- Email: [email protected]
Insights into the fate of a 13C labelled phenol pulse for stable isotope probing (SIP) experiments
Journal of Microbiological Methods 69 (2007) 340 – 344 www.elsevier.com/locate/jmicmeth
Insights into the fate of a 13 C labelled phenol pulse for stable isotope probing (SIP) experiments Mike Manefield a,⁎, Rob Griffiths b , Niall P. McNamara c , Darren Sleep c , Nick Ostle c , Andrew Whiteley b a
c
Centre for Marine Biofouling and Bioinnovation, University of New South Wales, Sydney, 2052, Australia b Centre for Ecology and Hydrology, Oxford, United Kingdom Centre for Ecology and Hydrology, Lancaster, Library Avenue, Bailrigg, Lancaster, LA1 4AP, United Kingdom Received 29 November 2006; received in revised form 22 January 2007; accepted 22 January 2007 Available online 21 February 2007
Abstract Stable isotope probing (SIP) using DNA or RNA as a biomarker has proven to be a useful method for attributing substrate utilisation to specific microbial taxa. In this study we followed the transfer of a 13C6-phenol pulse in an activated sludge micro-reactor to examine the resulting distribution of labelled carbon in the context of SIP. Most of the added phenol was metabolically converted within the first 100 min after 13C6phenol addition, with 49% incorporated into microbial biomass and 6% respired as CO2. Less than 1% of the total 13C labelled carbon supplied was incorporated into microbial RNA and DNA, with RNA labelling 6.5 times faster than DNA. The remainder of the added 13C was adsorbed and/or complexed to suspended solids within the sludge. The 13C content of nucleic acids increased beyond the initial consumption of the 13Cphenol pulse. This study confirms that RNA labels more efficiently than DNA and reveals that only a small proportion of a pulse is incorporated into nucleic acids. Evidence of continued 13C incorporation into nucleic acids suggests that cross-feeding of the SIP substrate was rapid. This highlights both the benefits of using a biomarker that is rapidly labelled and the importance of sampling within appropriate timescales to avoid or capture the effects of cross-feeding, depending on the goal of the study. © 2007 Elsevier B.V. All rights reserved. Keywords: Microbial biomass; DNA; RNA; Carbon; Respiration; CO2
1. Introduction One of the great challenges in microbiology lies in relating a vast genetic diversity to the wealth of metabolic activity encoded therein. Indeed, it has been a longstanding ambition for microbiologists to identify microbes responsible for particular environmental processes. This is no more apparent than in the case of activated sludge communities responsible for the degradation or sequestration of the myriad of environmentally deleterious compounds present in municipal or industrial wastewaters (Watanabe and Hino, 1996). The assignation of functions in wastewater treatment, such as the consumption and mineralisation of xenobiotic compounds, to microbial taxa, is a fundamental step in the rational en⁎ Corresponding author. Tel.: +61 293858287; fax: +61 293851779. E-mail address: [email protected] (M. Manefield). 0167-7012/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.mimet.2007.01.019
gineering of microbial communities for wastewater treatment (Curtis et al., 2003). Knowing which organisms are responsible for different processes in wastewater treatment enables educated bioaugmentation strategies to be developed if a reactor is performing poorly (Watanabe et al., 2002). Over the past decade, the sophistication of the microbial ecologists molecular toolbox has increased including the development of methods that identify specific microbial phylogenetic groups involved in biochemical transformations of applied and environmental interest. Of these methods, variations on the theme of Stable Isotope Probing (SIP) are widely appreciated as the key to unlocking the infamous microbial black box as evidenced by their broad application (Whiteley et al., 2006). SIP was originally conceived as a means of tracing 13C labelled compounds into the polar lipid derived fatty acids (PLFAs) of phylogenetic groups consuming these compounds (Boschker et al., 1998). SIP methodology then went through
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
two major modifications, the first innovation being the replacement of PLFAs with DNA as the target biomarker (Radajewski et al., 2000) and the second involving the replacement of DNA with the 16S rRNA molecule (Manefield et al., 2002a). The use of nucleic acids as a biomarker harnesses the superior phylogenetic resolution offered by nucleotide base sequences. The use of RNA makes use of the fact that this molecule is produced independently of cellular replication, so the activity of non-replicating cells can be detected. To date there are a limited number of applications of SIP to wastewater treatment. The first applications used RNA based SIP to identify bacteria responsible for phenol degradation in industrial wastewater treatment plants treating coking effluents generated by the steel manufacturing industry (Manefield et al., 2002a, 2005). The only other applications of SIP to wastewater treatment involved the identification of active methylotrophs and acetate oxidisers in denitrifying enrichment cultures and activated sludge respectively derived from a municipal treatment plant using DNA as a biomarker (Ginige et al., 2004, 2005). One of the limitations of DNA and RNA based SIP is the need to incorporate substantial amounts of stable isotope atoms into DNA and RNA to facilitate the density based separation of labelled nucleic acids from that which is unlabelled by centrifugation (Manefield et al., 2002b). Whilst it has been demonstrated that RNA labels more efficiently than DNA both in vitro and in vivo (Manefield et al., 2002a; Ostle et al., 2003), most researchers increase labelled substrate levels beyond naturally occurring concentrations and incubate for extended periods of time. Both of these actions increase the chances of labelling non-target organisms through trophic interactions or cross-feeding. In this study we wanted to