DNA-SIP Reveals That Syntrophaceae Play an Important Role in Methanogenic Hexadecane Degradation

The methanogenic degradation of linear alkanes is a common process in oil-impacted environments. However, little is known about the key players involved in this process. Here, the hexadecane-degrading organisms in a methanogenic, hexadecane-degrading consortium designated M82 obtained from Shengli oilfield and maintained at 35°C for over 4 years, were identified by DNA-stable isotope probing with UL-13C-hexadecane, followed by density-resolved terminal restriction fragment length polymorphism (T-RFLP) analysis, cloning and phylogenetic analysis of 16S rRNA gene fragments. Compared to the fractions of the 12C treatment, the relative abundance of two phylotypes significantly increased in the heavy fractions of the 13C-hexadecane incubated microcosm. One belongs to a uncultured member of the bacterial family Syntrophaceae, which show 95–97% rRNA sequence identity with Smithella propionica, and the other is affiliated with Methanoculleus receptaculi (>99% sequence identity). The results of the present study prove the significant role of uncultured Syntrophaceae in degradation of hexadecane, probably through syntrophic interactions with hydrogenotrophic methanogens.


Introduction
Crude oil is the world's major source of energy and a critical raw material for many chemical products. However, spills of such contaminated compounds during exploration, transport and production may cause environmental pollution. On the other hand, large amounts of oil are trapped in most oil reservoirs even with state-of-the-art oil production techniques [1,2]. Recently, the discovery of methanogenic degradation of hydrocarbons may offer an alternative route to clean up oil-contaminated environments and exploit difficult-to-use oil resources [3,4]. Zengler et al. (1999) was the first to report methanogenic degradation of hexadecane, a representative aliphatic hydrocarbon of crude oil. Larter and colleagues inferred that anaerobic microbial activity may predominate in deep subsurface oil reservoirs [5], and deduced that the anaerobic degradation of hydrocarbons prevails in biodegraded oil reservoirs [6]. Recently, they proved that methanogenic degradation of hydrocarbons widely occurs in oil reservoirs and CO 2 reduction is thought to be the dominant methane production pathway [7]. A large number of anaerobic microorganisms, including fermenting bacteria, nitratereducing bacteria, iron-reducing bacteria, sulphate-reducing bacteria and methanogens, have been detected in oil fields through culture-dependent and culture-independent approaches [8,9]. The methanogenic communities degrading petroleum hydrocarbons have been reported to occur not only in petroleum reservoirs [7,10,11,12,13], but also in oil sands tailings ponds [14,15], freshwater sediment [3] and oil-contaminated soils and sediments [16,17,18], suggesting that it is a common process in hydrocarbon-impacted environments. Thermodynamic analysis revealed the significant importance of syntrophic relationships between hydrocarbon-degraders and methanogenic archaea during methanogenic degradation of hydrocarbons [19]. Previous reports revealed that many uncultured phylotypes affiliated with syntrophic bacteria are present in petroleum hydrocarbon degrading methanogenic consortia [3,7,14,20]. However, little is known about the actually degrading microorganisms under methanogenic conditions.
Stable isotope probing is a powerful method that directly links functional microorganisms to a specific biogeochemical process. This method is based on the principle that key players assimilate stable isotope-labeled substrates and convert them into cell biomass (nucleic acids, phospholipid fatty acids, or proteins), that can be detected through molecular ecological techniques [21]. Many pure isolates degrading aromatic and aliphatic hydrocarbons under anoxic conditions have been characterized using culture-dependent methods [22]. Furthermore, Several reports suggested that diverse uncultured clades are involved in the anaerobic degradation of aromatic hydrocarbons under sulphatereducing [23,24], nitrate-reducing [25], iron-reducing [26], and methanogenic conditions [18,27]. However, limited studies have focused on long-chain alkane degraders under methanogenic conditions. Interestingly, recently an alkane-degrading sulphatereducing bacterium was reported to degrade hexadecane to methane in co-culture with a hydrogenotrophic methanogen [28]. A methanogenic hexadecane-degrading consortium M82 was enriched from Shengli oilfield, and it was revealed that members of uncultured Waste Water of Evry 1 (WWE1), Thermotogaceae, and/or Syntrophaceae were the most common bacterial phylotypes [29]. However, who are major contributors responsible for the hexadecane degradation is still unclear. In the present study, a time-resolved DNA-SIP experiment using UL-13 C-hexadecane was applied to this enriched consortium to unravel the key players involved in anaerobic hexadecane degradation and methane production.

