Edited by: András Táncsics, Szent István University, Hungary
Reviewed by: Krisztián Laczi, University of Szeged, Hungary; Attila Szabó, Eötvös Loránd University, Hungary
This article was submitted to Microbiotechnology, a section of the journal Frontiers in Microbiology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
There are two main strategies known how microorganisms regulate substrate utilization: specialization on one preferred substrate at high concentrations in batch cultures or simultaneous utilization of many substrates at low concentrations in chemostats. However, it remains unclear how microorganisms utilize substrates at low concentrations in the subsurface: do they focus on a single substrate and exhibit catabolite repression or do they de-repress regulation of all catabolic pathways? Here, we investigated the readiness of
The classical textbook knowledge of microbial growth states that microorganisms are only utilizing one substrate at a time while the consumption of other substrates is attenuated by catabolite repression leading to diauxic growth (Monod,
In previous studies with
Our experiments with extremely slow growth rates of 0.002 h−1 in retentostats revealed that
However, it remained unclear if the physiology of microorganisms in natural sediments can be fully simulated by chemostat or retentostat experiments. The difference to environmental conditions is the presence of high bacterial biomass and planktonic growth in chemostats, while natural conditions are characterized by very low biomass and mostly sessile life style in the presence of a mixed community. The main question remains if microorganisms in the environment utilize one substrate at a time or if they accommodate a de-repressed physiological state where all substrates are utilized simultaneously.
Hence, the aim of the current study was to elucidate how microorganisms utilize carbon sources in the environment. As an example of a typical microorganism found in groundwater, we cultivated
The columns were built by the mechanics workshop at Helmholtz Zentrum München. Columns were made from plexiglass, with a length of 50 cm and an inner diameter of 5 cm. Columns were sterilized with 70% ethanol for 1 h and filled with either natural sediment or coarse quartz sand with homogenous grain size of ~1 mm. Hereafter, quartz sand in Column 2 and 3 will be referred as sediment as it was used as a model for the groundwater sediment. Natural sediment was collected from a gravel pit in Bruckmühl near Munich, Germany and sieved with exclusion of grain sizes above 1 mm. The natural sediment and quartz sand were thoroughly washed five times in demineralized water, sterilized three times at 120°C and then dried at 105°C for 1 day. Columns were packed with dry sediment in a clean bench. To ensure homogeneous packing and to prevent air bubbles entrapment, columns were placed into a sonicator bath and mild sonication was applied after each stepwise filling with the sediment. In order to remove any remaining air bubbles, columns were flushed with sterile ultrapure MilliQ® water with a total organic carbon content (<10 μg L−1) at 0.25 ml h−1 for 1 week. Water or medium were injected at the columns bottom via a peristaltic pump. Due to the experimental set up, conditions were not sterile which also led to the colonization of the columns with other microorganisms.
In order to monitor stability of anoxic conditions for cultivation of
After conducting a tracer experiment and establishment of anoxic conditions, columns were equilibrated with sterile medium at 18 ml h−1 for 2 days. Then, ~140 ml of
A bromide tracer experiment was conducted to check the packing and performance of the columns prior to inoculation with bacteria. The columns were flushed with ultrapure water (MilliQ) at 0.3 ml min −1 and 3 ml of a Br− tracer solution (KBr, Sigma Aldrich, St. Louis, Missouri, USA) (70 mg L−1) was injected with a syringe through a septum. The outflowing water was collected in 30 min fractions using a fraction collector. The bromide tracer experiment was performed for Columns 1 and 2 and confirmed that the sediment was packed homogenously (
Benzoate was measured by HPLC (Shimadzu, Japan) using a PFP Kinetex column (75 x 4.6 mm) (Phenomenex Inc., USA). Elution was isocratic with 1% acetic acid in MilliQ water (solvent A) and 1% acetic acid in methanol (solvent B) (50:50, v:v) at a flow rate of 0.7 ml min−1 (UV detection at 236 nm). Column temperature was set to 30°C.