examine the potential for microbial cross-feeding and to determine appropriate sampling timescales for SIP. This was achieved using a 13C6-phenol pulse into an activated sludge bioreactor with measurements of the transfer of 13 C from the pulse into microbial biomass, DNA and RNA, and respired as carbon dioxide (CO2). Supplementary measurements of phenol and oxygen concentrations were also made. 2. Materials and methods 2.1. Phenol pulse conditions and sampling regime Activated sludge was collected from a Vitox® reactor (Basin 1) at an industrial wastewater treatment plant (Corus Steelworks, Scunthorpe, UK) and transported immediately back to the laboratory at room temperature. Four 800 ml reactors containing 500 ml of activated sludge each were fed with 5 ml of inlet liquor from the treatment plant and incubated overnight at 30 °C. Oxygenation of the reactors was facilitated by pumping CO2 free air through air stones into the reactors at 500 ml/min. Twenty-four hours after sample collection the reactors were again fed with 5 ml of inlet liquor and 1 mM phenol. The bioreactors were maintained at 25 °C throughout the incubation and sampling periods. Three reactors were then fed with 1 mM 13C6-phenol, whilst the fourth was fed with 1 mM unlabelled phenol. Samples (1 ml) were taken for quantification of phenol every 10 min over
341
100 min. Duplicate 1 ml activated sludge samples were taken from each reactor every 10 min over 100 min for quantification of the 13C content of DNA, RNA and biomass. The biomass in these samples was pelleted and washed twice in chilled filtered supernatant from the Vitox® reactor before being stored at − 20 °C. Headspace gas samples were taken every 5 min over 100 min for quantification of CO2 and 13C–CO2. The fourth reactor was used to monitor the relative oxygen concentration in the activated sludge during the consumption of the phenol pulse using a dissolved oxygen probe. 2.2. Nucleic acid extraction Nucleic acids were extracted as previously described (Manefield et al., 2002a) using bead beating, phenol/chloroform and Qiagen Rneasy minikit protocols. DNA was separated from RNA samples by Qiagen RNA/DNA mini kit protocols (Qiagen). DNA and RNA were quantified and checked for purity using ultramicro volume cuvettes in a GeneQuant pro RNA/DNA calculator (Amersham Pharmacia Biotech). 2.3.
13
C/12C isotope ratio mass spectrometry (IRMS) analysis
DNA and RNA samples were prepared for isotope ratio mass spectrometry (IRMS) analysis by cutting with glucose as previously described by Manefield et al. (2002b) to ensure sufficient material for IRMS. Nucleic acid and biomass samples were placed into 6 × 4 mm tin cups (Elemental Microanalysis, Okehampton, UK) and freeze-dried for 16 h (Christ, Germany) before being analysed for 13C content by continuous flow IRMS at the NERC Life Sciences Mass Spectrometry Facility located at CEH Lancaster UK, using an Elemental Analyser (Carlo Erba, Milano, Italy) linked to a modified Dennis Leigh Technology IRMS (Provac Services, Cheshire, UK). 2.4. CO2 quantification and isotope determination Headspace gas samples (8 cm3) were taken every 5 min and transferred into 3.6 cm3 Exetainers (Labco Ltd., UK) via a septum port. CO2 concentrations were determined using a Perkin Elmer Autosystem XL Gas Chromatograph fitted with a flame ionisation detector fitted with a methaniser operated at 350 °C. CO2 was separated from the gas sample isothermally on a 2 m Poropak Q packed column at 40 °C, with N2 as the carrier gas flowing at 30 cm3 min− 1. 12C/13C ratios were quantified by using a Trace Gas Preconcentrator coupled to an Isoprime IRMS (GV Instruments, UK) at CEH Lancaster. CO2 and 13C concentrations were interpolated by means of a cubic spline model. Amount of isotope found in respired CO2 was calculated by means of a binary mass balance mixing model (e.g. Ostle et al., 2003). 3. Results 3.1.
13
C-phenol oxidation
Fig. 1 illustrates the phenol concentration in each of the triplicate reactors over time. Despite best efforts to treat each
342
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
1 mM. Ultimately, this analysis accounts for 72% of the 13C from the pulse. The incorporation of 13C from phenol into RNA and DNA was less substantial than incorporation into biomass but most rapid in the first 60 min after 13C6-phenol labelling (Fig. 2). As observed previously, the rate of incorporation of 13C into RNA was more rapid at 0.0026 at.% per minute compared with the rate of incorporation into DNA at 0.0004 at.% per minute, representing a 6.5 fold difference. The average RNA yield from 1 ml activated sludge samples was 15 μg. Assuming a 50 % extraction efficiency, the total amount of carbon in RNA in the reactors was 5.4 mg. The 0.2% increase in 13C at.% that occurred over the 100 min incubation therefore represents the incorporation of approximately 10 μg of carbon from the pulse equating to less than 0.1% of the 13C added. Based on the observation that the average DNA yield was similar to the average RNA yield and on the fact that bacterial cells contain more RNA than DNA (six times more in Escherichia coli cells growing exponentially) the proportion of the carbon from the pulse that was incorporated into DNA will be less than 0.1%. Fig. 1. (A) Phenol degradation over time in replicate activated sludge samples. Replicate 1 (squares), replicate 2 (circles) and replicate 3 (triangles). (B) Relative oxygen concentration throughout the consumption of the phenol pulse.
reactor in an identical fashion the rate of phenol consumption differed. The 1 mM pulse representing 36.2 mg of carbon, was consumed rapidly, within 40 min for one reactor and within 60 min for another. The rate of phenol consumption appeared to accelerate after approximately 30 min. Fig. 1 also illustrates the relative concentration of oxygen in the reactors during the phenol pulse. A rapid consumption of oxygen was associated with addition of the phenol pulse with the concentration reaching its lowest level over 20 min and reaching a plateau close to pre-pulse concentrations after 60 min when the phenol had been consumed. 3.2.