SIP Incubation
Aliquots of 7 mL anoxic freshwater medium without sulphate and nitrate [30] were prepared and dispensed into 50 mL glass vials using Hungate anaerobic technique [31], in which Na 2 -S.9H 2 O (0.3 g L 21 ), NaHCO 3 (2.5 g L 21 ) and oil-contaminated soil extract (5 ml L 21 ) [32] were added. Resazurin (1 mg L 21 ) was Figure 1. Time course of methane production and 13 C/ 12 C isotopic composition of methane of the consortia grown with hexadecane. A: Methane prodution; B: 13 C/ 12 C isotopic composition of methane 12 C-hexadecane: the consortium amended with unlabeled hexadecane; 13 C-hexadecane: the consortium amended with 13C-labeled hexadecane; hexadecane (-): the controls without hexadecane addition, arrows indicate sampling points for the isopycnic centrifugation of genomic DNA. Error bars indicate standard deviations from three triplicates. doi:10.1371/journal.pone.0066784.g001 used as a redox indicator, A hexadecane-degrading methanogenic consortium M82, maintained at 35uC in our laboratory for several years, which was obtained from Shengli oilfield, was selected for SIP incubation with 30% inoculum (v/v) [29]. A total of three sets of treatments were prepared: one set of microcosms was incubated by adding 1 mL of mixture of UL-13 C-hexadecane (99 atom%, Sigma-Aldrich) and 2,2,4,4,6,8,8-heptamethylnonane (HMN) (v/ v, 5: 1000) (13 replicates), the other set was amended with the same volume of unlabeled hexadecane and HMN (21 replicates). While the control group only received 1 mL HMN (17 replicates).The experiment was set up in a gas atmosphere of 80% N 2 and 20% CO 2 and statically incubated at 35uC in the dark.

Process Measurements
The time course of methane production was measured using a gas chromatograph with a thermal conductivity detector as previously described [33]. The isotopic composition of methane was determined through a TraceGas system interfaced with a mass spectrometer (IsoPrime100, United Kingdom), briefly, gas samples (ca. 0.5 mL) withdrawn from the headspace of the culture vials were diluted and flushed into an IsoPrime trace gas pre-concentrator, CO 2 and trace amount of water were removed through a catalyst (Ascarite+Mg(ClO 4 ) 2 ) and automated cold trap. The purified methane (ca. 0.7 mL) was converted into CO 2 in a combustion furnace at a temperature of 1050uC. Helium (99.999%) was used as a carrier gas at a flow rate of 20 ml min 21 . The carbon isotopic abundance of CO 2 was reported relative to the Vienna PeeDee Belemnite.
One-way analysis of variance (ANOVA) for comparison of the difference between groups with or without hexadecane was accomplished using SPSS for windows 16.0 (The mean difference is significance at the 0.05 level).

Nucleic Acid Extraction and Ultracentrifugation
Triplicate vials were sacrificed periodically for molecular analysis during the experiment (The 13 C-hexadecane group: day 126, 166 and 218, The 12 C-hexadecane group: day 126, 166, 198 and 218; The hexadecane-free group: day 126, 166, 198, and 218). Methanogenic cultures (4 mL) were shaken well before use and centrifuged at 17, 700 g for 4 min at 4uC, and stored at 280uC with the protection of RNAlater (Invitrogen, USA). Total genomic DNA was extracted via a bead-beating method [34], purified using the Promega Wizard DNA cleanup system (Promega, USA), and    [35]. Isopycnic centrifugation was performed at 45,000 rpm for 40 h at 20uC with a Beckman 90 Ti rotor using an Optima TM L-80 XP ultracentrifuge (Beckman, German). Eighteen equal gradient fractions (ca. 500 mL) of density-separated DNA were collected using a syringe pump (Leifu, China), and buoyant density was measured using an AR200 digital refractometer (Reichert, USA). The fractionated DNA recovered in fractions was precipitated by adding two volumes of polyethylene glycol and 1 mL glycogen as previously described [21].

