enrichment and was distantly related (64% sequence similarity) to a pure culture of Desulfatibacillum alkenivorans AK-01, which possessed an identical conserved cysteine residue.

Figure 6.

Phylogenetic tree of the inferred amino acid sequences (c. 154 aa) of assA genes. The sequences from this study were retrieved from the first and fourth transfers of hexadecane-degrading methanogenic consortia. Numbers in the parentheses indicate the clone number from the first (in green) and fourth (in red) transfers. The tree was constructed using mega5.0 software with 1000 bootstrap replications. The bootstrap values are given at nodes when > 50%. The putative amino acid sequence of bssA of Geobacter toluenoxydans (EF123666) was used as the outgroup. The scale bar represents 5% sequence divergence.

Discussion

In the present study, we enriched hexadecane-degrading consortia from Chinese Shengli oilfield under methanogenic conditions. A long growth period (c. 200 days) was required to degrade hexadecane under methanogenic conditions. This result is consistent with previous studies (Zengler et al., 1999; Widdel et al., 2010; Siddique et al., 2011), and may be caused by the insolubility and chemical inertness of alkanes and the slow growth rate of alkane degraders. Approximately 56.8 ± 4% of hexadecane was converted to methane according to the theoretical equation of the methanogenic degradation of hexadecane (Eqn. (1)), which is consistent with previous reports on incomplete recovery, most likely due to the long period of incubation, absorption by the rubber stopper or unknown reasons (1999a, 1999bSo & Young, 1999a; Zengler et al., 1999; Winderl et al., 2010).

display math(1)

Five possible methanogenic pathways have been proposed for alkane degradation based on thermodynamic analysis (Dolfing et al., 2008), which was partially confirmed through the phylogenetic analysis of methanogenic communities involved in alkane degradation (Mbadinga et al., 2011). It is noteworthy that a shift in the methanogenic structure during successive transfers was observed in this study, indicating that the methanogenic pathway shifted from acetate fermentation to CO2 reduction. This observed phenomenon may contribute to the availability of methanogen precursors, suggesting that H2/CO2 rather than acetate may be the main intermediate of hexadecane degradation. The dynamics of aceticlasitc methanogenic populations were also monitored in this study, which may be affected by the incubation conditions. However, further research is required to evaluate this hypothesis.

The bacterial community changed from a diverse composition to the enrichment of organisms related to the uncultured WWE1 bacterium, Thermotogaceae, Syntrophaceae, Spirochaetaceae and Bacteroidales in the later transfer incubations. The uncultured WWE1 bacterium (characterised by the 161 bp T-RF) was most likely the most abundant bacterial member in the highly enriched culture. The clone sequences retrieved in the present experiment were closely related to Candidatus Cloacamonas acidaminovorans, whose genome has been reconstructed via the metagenomic method (Pelletier et al., 2008). Bioinformatics analysis revealed that this organism harbours genomic machinery for the anaerobic syntrophic oxidation of propionate. It also contains a large proportion of genes of unknown functions, suggesting unrevealed metabolic potential. However, its role in alkane degradation has not been determined. The second most dominant group, which was characterised by T-RF 164 bp, was most likely related to members of Thermotogceae. The sequences retrieved in our study showed 97–99% sequence similarity to Candidatus Mesotoga sulfurreducens (Ben Hania et al., 2011) and Mesotoga prima (Nesbø et al., 2012), which belong to the ‘Mesotoga’ lineage of Thermotogaceae, and have the ability of degrading proteinaceous substrates, some sugars and lactate. Thermotogales-related phylotypes were theorised to be involved in the biodegradation of hydrocarbons via syntrophic or fermentation oxidation under methanogenic (Gieg et al., 2010) and sulphate-reducing conditions (Berlendis et al., 2010) or syntrophic acetate oxidation with hydrogenotrophic methanogens (Balk et al., 2002). However, there has been no evidence of a pure culture of Thermotogaceae capable of hydrocarbon degradation under methanogenic conditions. The pure isolates cannot degrade long-chain fatty acids in co-culture with hydrogenotrophic methanogens (Sousa et al., 2009; Ben Hania et al., 2011).