13
3.3.
13
C incorporation into CO2
The CO2 concentration in the headspace of the reaction vessels followed a similar pattern between triplicates. It rose steadily during the first 35 min of the pulse at a rate of 4.5 μg/ml/min from a basal level of 390 μg/ml (Fig. 3). From this point onwards the CO2 concentration rose steeply at a rate of 24.5 μg/ml/min until the pulse was consumed. This corresponds with the period of rapid phenol consumption observed. The CO2 concentration then fell rapidly to reach a concentration approximately 100 μg/ml higher than the basal level after 70 min, suggestive of a priming affect from the 13C6-phenol pulse. The 13C at.% of the CO2 in the headspace from the triplicate reactors displayed a considerable amount of variability (Fig. 3). In replicate one, the proportion of carbon that came from the
C incorporation into biomass and nucleic acids
Fig. 2 illustrates increases in 13C at.% in the activated sludge biomass, DNA and RNA, during consumption of the labelled phenol pulse in a single representative reactor (Replicate 2). Before the pulse, the activated sludge biomass had a 13C content of 1.09 at.%. Immediately after the pulse the biomass already displayed an elevated 13C content of 1.35 at.% indicating that some 13C6-phenol had adsorbed to organic matter and the reactor biomass. As the pulse was consumed, the 13C content of the biomass rose to a plateau of 1.90 at.% around 50 min, corresponding well with the point at which the phenol pulse was completely consumed. Given that the dry weight of the sludge was 13.72 mg/ml and that the carbon content of the biomass was 47%, the observed increase in 13C at.% (0.55% without adsorbed carbon) indicates that 49% of the carbon from the phenol pulse was incorporated into biomass within 100 min. Interestingly, 23% of the carbon from the phenol pulse appears to have adsorbed to the biomass, potentially explaining why the phenol concentrations immediately after addition were less than
Fig. 2. Increases in 13C at.% over time in (A) total biomass and (B) DNA (triangles) and RNA (circles) for replicate 2.
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
Fig. 3. Conversion of 13C stable isotope labelled phenol to carbon dioxide by microbial respiration within the activated sludge. (A) Carbon dioxide concentrations in ppm in the headspace of replicate reaction vessels. (B) Proportion of the carbon dioxide molecules in the headspace of replicate reaction vessels that is 13C as a percentage. (C) Extrapolation of the amount of carbon as carbon dioxide in the headspace of replicate reaction vessels derived from the stable isotope labelled phenol pulse. Replicate 1 (squares), replicate 2 (circles) and replicate 3 (triangles).
phenol pulse that was incorporated into CO2 rose relatively steeply until the pulse was consumed and then reached a plateau at 33 13C at.%. In replicate two, the level of CO2 derived from the pulse hovered between 3 and 10 13C at.% with no obvious peak. In replicate three, the proportion of labelled carbon in the headspace reached a peak of 22 13C at.% at the point of consumption of the pulse and thereafter dropped to 7 13C at.% after 100 min. This variability was further evident in an extrapolation of the amount of carbon derived from the pulse that was incorporated into CO2 (Fig. 3). The percentage of the carbon from the pulse that was lost from the system due to respiration was 9.1, 2.9 and 5.0% for replicates 1, 2 and 3, respectively, generating an average of 5.7%. 4. Discussion Stable isotope probing exploiting DNA or RNA as a biomarker has now been applied in over 35 studies seeking to link biological processes to specific taxa in complex microbial communities. The technique has been applied successfully in both ecological studies aiming to identify key organisms in biogeochemical cycles (Ostle et al., 2003) and biotechnological studies aiming to identify organisms of use in wastewater
343
treatment and site bioremediation (Whiteley et al., 2006). In this study, the fate of a stable isotope probing pulse was characterised to quantify the proportion of the pulse that is incorporated into nucleic acids and to gain insight into the most appropriate sampling regime for nucleic acid extraction. Immediately after addition of the 1 mM 13C6-phenol pulse, only 59% of this was detectable in the cell free supernatant by the colorimetric assay employed. This suggests that 41% of the pulse adsorbed to biomass and/or surfaces in the reaction vessel. The biomass data suggest that 23% of the pulse adsorbed to the activated sludge biomass indicating by subtraction that 18% adsorbed to reactor surfaces. IRMS analysis revealed that 49% of 13C from the pulse was incorporated into biomass and 6% of carbon from the pulse was respired as CO2, accounting in total for 96% of the pulse. Activated sludge supernatants were not analysed for 13C content, so it is unclear what proportion of the pulse was converted into soluble extracellular CO2, degradation intermediates and metabolites, but it is likely to be less than 4%. IRMS analysis of DNA and RNA samples taken throughout the pulse period confirmed that RNA labels at a higher rate than DNA but that only a fraction of a percent of the carbon added was incorporated into nucleic acids. Given that RNA accounts for 20% of the dry weight of an exponentially growing E. coli cell (Neidhart and Umbarger, 1996) and that approximately half of the carbon from the phenol pulse was incorporated into biomass, the finding that less than 0.1% of the carbon was found in RNA was surprising. The lower than expected level of incorporation into nucleic acids is likely indicative of the specificity of carbon flow into functional microbes which may make up only a small proportion of the entire community. Unlike incorporation into microbial biomass, the 13C content of RNA and DNA continued to rise after the phenol pulse was consumed (i.e. after 60 min). This is likely due to nucleotide synthesis drawing on the labelled pool of carbon within the biomass rather than on phenol directly. This finding indicates that sampling regimes for SIP analysis should not extend beyond the period of pulse consumption if primary feeding is of interest. In studies aiming to track the transfer of carbon to secondary feeders through cross-feeding, sampling beyond the consumption of the pulse is recommended. The CO2 generated directly from the phenol pulse accounted for between 3 and 9% of the carbon from the pulse, with an average of approximately 6%. Based on the meta cleavage pathway transformation of phenol to pyruvate and acetaldeyhyde, the metabolism of 49% of the pulse would generate approximately 11 mg of CO2. This represents approximately 3 mg of carbon or 8% of the carbon from the pulse in good agreement with the figure obtained empirically. Padmanabhan et al. (2003) recovered 23% of carbon from a labelled phenol pulse into soil after 48 h. Similarly, DeRito et al. (2005) recovered 18% of carbon from a labelled phenol pulse into soil after 30 h. Given that the ratio of 13CO2 to 12CO2 was still elevated at the end of the 100 min monitoring period of the experiment conducted here continued monitoring would reveal additional evolution of 13CO2. From a biochemical perspective, this can be accounted for by the conversion of pyruvate into acetyl-CoA.