Quantification and T-RFLP Analysis of rRNA Genes in the Density Gradient Fractions
Bacterial 16S rRNA genes in genomic DNA retrieved from the gradient fractions was quantified using a CFX96 touch TM realtime PCR detection system (Bio-Rad, USA) with a primer pair B519f/907r [36]. The PCR reaction mixtures (20 mL) consisted of 2 mL DNA, 7 mL ddH 2 0, 10 mL SsoFast TM EvaGreen supermix (Bio-Rad, USA), and 0.5 mL each of forward and reverse primers (5 pmol). The PCR cycle was carried out as follows: 94uC for 3 min, followed by 40 cycles of 94uC for 30 s, 54uC for 30 s, and 70uC for 30 s. Finally the melting curve was analyzed between 65 and 95uC in 0.5uC increment. Almost full-length 16S rRNA genes of Escherichia coli strain JM109 were diluted (10 5 -10 10 copies mL 21 ) and used to construct a standard curve [37]. Quantitative PCR (qPCR) assays for Syntrophaceae, Methanoculleus and Methanosaetaceae were carried out using primer sets Syn826f/Syn1263r [20], 298f/ 586R [38] and Mst702f/Mst826r [39], respectively. The PCR reactions were performed as described above, but with a PCR annealing temperature of 55uC for Syntrophaceae, 58uC for Methanoculleus and Methanosaetaceae. Clone sequences of BM_12 (KC460267), A5_2 (HQ689161) and A5_7 (HQ689186) were also diluted (10 3 -10 9 copies mL 21 ) and used to construct standard curves, respectively. For the T-RFLP analysis, the density-resolved DNA fractions were PCR amplified using the primer pairs B27f/ B907r for bacteria and A109f/A934r for archaea [34,40]. The 59 end of primers B27f and A934r were labeled with 6-carboxyfluorescein. The PCR products were purified with TIAN Quick Midi Purification Kit (Tiangen, China), and then digested at 37uC using Hae III for bacterial DNA (TakaRa, Japan) and at 65uC using Taq I for archaeal DNA (TakaRa, Japan). The following procedure for T-RFLP analyses was performed using identical settings to previous report [29].

Cloning, Sequencing and Phylogenetic Analysis
The PCR mixtures and procedures for cloning were identical to the aforementioned T-RFLP analysis, except the primers that were not fluorescently labeled. The PCR products were gel purified and cloned into E. coli competent cells JM109 (Takara, Japan) using pMD TM 19-T vector (Takara, Japan) according to the manufacturer's instructions. Single colonies were picked and sequenced using a Genetic Analyzer (ABI, USA) [34].
The sequences were checked using the ''Chimera check with Bellerophon'' program of the greengene database [41] and aligned using the Clustal X software [42]. Distance matrices were calculated using the DNAdist software within the PHYLIP software package version 3.69 with the F84 correction [43]. The aligned sequences were grouped into operational taxonomic units using the farthest-neighbor clustering algorithm of the DOTUR software with a 97% threshold [44]. Diversity coverage of the constructed clone library was calculated using Good's formula [45]. The representative clones from each OTUs were compared to RDP type species using the seqmatch program to determine the closest relatives [46]. Phylogenetic trees of the bacterial and archaeal 16S rRNA genes were created using the neighbor-joining method of Mega 5.1, the genetic distance matrix was estimated using maximum composite likelihood method [47]. Bootstrap values were calculated after 1,000 replications. The GenBank accession numbers of the 16S rRNA gene sequences generated in this study are JX088262 to JX088362 and JX473480 to JX473581.