The third probable group is related to Syntrophaceae or Desulfovibrionaceae. Both of these organisms were also characterised by T-RF 164 bp (shared with Thermotogaceae). The Syntrophaceae sequences retrieved in our study showed 96–97% similarity to the syntrophic propionate-oxidising bacterium Smithella propionica LYP (Liu et al., 1999). Similar sequences were previously detected in methanogenic alkane-degrading enrichment cultures sampled from different geological locations (Zengler et al., 1999; Jones et al., 2008; Callaghan et al., 2010; Siddique et al., 2011; Wang et al., 2011). Recently, Gray et al. (2011) proposed that Syntrophaceae-related members played a key role in complete oxidation of crude oil alkanes to acetate and/or hydrogen in syntrophic partnership with methanogens. Two Desulfovibrio- and Soehngenia-like strains (characterised by T-RF 164 and 301 bp, respectively) have been isolated from this methanogenic, hexadecane-degrading consortium in our laboratory. However, no growth was observed using hexadecane as a sole substrate in co-culture with hydrogenotrophic methanogens (data not shown).

The members of Spirochaetaceae (209 and 212 bp) and uncultured JS1 bacterium (154 and 207 bp) were particularly abundant in the early transfers but decreased in later transfers. A representative clone of Spirochaetaceae shared 87% sequence similarity with Treponema primitia ZAS-2, which is a H2-utilising homoacetogen isolated from termite guts (Leadbetter et al., 1999). Although JS1-related bacterium have not yet been cultivated, Webster et al. (2006) proposed that uncultured JS1 bacterium may metabolise acetate, a key intermediate during the methanogenic degradation of alkanes (Dolfing et al., 2008). Other organisms that were occasionally abundant in transfer incubations include Anaerolineaceae (T-RFs 517 and 518 bp), Synergistaceae (T-RF 179 bp) and Soehngenia (T-RFs 300 and 301 bp), suggesting their uncertain role in alkane degradation.

Fifteen clones retrieved from the first and second transfers were related to D. alkenivorans AK-01 (sequence similarity of 94–95%; So & Young, 1999a). The characteristic T-RF was 166 bp, and it was present in early transfers but decreased in the later transfers. Interestingly, we recovered 37 sequences of the assA gene (encoding alkylsuccinate synthase) from the first and fourth transfers that were also distantly related to D. alkenivorans AK-01. This organism was isolated from petroleum-contaminated estuarine sediment and degraded alkanes (C13–C18) and 1-alkenes (C15 and C16) under sulphate-reducing conditions (So & Young, 1999b; Callaghan et al., 2006, 2008). Genome analysis revealed that they utilised the fumarate addition reaction for the activation of alkanes and degraded hexadecane in co-cultures with hydrogenotrophic methanogens (Callaghan et al., 2012). The detection of putative assA genes in the present experiment implies that the fumarate addition reaction may also play a role in the methanogenic degradation of hexadecane. However, further studies are necessary to conclude this point because of the low similarity to D. alkenivorans AK-01.

In summary, we demonstrated that both archaeal and bacterial communities shifted during the enrichment cultivation of hexadecane-degrading consortia. Archaeal populations changed with a substantial increase in the hydrogenotrophic methanogen Methanoculleus in the later stage of enrichment cultivation. The bacterial populations related to the uncultured WWE1 bacterium, Thermotogaceae and Syntrophaceae, most likely played an important role in hexadecane degradation in the highly enriched cultures. All of these organisms are potentially capable of syntrophic interactions with methanogen partners during the degradation of hexadecane. However, the concerted function of other anaerobes, such as Spirochaetaceae, Anaerolineaceae, Synergistaceae and Bacteroidales cannot be ruled out, especially in the early stages of enrichment cultivation. The detection of assA-like sequences leads us to propose that the fumarate addition reaction may function in the methanogenic degradation of hexadecane. Given that most sequences retrieved in the present study are related to uncultured organisms, it is necessary to isolate the key organisms in the future to reveal their degradation mechanisms. It may also be important to employ other molecular methods such as SIP (Winderl et al., 2010) and GeneFISH (Moraru et al., 2010), to detect the active players in natural environments to reveal the ecophysiology of these organisms.

Acknowledgements

We thank Quan Yuan for technical assistance, Pengfei Liu for the introduction of arb software, and Xia Li for help with methane determination. This study was supported by the National Nature Science Foundation of China (40830534; 40625003; 30900049; 40973059).

Ancillary