344
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
When seeking to identify micro-organisms responsible for processing substrates in natural environments by stable isotope probing the objective is to label RNA or DNA to the highest extent possible whilst minimising the transfer of carbon to secondary assimilators through cross-feeding. This study confirms that RNA labels more efficiently than DNA and reveals that only a small proportion of a pulse is incorporated into nucleic acids and that this incorporation continues after a pulse has been consumed. Along with the need to use realistic concentrations of labelled substrates, this highlights the delicacy of the challenge in applying substrate pulses for SIP and further emphasises the benefits of using a biomarker that is rapidly labelled. References Boschker, H.T.S., Nold, S.C., Wellsbury, P., Bos, D., de Graaf, W., Pel, R., Parkes, R.J., Cappenberg, T.E., 1998. Direct linking of microbial populations to specific biogeochemical processes by C-13-labelling of biomarkers. Nature 392, 801–805. Curtis, T.P., Head, I.M., Graham, D.W., 2003. Theoretical ecology for engineering biology. Environ. Sci. Technol. 37, 64–70. DeRito, C.M., Pumphrey, G.M., Madsen, E.L., 2005. Use of field-based stable isotope probing to identify adapted populations and track carbon flow through a phenol-degrading soil microbial community. Appl. Environ. Microbiol. 71, 7858–7865. Ginige, M.P., Hugenholtz, P., Daims, H., Wagner, M., Keller, J., Blackall, L.L., 2004. Use of stable-isotope probing, full-cycle rRNA analysis, and fluorescence in situ hybridization-microautoradiography to study a methanol-fed denitrifying microbial community. Appl. Environ. Microbiol. 70, 588–596. Ginige, M.P., Keller, J., Blackall, L.L., 2005. Investigation of an acetate-fed denitrifying microbial community by stable isotope probing, full-cycle
rRNA analysis, and fluorescent in situ hybridization-microautoradiography. Appl. Environ. Microbiol. 71, 8683–8691. Manefield, M., Whiteley, A.S., Griffiths, R.I., Bailey, M.J., 2002a. RNA stable isotope probing, a novel means of linking microbial community function to phylogeny. Appl. Environ. Microbiol. 68, 5367–5373. Manefield, M., Whiteley, A.S., Ostle, N., Ineson, P., Bailey, M.J., 2002b. Technical considerations for RNA based stable isotope probing: an approach to associating microbial diversity with microbial community function. Rapid Commun. Mass Spectrom. 16, 2179–2183. Manefield, M., Griffiths, R.I., Leigh, M.B., Fisher, R., Whiteley, A.S., 2005. Functional and compositional comparison of two activated sludge communities remediating coking effluent. Environ. Microbiol. 7, 715–722. Neidhart, F.C., Umbarger, H.E., 1996. Chemical composition of Escherichia coli. In: Neidhart, F.C. (Ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. ASM Press, Washington, DC, pp. 13–16. Ostle, N., Whiteley, A.S., Bailey, M.J., Sleep, D., Ineson, P., Manefield, M., 2003. Active microbial RNA turnover in a grassland soil estimated using a 13 CO2 spike. Soil Biol. Biochem. 35, 877–885. Padmanabhan, P., Padmanabhan, S., DeRito, C., Gray, A., Gannon, D., Snape, J.R., Tsai, C.S., Park, W., Jeon, C., Madsen, E.L., 2003. Respiration of 13 C-labeled substrates added to soil in the field and subsequent 16S rRNA gene analysis of 13C-labeled soil DNA. Appl. Environ. Microbiol. 69, 1614–1622. Radajewski, S., Ineson, P., Parekh, N.R., Murrell, J.C., 2000. Stable-isotope probing as a tool in microbial ecology. Nature 403, 646–649. Watanabe, K., Hino, S., 1996. Identification of a functionally important population in phenol-digesting activated sludge with antisera raised against isolated bacterial strains. Appl. Environ. Microbiol. 62, 3901–3904. Watanabe, K., Teramoto, M., Harayama, S., 2002. Stable augmentation of activated sludge with foreign catabolic genes harboured by an indigenous dominant bacterium. Environ. Microbiol. 4, 577–583. Whiteley, A.S., Manefield, M., Lueders, T., 2006. Unlocking the ‘microbial black box’ using RNA-based stable isotope probing technologies. Curr. Opin. Biotechnol. 17, 67–71.