Hexadecane Degradation under Methanogenic Conditions
The 12 C-hexadecane-degrading consortium began to accumulate methane after 126 days of incubation relative to the hexadecane free control incubations (p = 0.001), and generated a total of 21867 mmol of methane after 218 days of incubation. By contrast, the control without hexadecane addition produced a negligible amount of methane that is only 1661 mmol at day 218 (Fig. 1A). From 5 mL (ca. 17.1 mmol) hexadecane added, about 210 mmol of methane would be produced theoretically according to the stoichiometric conversion of hexadecane into methane and carbon dioxide as calculated by Symons and Buswell equation (Equation I), suggesting 9663% of predicted methane was produced ((substrate-amended methane minus substrate-unamended methane)/theoretical methane). Compared to the group in which 12 C-hexadecane was added, the enrichment cultures with 13 C-hexadecane probably exhibited a lag of greater than 126 days before significant methane production was observed relative to hexadecane free control incubations (p = 0.06). It began to produce a substantial increase of methane production after 198 days of incubation (Fig. 1A), and generated about 120618 mmol of methane at day 218 relative to the substrate-unamended control; corresponding to 5769% theoretically predicted maximum possible methane production. The isotopic composition of 13 CH 4 generated from the 13 C-hexadecane consortium increased up to 45.160.3% after 218 days of incubation (Fig. 1B), while the value retrieved from unlabeled hexadecane consortium didn't exceed 1.05160.001% during 218 days of incubation. These results confirmed the process of methanogenic degradation of hexadecane. Four replicates of enrichment cultures amended with 13 C-hexadecane (total thirteen replicates) didn't produce methane within ca. 200 days of incubation for unknown reasons. Quantitative and Qualitative Analyses of Comparative SIP Gradients As methane production progressed, several different vials were sacrificed for the molecular analysis at different time points (The 13 C-hexadecane group: day 126, 166 and 218; the 12 Chexadecane group: day 126, 166, 198 and 218; the hexadecanefree group: day 126, 166, 198, and 218). T-RFLP analysis revealed that the bacterial communities was mainly composed of T-RFs 77, 160, 203, 207, 215 and 337 bp (Fig. S1A). T-RF 207 bp increased Figure 5. Phylogenetic tree based on bacterial 16S rRNA gene and related type strain and environmental clone sequences using the neighbor-joining method. The sequence of Methanoculleus receptaculi (DQ787476) was used as the outgroup. Scale bar, 2% estimated difference in nucleotide sequence. Numbers in the first parenthesis indicate GenBank number of type clone, the second represent the length of in silico T-RFs; and the thrid denote the clone number, the ''light'' clone sequences were labeled on a light gray background and the ''heavy'' were underlined on a dark gray background. doi:10.1371/journal.pone.0066784.g005 Table 2. Phylogenetic affiliations and numbers of archaeal 16S rRNA sequences retrieved from clone libraries generated from different fractions.

Phylogenetic group
No. of clones

In silico T-RF (bp)
Type clone the closest type strains (similarity) 12 C-hexadecane (D126-1.565) 13  markedly over time in abundance in 12 C-or 13 C-hexadecane microcosms compared to the hexadecane-free group (Fig. S1A). The archaeal T-RFLP profiles of the 16S rRNA genes were mainly composed of four T-RFs of 186, 228, 284 and 495 bp in the hexadecane free, 12 C-hexadecane, and 13 C-hexadecane microcosms (Fig. S1B), the 186-bp T-RF increased over time in hexadecane addition group (Fig. S1B). For the difference of lag period between 12 C-and 13 Chexadecane microcosms, the genomic DNA for isopycnic centrifugation was extracted from exponential phase of both enrichment cultures, not from the same day of incubation. qPCR analysis of the bacterial 16S rRNA gene copies in each individual fraction revealed that the peak of the bulk DNA, with 12 C-hexadecane incubation at day 126 and a buoyant density of 1.565 g mL 21 , was similar to that at day 166. However, a peak at 1.571 g mL 21 was detected at day 198 in the 13 C-hexadecane consortium, shifted slightly to 1.574 g mL 21 after 218 days of incubation (Fig. 2). The relative abundances of bacterial and archaeal 16S rRNA genes from different gradient fraction were analyzed through T-RFLP fingerprinting. For the 13 C-hexadecane incubation, bacterial T-RF 207 bp was significantly enriched in the ''heavy'' fractions of the density gradient after 198 and 218 days of incubation. By comparison, this fragment became dominant in all of the density fractions of 12 C-hexadecane consortium from 126 to 166 days of incubation, and exhibits a decrease trend from light to heavy fractions of the density gradient. Other T-RFs, such as those with 77-and 160-bp, were slightly enriched at the beginning of the incubations (day 126) with 12 C-and 13 C-hexadecane, respectively (Fig. 3). The archaeal T-RF 186 bp dominated the heavy density fractions after successive incubations (day 198 and 218) in the 13 Chexadecane incubation (Fig. 4). On the contrary, this fragment became dominant in all of the observed fraction densities with similar abundance levels in 12 C-hexadecane microcosms after 166 days of incubation.