Insights into the fate of a 13 C labelled phenol pulse for stable isotope probing (SIP) experiments Mike Manefield a,⁎, Rob Griffiths b , Niall P. McNamara c , Darren Sleep c , Nick Ostle c , Andrew Whiteley b a
c
Centre for Marine Biofouling and Bioinnovation, University of New South Wales, Sydney, 2052, Australia b Centre for Ecology and Hydrology, Oxford, United Kingdom Centre for Ecology and Hydrology, Lancaster, Library Avenue, Bailrigg, Lancaster, LA1 4AP, United Kingdom Received 29 November 2006; received in revised form 22 January 2007; accepted 22 January 2007 Available online 21 February 2007
Abstract Stable isotope probing (SIP) using DNA or RNA as a biomarker has proven to be a useful method for attributing substrate utilisation to specific microbial taxa. In this study we followed the transfer of a 13C6-phenol pulse in an activated sludge micro-reactor to examine the resulting distribution of labelled carbon in the context of SIP. Most of the added phenol was metabolically converted within the first 100 min after 13C6phenol addition, with 49% incorporated into microbial biomass and 6% respired as CO2. Less than 1% of the total 13C labelled carbon supplied was incorporated into microbial RNA and DNA, with RNA labelling 6.5 times faster than DNA. The remainder of the added 13C was adsorbed and/or complexed to suspended solids within the sludge. The 13C content of nucleic acids increased beyond the initial consumption of the 13Cphenol pulse. This study confirms that RNA labels more efficiently than DNA and reveals that only a small proportion of a pulse is incorporated into nucleic acids. Evidence of continued 13C incorporation into nucleic acids suggests that cross-feeding of the SIP substrate was rapid. This highlights both the benefits of using a biomarker that is rapidly labelled and the importance of sampling within appropriate timescales to avoid or capture the effects of cross-feeding, depending on the goal of the study. © 2007 Elsevier B.V. All rights reserved. Keywords: Microbial biomass; DNA; RNA; Carbon; Respiration; CO2
1. Introduction One of the great challenges in microbiology lies in relating a vast genetic diversity to the wealth of metabolic activity encoded therein. Indeed, it has been a longstanding ambition for microbiologists to identify microbes responsible for particular environmental processes. This is no more apparent than in the case of activated sludge communities responsible for the degradation or sequestration of the myriad of environmentally deleterious compounds present in municipal or industrial wastewaters (Watanabe and Hino, 1996). The assignation of functions in wastewater treatment, such as the consumption and mineralisation of xenobiotic compounds, to microbial taxa, is a fundamental step in the rational en⁎ Corresponding author. Tel.: +61 293858287; fax: +61 293851779. E-mail address: [email protected] (M. Manefield). 0167-7012/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.mimet.2007.01.019
gineering of microbial communities for wastewater treatment (Curtis et al., 2003). Knowing which organisms are responsible for different processes in wastewater treatment enables educated bioaugmentation strategies to be developed if a reactor is performing poorly (Watanabe et al., 2002). Over the past decade, the sophistication of the microbial ecologists molecular toolbox has increased including the development of methods that identify specific microbial phylogenetic groups involved in biochemical transformations of applied and environmental interest. Of these methods, variations on the theme of Stable Isotope Probing (SIP) are widely appreciated as the key to unlocking the infamous microbial black box as evidenced by their broad application (Whiteley et al., 2006). SIP was originally conceived as a means of tracing 13C labelled compounds into the polar lipid derived fatty acids (PLFAs) of phylogenetic groups consuming these compounds (Boschker et al., 1998). SIP methodology then went through
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
two major modifications, the first innovation being the replacement of PLFAs with DNA as the target biomarker (Radajewski et al., 2000) and the second involving the replacement of DNA with the 16S rRNA molecule (Manefield et al., 2002a). The use of nucleic acids as a biomarker harnesses the superior phylogenetic resolution offered by nucleotide base sequences. The use of RNA makes use of the fact that this molecule is produced independently of cellular replication, so the activity of non-replicating cells can be detected. To date there are a limited number of applications of SIP to wastewater treatment. The first applications used RNA based SIP to identify bacteria responsible for phenol degradation in industrial wastewater treatment plants treating coking effluents generated by the steel manufacturing industry (Manefield et al., 2002a, 2005). The only other applications of SIP to wastewater treatment involved the identification of active methylotrophs and acetate oxidisers in denitrifying enrichment cultures and activated sludge respectively derived from a municipal treatment plant using DNA as a biomarker (Ginige et al., 2004, 2005). One of the limitations of DNA and RNA based SIP is the need to incorporate substantial amounts of stable isotope atoms into DNA and RNA to facilitate the density based separation of labelled nucleic acids from that which is unlabelled by centrifugation (Manefield et al., 2002b). Whilst it has been demonstrated that RNA labels more efficiently than DNA both in vitro and in vivo (Manefield et al., 2002a; Ostle et al., 2003), most researchers increase labelled substrate levels beyond naturally occurring concentrations and incubate for extended periods of time. Both of these actions increase the chances of labelling non-target organisms through trophic interactions or cross-feeding. In this study we wanted to examine the potential for microbial cross-feeding and to determine appropriate sampling timescales for SIP. This was achieved using a 13C6-phenol pulse into an activated sludge bioreactor with measurements of the transfer of 13 C from the pulse into microbial biomass, DNA and RNA, and respired as carbon dioxide (CO2). Supplementary measurements of phenol and oxygen concentrations were also made. 2. Materials and methods 2.1. Phenol pulse conditions and sampling regime Activated sludge was collected from a Vitox® reactor (Basin 1) at an industrial wastewater treatment plant (Corus Steelworks, Scunthorpe, UK) and transported immediately back to the laboratory at room temperature. Four 800 ml reactors containing 500 ml of activated sludge each were fed with 5 ml of inlet liquor from the treatment plant and incubated overnight at 30 °C. Oxygenation of the reactors was facilitated by pumping CO2 free air through air stones into the reactors at 500 ml/min. Twenty-four hours after sample collection the reactors were again fed with 5 ml of inlet liquor and 1 mM phenol. The bioreactors were maintained at 25 °C throughout the incubation and sampling periods. Three reactors were then fed with 1 mM 13C6-phenol, whilst the fourth was fed with 1 mM unlabelled phenol. Samples (1 ml) were taken for quantification of phenol every 10 min over
341
100 min. Duplicate 1 ml activated sludge samples were taken from each reactor every 10 min over 100 min for quantification of the 13C content of DNA, RNA and biomass. The biomass in these samples was pelleted and washed twice in chilled filtered supernatant from the Vitox® reactor before being stored at − 20 °C. Headspace gas samples were taken every 5 min over 100 min for quantification of CO2 and 13C–CO2. The fourth reactor was used to monitor the relative oxygen concentration in the activated sludge during the consumption of the phenol pulse using a dissolved oxygen probe. 2.2. Nucleic acid extraction Nucleic acids were extracted as previously described (Manefield et al., 2002a) using bead beating, phenol/chloroform and Qiagen Rneasy minikit protocols. DNA was separated from RNA samples by Qiagen RNA/DNA mini kit protocols (Qiagen). DNA and RNA were quantified and checked for purity using ultramicro volume cuvettes in a GeneQuant pro RNA/DNA calculator (Amersham Pharmacia Biotech). 2.3.
13
C/12C isotope ratio mass spectrometry (IRMS) analysis
DNA and RNA samples were prepared for isotope ratio mass spectrometry (IRMS) analysis by cutting with glucose as previously described by Manefield et al. (2002b) to ensure sufficient material for IRMS. Nucleic acid and biomass samples were placed into 6 × 4 mm tin cups (Elemental Microanalysis, Okehampton, UK) and freeze-dried for 16 h (Christ, Germany) before being analysed for 13C content by continuous flow IRMS at the NERC Life Sciences Mass Spectrometry Facility located at CEH Lancaster UK, using an Elemental Analyser (Carlo Erba, Milano, Italy) linked to a modified Dennis Leigh Technology IRMS (Provac Services, Cheshire, UK). 2.4. CO2 quantification and isotope determination Headspace gas samples (8 cm3) were taken every 5 min and transferred into 3.6 cm3 Exetainers (Labco Ltd., UK) via a septum port. CO2 concentrations were determined using a Perkin Elmer Autosystem XL Gas Chromatograph fitted with a flame ionisation detector fitted with a methaniser operated at 350 °C. CO2 was separated from the gas sample isothermally on a 2 m Poropak Q packed column at 40 °C, with N2 as the carrier gas flowing at 30 cm3 min− 1. 12C/13C ratios were quantified by using a Trace Gas Preconcentrator coupled to an Isoprime IRMS (GV Instruments, UK) at CEH Lancaster. CO2 and 13C concentrations were interpolated by means of a cubic spline model. Amount of isotope found in respired CO2 was calculated by means of a binary mass balance mixing model (e.g. Ostle et al., 2003). 3. Results 3.1.
13
C-phenol oxidation
Fig. 1 illustrates the phenol concentration in each of the triplicate reactors over time. Despite best efforts to treat each
342
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
1 mM. Ultimately, this analysis accounts for 72% of the 13C from the pulse. The incorporation of 13C from phenol into RNA and DNA was less substantial than incorporation into biomass but most rapid in the first 60 min after 13C6-phenol labelling (Fig. 2). As observed previously, the rate of incorporation of 13C into RNA was more rapid at 0.0026 at.% per minute compared with the rate of incorporation into DNA at 0.0004 at.% per minute, representing a 6.5 fold difference. The average RNA yield from 1 ml activated sludge samples was 15 μg. Assuming a 50 % extraction efficiency, the total amount of carbon in RNA in the reactors was 5.4 mg. The 0.2% increase in 13C at.% that occurred over the 100 min incubation therefore represents the incorporation of approximately 10 μg of carbon from the pulse equating to less than 0.1% of the 13C added. Based on the observation that the average DNA yield was similar to the average RNA yield and on the fact that bacterial cells contain more RNA than DNA (six times more in Escherichia coli cells growing exponentially) the proportion of the carbon from the pulse that was incorporated into DNA will be less than 0.1%. Fig. 1. (A) Phenol degradation over time in replicate activated sludge samples. Replicate 1 (squares), replicate 2 (circles) and replicate 3 (triangles). (B) Relative oxygen concentration throughout the consumption of the phenol pulse.
reactor in an identical fashion the rate of phenol consumption differed. The 1 mM pulse representing 36.2 mg of carbon, was consumed rapidly, within 40 min for one reactor and within 60 min for another. The rate of phenol consumption appeared to accelerate after approximately 30 min. Fig. 1 also illustrates the relative concentration of oxygen in the reactors during the phenol pulse. A rapid consumption of oxygen was associated with addition of the phenol pulse with the concentration reaching its lowest level over 20 min and reaching a plateau close to pre-pulse concentrations after 60 min when the phenol had been consumed. 3.2.