Phylogenetic Identification of the Labeled 16S rRNA Gene
The in vivo T-RFs representing specific microbial lineages were characterized by cloning and sequencing of 16S rRNA genes. Two bacterial 16S rRNA gene clone libraries (total 154 clones) were generated from gradient fractions containing either ''light'' (unlabeled consortium on day 126, BD = 1.565 g mL 21 ) or ''heavy'' fraction ( 13 C-labeled consortium on day 218, BD = 1.582 g ml 21 ) of nucleic acids as indicated by arrows in Fig. 2. The ''heavy'' bacterial clone library reached almost complete coverage (91.3%) based on the Good's coverage estimator, and 75.3% was detected in the ''light'' clone library (Table 1, Fig. S2). Phylogenetic analysis revealed that over threefourths of sequences were clustered into Deltaproteobacteria, Thermotogaceae, Synergistaceae, and Spirochaetaceae in both clone libraries Figure 6. Phylogenetic tree based on the archaeal 16S rRNA gene and related type strain sequences using the neighbor-joining method. The bootstrap values are given at nodes when .50%. The sequence of Thermococcus mexicalis (Z75218) was used as the outgroup. Scale bar, 2% estimated difference in nucleotide sequence. Numbers in the first parenthesis indicate GenBank number of type clone, the second represent the length of in silico T-RFs; and the thrid denote the clone number, the ''light'' clone sequences were labeled on a light gray background and the ''heavy'' were underlined on a dark gray background. doi:10.1371/journal.pone.0066784.g006 (Fig. 5). Members of Thermotogaceae (20.5%), Spirochaetaceae (20.5%) Synergistaceae (14.8%) and Desulfovibrionaceae (9.6%) dominated the ''light'' community, while sequences affiliated with Syntrophaceae (35.8%), Synergistaceae (14.8%), Thermotogaceae (9.9%) dominated the ''heavy'' library (Table 1). In silico analysis of clone sequences revealed that the 207-bp T-RF represented members of Syntrophaceae, the 77-bp T-RF mainly represented members of Desulfovibrionaceae, 3 of 81 clones generated from ''heavy'' library belong to family Syntrophaceae also account for T-RF 77 bp. The Spirochaetaceaeand Synergistaceae-related members were mainly characterized by T-RF 215-bp and 160-bp, respectively. The uncultured WWE1 bacterium (Waste Water of Evry 1) was characterized by T-RF 337 bp. The Thermotogaceae-affiliated clones mainly represented a T-RF of 203 bp and dominated the ''light'' clone library (Table 1, Fig. 5).
Two archaeal 16S rRNA gene clone libraries (49 clones) were also retrieved from the ''light'' and ''heavy'' fractions as indicated by arrows in Fig. 2, respectively. All of the archaeal sequences belonged to the phylum Euryarchaeota (Table 2, Fig. 6), In the ''light'' clone library, the majority of archaeal clones (73.1%) were clustered into acetoclastic Methanosaeta, and showed quite high 16S rRNA sequence identity (.99%) to Methanosaeta concilii (T-RF 284 bp) or Methanosaeta harundinacea (T-RF 495 bp). The remaining clones showed 99% 16S rRNA sequence identity to Methanoculleus receptaculi, represented by a T-RF of 186 bp. By contrast, all the clones in the ''heavy'' library belonged to members of M. receptaculi (.99%). No clones accounting for 228-and 393-bp T-RF were detected in both microcosms, but these two fragments may be assigned to Methanosaetaceae and Methanomicrobiaceae respectively, based on our previous study using the same inoculum [29].
qPCR analysis across all fractions revealed that the peak of 16S rRNA gene of Syntrophaceae was primarily detected in the fraction of 1.565 g mL 21 in 12 C-hexadecane incubation, shifted to fractions above 1.571 g mL 21 in 13 C-hexadecane incubation (Fig. S3A). Methanoculleus and Methanosaetaceae also shifted from unlabeled to 13 C-labeled microcosms (Figs. S3B and S3C). A tailing of DNA indicative of Methanoculleus into the heavy gradient fraction (1.574 g mL 21 ) of 12 C-hexadecane microcosm at day 166 was also observed (Fig. S3C). Further analysis revealed that Log 16S rRNA gene of Syntrophaceae increased by 1.1360.69 log units in 12 C-hexadecane microcosm (per mL) from day 126 to 218 ( Fig.  S4A), the average increase in abundance of Methanoculleus was 1.5360.49 log units, which is higher than Methanosaetaceae (0.6260.41 log units) ( (Figs. S4B and S4C).