13
3.3.
13
C incorporation into CO2
The CO2 concentration in the headspace of the reaction vessels followed a similar pattern between triplicates. It rose steadily during the first 35 min of the pulse at a rate of 4.5 μg/ml/min from a basal level of 390 μg/ml (Fig. 3). From this point onwards the CO2 concentration rose steeply at a rate of 24.5 μg/ml/min until the pulse was consumed. This corresponds with the period of rapid phenol consumption observed. The CO2 concentration then fell rapidly to reach a concentration approximately 100 μg/ml higher than the basal level after 70 min, suggestive of a priming affect from the 13C6-phenol pulse. The 13C at.% of the CO2 in the headspace from the triplicate reactors displayed a considerable amount of variability (Fig. 3). In replicate one, the proportion of carbon that came from the
C incorporation into biomass and nucleic acids
Fig. 2 illustrates increases in 13C at.% in the activated sludge biomass, DNA and RNA, during consumption of the labelled phenol pulse in a single representative reactor (Replicate 2). Before the pulse, the activated sludge biomass had a 13C content of 1.09 at.%. Immediately after the pulse the biomass already displayed an elevated 13C content of 1.35 at.% indicating that some 13C6-phenol had adsorbed to organic matter and the reactor biomass. As the pulse was consumed, the 13C content of the biomass rose to a plateau of 1.90 at.% around 50 min, corresponding well with the point at which the phenol pulse was completely consumed. Given that the dry weight of the sludge was 13.72 mg/ml and that the carbon content of the biomass was 47%, the observed increase in 13C at.% (0.55% without adsorbed carbon) indicates that 49% of the carbon from the phenol pulse was incorporated into biomass within 100 min. Interestingly, 23% of the carbon from the phenol pulse appears to have adsorbed to the biomass, potentially explaining why the phenol concentrations immediately after addition were less than
Fig. 2. Increases in 13C at.% over time in (A) total biomass and (B) DNA (triangles) and RNA (circles) for replicate 2.
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
Fig. 3. Conversion of 13C stable isotope labelled phenol to carbon dioxide by microbial respiration within the activated sludge. (A) Carbon dioxide concentrations in ppm in the headspace of replicate reaction vessels. (B) Proportion of the carbon dioxide molecules in the headspace of replicate reaction vessels that is 13C as a percentage. (C) Extrapolation of the amount of carbon as carbon dioxide in the headspace of replicate reaction vessels derived from the stable isotope labelled phenol pulse. Replicate 1 (squares), replicate 2 (circles) and replicate 3 (triangles).
phenol pulse that was incorporated into CO2 rose relatively steeply until the pulse was consumed and then reached a plateau at 33 13C at.%. In replicate two, the level of CO2 derived from the pulse hovered between 3 and 10 13C at.% with no obvious peak. In replicate three, the proportion of labelled carbon in the headspace reached a peak of 22 13C at.% at the point of consumption of the pulse and thereafter dropped to 7 13C at.% after 100 min. This variability was further evident in an extrapolation of the amount of carbon derived from the pulse that was incorporated into CO2 (Fig. 3). The percentage of the carbon from the pulse that was lost from the system due to respiration was 9.1, 2.9 and 5.0% for replicates 1, 2 and 3, respectively, generating an average of 5.7%. 4. Discussion Stable isotope probing exploiting DNA or RNA as a biomarker has now been applied in over 35 studies seeking to link biological processes to specific taxa in complex microbial communities. The technique has been applied successfully in both ecological studies aiming to identify key organisms in biogeochemical cycles (Ostle et al., 2003) and biotechnological studies aiming to identify organisms of use in wastewater
343
treatment and site bioremediation (Whiteley et al., 2006). In this study, the fate of a stable isotope probing pulse was characterised to quantify the proportion of the pulse that is incorporated into nucleic acids and to gain insight into the most appropriate sampling regime for nucleic acid extraction. Immediately after addition of the 1 mM 13C6-phenol pulse, only 59% of this was detectable in the cell free supernatant by the colorimetric assay employed. This suggests that 41% of the pulse adsorbed to biomass and/or surfaces in the reaction vessel. The biomass data suggest that 23% of the pulse adsorbed to the activated sludge biomass indicating by subtraction that 18% adsorbed to reactor surfaces. IRMS analysis revealed that 49% of 13C from the pulse was incorporated into biomass and 6% of carbon from the pulse was respired as CO2, accounting in total for 96% of the pulse. Activated sludge supernatants were not analysed for 13C content, so it is unclear what proportion of the pulse was converted into soluble extracellular CO2, degradation intermediates and metabolites, but it is likely to be less than 4%. IRMS analysis of DNA and RNA samples taken throughout the pulse period confirmed that RNA labels at a higher rate than DNA but that only a fraction of a percent of the carbon added was incorporated into nucleic acids. Given that RNA accounts for 20% of the dry weight of an exponentially growing E. coli cell (Neidhart and Umbarger, 1996) and that approximately half of the carbon from the phenol pulse was incorporated into biomass, the finding that less than 0.1% of the carbon was found in RNA was surprising. The lower than expected level of incorporation into nucleic acids is likely indicative of the specificity of carbon flow into functional microbes which may make up only a small proportion of the entire community. Unlike incorporation into microbial biomass, the 13C content of RNA and DNA continued to rise after the phenol pulse was consumed (i.e. after 60 min). This is likely due to nucleotide synthesis drawing on the labelled pool of carbon within the biomass rather than on phenol directly. This finding indicates that sampling regimes for SIP analysis should not extend beyond the period of pulse consumption if primary feeding is of interest. In studies aiming to track the transfer of carbon to secondary feeders through cross-feeding, sampling beyond the consumption of the pulse is recommended. The CO2 generated directly from the phenol pulse accounted for between 3 and 9% of the carbon from the pulse, with an average of approximately 6%. Based on the meta cleavage pathway transformation of phenol to pyruvate and acetaldeyhyde, the metabolism of 49% of the pulse would generate approximately 11 mg of CO2. This represents approximately 3 mg of carbon or 8% of the carbon from the pulse in good agreement with the figure obtained empirically. Padmanabhan et al. (2003) recovered 23% of carbon from a labelled phenol pulse into soil after 48 h. Similarly, DeRito et al. (2005) recovered 18% of carbon from a labelled phenol pulse into soil after 30 h. Given that the ratio of 13CO2 to 12CO2 was still elevated at the end of the 100 min monitoring period of the experiment conducted here continued monitoring would reveal additional evolution of 13CO2. From a biochemical perspective, this can be accounted for by the conversion of pyruvate into acetyl-CoA.