Discussion
To our knowledge, the present study is the first to characterize the key players involved in the methanogenic degradation of long chain n-alkane using DNA-based SIP. 13 C-hexadecane degradation to methane was retarded compared to 12 C-hexadecane (Fig. 1A), which was also reported previously [3]. However, Fowler et al. reported similar growth curves in a methanogenic consortium degrading 12 C-toluene and 13 C-toluene, respectively [48]. One may speculate that there are differences between organisms: some may be more sensitive and some more ''tolerant'' towards nonnatural isotope contents in their substrates (personal communications with Dr. Friedrich Widdel). Members of Syntrophaceae accounting for bacterial T-RF 207 bp may have a more active role in hexadecane degradation as the majority of 13 C-labeled carbon was incorporated into DNA from this phylotype. This was confirmed by the fact that the 207-bp T-RF became predominant in the 12 C-hexadecane consortium at the later logarithmic growth phase. The closest cultured representative of the Syntrophaceae sequences (95%-97% 16S rRNA sequence identity) retrieved in this study is Smithella propionica LYP, a syntrophic bacterium isolated from an anaerobic digester, which dismutated propionate to acetate and butyrate [49,50]. Syntrophaceae-affiliated clones have been detected in a vast number of methanogenic consortia degrading alkanes and petroleum hydrocarbon [3,7,12,14,15,16,51], suggesting their potential ecophysiological role in the process of hydrocarbon degradation under methanogenic conditions. Gray et al. revealed a significant positive correlation between Syntrophaceae-affiliated clones and methane production from a petroleum hydrocarbon-degrading consortium, and proposed that the phylotype played an important role in complete oxidation of crude oil alkanes to acetate and/or hydrogen [20]. In this study, we further identified a novel uncultured member of Syntrophaceae that probably played a key role in hexadecane degradation under methanogentic conditions using DNA-SIP.
Other bacterial T-RFs such as those of 77, 160, 203, 215 and 337 bp were also detected in the 12 C-and 13 C-hexadecane enrichment cultures. Dissimilatory sulfate-reducing bacterium Desulfovibrio africanus DSM 2603 T shows 98% 16S rRNA sequence identity to type clone HB1_5 (accounts for T-RF 77 bp), which exhibits syntrophic oxidation of organic acids under sulphatelimited conditions in co-culture with methanogens, and could use H 2 as an electron donor and assimilate acetate and CO 2 as carbon sources [52]. The Desulfovibrio-like bacterium ACE-8 (GenBank No. JX477133) was isolated from this consortium in our laboratory. No growth occurred after nine months of incubation when cocultured with M. recepatculi ZC-2 using hexadecane as a sole substrate at 35uC (data not shown). This implies that the Desulfovibrio-like organism may metabolize intermediates generated from hexadecane degradation. However, a few clones accounting for T-RF 77 bp also belong to Syntrophaceae, and the utilization of hexadecane for growth by this organism can't be excluded.
Clones (type clone HB1_6) clustered in Synergistaceae shared a T-RF of 160 bp, and has 93% 16S rRNA gene sequence identity to Thermovirga lienii, which degrades proteinaceous substrates, amino acids, and a limited range of organic acids, but not sugars, fatty acids or alcohols. It produces ethanol, acetate, propionate, isovalerate/2-methylbutyrate, H 2 and CO 2 [53]. The Thermotogaceae-related member (type clone HB1_28) representing a T-RF of 203 bp, shares 98% sequence similarity to Mesotoga prima and Candidatus Mesotoga sulfurreducens [54,55], which is a novel phylogenetic lineage in the order Thermotogales and grows on sugars and some proteinaceous substrates with acetate as a major fermentation product. The Spirochaetaceae-affiliated member (mainly accounts for T-RF 215 bp) exhibits 88% 16S rRNA sequence identity with the homoacetogenic Treponema primitia ZAS-2 T [56], which has the physiological potential to utilize H 2 and CO 2 as alternative energy sources, and possibly possesses the capacity to compete with hydrogenotrophic methanogens. Uncultured WWE1 bacterium (T-RF 337 bp) always predominated in the methanogenic hexadecane-degrading consortium during transfer incubations [29], became less abundant after HMN addition as a carrier of hexadecane. Genome analysis of Candidatus Cloacamonas acidaminovorans, an uncultured representative bacterium of the WWE1 bacterium, revealed the potential of oxidation of