344
M. Manefield et al. / Journal of Microbiological Methods 69 (2007) 340–344
When seeking to identify micro-organisms responsible for processing substrates in natural environments by stable isotope probing the objective is to label RNA or DNA to the highest extent possible whilst minimising the transfer of carbon to secondary assimilators through cross-feeding. This study confirms that RNA labels more efficiently than DNA and reveals that only a small proportion of a pulse is incorporated into nucleic acids and that this incorporation continues after a pulse has been consumed. Along with the need to use realistic concentrations of labelled substrates, this highlights the delicacy of the challenge in applying substrate pulses for SIP and further emphasises the benefits of using a biomarker that is rapidly labelled. References Boschker, H.T.S., Nold, S.C., Wellsbury, P., Bos, D., de Graaf, W., Pel, R., Parkes, R.J., Cappenberg, T.E., 1998. Direct linking of microbial populations to specific biogeochemical processes by C-13-labelling of biomarkers. Nature 392, 801–805. Curtis, T.P., Head, I.M., Graham, D.W., 2003. Theoretical ecology for engineering biology. Environ. Sci. Technol. 37, 64–70. DeRito, C.M., Pumphrey, G.M., Madsen, E.L., 2005. Use of field-based stable isotope probing to identify adapted populations and track carbon flow through a phenol-degrading soil microbial community. Appl. Environ. Microbiol. 71, 7858–7865. Ginige, M.P., Hugenholtz, P., Daims, H., Wagner, M., Keller, J., Blackall, L.L., 2004. Use of stable-isotope probing, full-cycle rRNA analysis, and fluorescence in situ hybridization-microautoradiography to study a methanol-fed denitrifying microbial community. Appl. Environ. Microbiol. 70, 588–596. Ginige, M.P., Keller, J., Blackall, L.L., 2005. Investigation of an acetate-fed denitrifying microbial community by stable isotope probing, full-cycle
rRNA analysis, and fluorescent in situ hybridization-microautoradiography. Appl. Environ. Microbiol. 71, 8683–8691. Manefield, M., Whiteley, A.S., Griffiths, R.I., Bailey, M.J., 2002a. RNA stable isotope probing, a novel means of linking microbial community function to phylogeny. Appl. Environ. Microbiol. 68, 5367–5373. Manefield, M., Whiteley, A.S., Ostle, N., Ineson, P., Bailey, M.J., 2002b. Technical considerations for RNA based stable isotope probing: an approach to associating microbial diversity with microbial community function. Rapid Commun. Mass Spectrom. 16, 2179–2183. Manefield, M., Griffiths, R.I., Leigh, M.B., Fisher, R., Whiteley, A.S., 2005. Functional and compositional comparison of two activated sludge communities remediating coking effluent. Environ. Microbiol. 7, 715–722. Neidhart, F.C., Umbarger, H.E., 1996. Chemical composition of Escherichia coli. In: Neidhart, F.C. (Ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. ASM Press, Washington, DC, pp. 13–16. Ostle, N., Whiteley, A.S., Bailey, M.J., Sleep, D., Ineson, P., Manefield, M., 2003. Active microbial RNA turnover in a grassland soil estimated using a 13 CO2 spike. Soil Biol. Biochem. 35, 877–885. Padmanabhan, P., Padmanabhan, S., DeRito, C., Gray, A., Gannon, D., Snape, J.R., Tsai, C.S., Park, W., Jeon, C., Madsen, E.L., 2003. Respiration of 13 C-labeled substrates added to soil in the field and subsequent 16S rRNA gene analysis of 13C-labeled soil DNA. Appl. Environ. Microbiol. 69, 1614–1622. Radajewski, S., Ineson, P., Parekh, N.R., Murrell, J.C., 2000. Stable-isotope probing as a tool in microbial ecology. Nature 403, 646–649. Watanabe, K., Hino, S., 1996. Identification of a functionally important population in phenol-digesting activated sludge with antisera raised against isolated bacterial strains. Appl. Environ. Microbiol. 62, 3901–3904. Watanabe, K., Teramoto, M., Harayama, S., 2002. Stable augmentation of activated sludge with foreign catabolic genes harboured by an indigenous dominant bacterium. Environ. Microbiol. 4, 577–583. Whiteley, A.S., Manefield, M., Lueders, T., 2006. Unlocking the ‘microbial black box’ using RNA-based stable isotope probing technologies. Curr. Opin. Biotechnol. 17, 67–71.