propionate into acetate and carbon dioxide [57]. From the wide distribution but low abundance of members of Desulfovibrionaceae, Thermotogaceae, Spirochaetaceae, Synergistaceae and uncultured WWE1 bacterium in this study, we proposed that the ecophysiological role of these phylotypes may be associated with the utilization of intermediates from hexadecane degradation and even dead biomass, and were by far not as important as the members of the Syntrophaceae counterparts detected in our SIP incubation. However, their ecological significance in oil-impacted environments should be further investigated due to their wide distribution [9,10,58,59,60]. The Desulfobacteraceae-affiliated member (4 in 81 clones) detected in ''heavy'' clone library (Table 1), was most closely related to Desulfatibacillum alkenivorans, a known alkane-degrader under sulphate-reducing condition, which could degrade hexadecane in co-culture with hydrogenotrophic methanogen [28]. However, the corresponding T-RF 209 bp was not detected with high abundance (,1%) through T-RFLP analysis during SIP incubation, suggesting their minor role in hexadecane conversion.
The archaeal community with 13 C-labeled rRNA genes in heavy fractions only consisted of M. receptaculi-affiliated organisms (T-RF 186 bp). M. receptaculi, which was isolated from Shengli oilfield, is a hydrogenotrophic methanogen that uses hydrogen and carbon dioxide as energy and carbon sources, grows best at 55uC, and demonstrates methanogenic activity under mesophilic conditions [33]. qPCR assay also confirmed that the increase of Methanoculles species is much higher than member s of Methanosaetaceae (Figs. S4B and 4C). The results demonstrate that, the principal pathway of methane production may be mainly through carbon dioxide reduction in our incubation of hexadecane degradation. However, a certain abundance of Methanosaeta species in hexadecane incubation may suggest that acetoclastic methanogens use small amounts of acetate from 13 C-hexadecane. This is consistent with the findings of Gray et al., 2011 who demonstrated that both acetoclastic and CO 2 -reducing methanogens were enriched during methanogenic degradation of crude oil alkanes, but CO 2 -reducing methanogens were enriched to a greater extent.
Moreover this is what would be expected if syntrophic acetate oxidation was an important process in methanogenic alkane degradation as previously suggested [20] This is also consistent with modeling of gas isotopic composition in biodegraded petroleum reservoirs that suggested that most, but not all methanogenic hydrocarbon degradation was channeled through the CO 2 reduction pathway [7].
Methanogenic degradation of complex compounds is a common process under electron-acceptor limited conditions, and the ability of bacteria and methanogenic archaea to degrade hydrocarbons collectively via syntrophic cooperation is definitely necessary [19]. In the present study, so-far uncultured members of Syntrophaceae were shown to play a vital role in hexadecane degradation, mainly coupled with hydrogenotrophic M. receptaculirelated archaea. Figure S1 The T-RFLP profiles at different days of incubation. A: bacterial domain; B: archaeal domain. 12 Chexadecane: the consortium amended with unlabeled hexadecane; 13 C-hexadecane: the consortium amended with 13 C-labeled hexadecane; hexadecane (2): the controls without hexadecane addition, error bars indicate standard deviations from three triplicates.

Supporting Information
(TIF) Figure S2 Rarefaction curves constructed from two bacterial 16S rRNA gene libraries based on OTU cutoff of equal or higher 97%. 12 C-hexadecane: the consortium amended with unlabeled hexadecane; 13 C-hexadecane: the consortium amended with 13  Genomic DNA retrieved from 13C-hexadecane microcosm at day 198; 13 C-218 days (m): Genomic DNA retrieved from 13 Chexadecane microcosm at day 218; 12 C-126 days (g): Genomic DNA retrieved from 12 C-hexadecane microcosm at day 126; 12 C-166 days (e): Genomic DNA retrieved from 12 C-hexadecane microcosm at day 166. We purified genomic DNA by washing twice the filtration column (first: 50 mL; second: 30 mL) when using Promega Wizard DNA cleanup system (Promega, USA). The entire first washed DNA sampled from 13 C-hexadecane microcosm at day 126 was used for isopycnic centrifugation, and the second washed DNA was used for T-RFLP analysis. So the data for quantitative analysis of gene copies of 13 C-hexadecane microcosm at day 126 was not shown, error bars indicate standard deviations from three triplicates. (TIF)