Bromide was analyzed by ion chromatography on a DX-100 (Dionex, Germering, Germany) as described in Stoewer et al. (
Duplicate 0.5 ml aliquots of each sediment fraction were fixed with 2.5% glutardialdehyde and stored at 4°C until further analysis. Sample preparation and flow cytometer analysis were done as described in Anneser et al. (
Sediment collected for proteomics analysis was immediately put at −80°C and stored until further analysis. Proteins were extracted according to a protocol modified after (Taylor and Williams,
Cells from batch cultures were harvested during exponential phase and extracted according to Marozava et al. (
Protein concentration was determined using the Bradford protein assay (Bio-Rad, Munich, Germany) with bovine serum albumin as the standard (Bradford,
Each 10 μg of every column fraction and of the batch sample were digested using a modified filter-aided sample preparation procedure (Wiśniewski et al.,
Samples were measured on a LTQ OrbitrapXL mass spectrometer (Thermo Scientific) online coupled to an Ultimate 3000 nano-RSLC (Dionex) as described previously (Hauck et al.,
Generated raw files were analyzed using Progenesis QI for proteomics (version 2.0, Non-linear Dynamics, part of Waters) for label-free quantification as described (Hauck et al.,
Normalized and log10 transformed protein abundances of
Since
For metagenomic analysis, DNA was extracted from sediments of different column fractions (30 samples in total) as described in Marozava et al. (
Assembly-based metagenomics was performed on DNA samples extracted from the sediments. Nine DNA samples were sent to GATC (Konstanz, Germany) for library preparation and 150-bps paired-end Illumina HiSeq sequencing. Raw reads were processed by trimming and quality filtering with bbduk (
Column 1 (packed with natural sediment) was harvested after 12.5 days of cultivation, while the other two columns (packed with quartz sand also named sediment hereafter) were harvested after 22 (Column 2) and 23 (Column 3) days of cultivation, respectively. Degradation of benzoate in Column 1 produced an almost linear gradient from 0.9 mM at the inlet to non-detectable concentrations at the outlet (
Benzoate concentrations and bacterial abundances in the three sediment columns at the end of the cultivation at 300, 528, and 552 h for columns 1 (squares), 2 (circles), and 3 (triangles), respectively. Concentrations of benzoate
We added an excess of nitrate to ensure that the system is electron donor-limited at any spot in the columns. Hence, concentrations in the columns did not decline significantly, whereas 6 mM nitrate are enough to oxidize 1 mM of benzoate to CO2.
The total bacterial abundances decreased only slightly from the inlet region of 0–10 cm depth to the outlet region of the three columns at 40–50 cm depth. Bacterial abundances ranged from 4·106 to 3·107 cells per g of sediment (Column 1), 7·107 to 7·109 cells per g of sediment (Column 2), and 1·108 to 6·108 cells per g of sediment (Column 3) (
At the end of the cultivation, the proteins were extracted from the sediment including pore water from every 5 cm depth and analyzed by mass spectrometry. Since the percentage of sessile organisms in sediments vastly exceeds the planktonic ones in the pore water, mostly sessile bacteria were analyzed in the metagenomic and metaproteomic analysis. Metagenomic analysis was conducted to assist differentiating
In total, proteins from 87 different bacterial taxa were identified via the metaproteomic approach where peptide search was done against an in house metagenomic database created via metagenomic analysis of the selected column fractions (Data Sheet S1). Several identified taxa are known benzoate degraders affiliated to the genera
3,741 proteins were detected in the total proteome, 1,004 of which belonged to
Non-parametric clustering of all proteins detected at each column depth showed that Column 1 (natural sediment) differed from Columns 2 and 3 (quartz sediment) and all columns showed different protein expression than the batch cultures, which were used as a reference (where
Schematic representation of selected depths for statistical analysis: Different shaded areas on the columns represent different benzoate concentration ranges which were used further for the pairwise comparisons (as indicated in the legend). The columns are represented with the following symbols: columns 1 (squares), 2 (circles), and 3 (triangles), respectively.
The proteomes of the different zones of the columns were compared to the high benzoate zone in Column 1 (~0.9 mM) (
Number of
Most of the predicted proteins for acetate metabolism were detected in the three columns. Acetate kinase (Gmet_1034) showed increased abundances at the column areas with low benzoate concentrations, relative to high (
List of differentially expressed catabolic proteins identified in Bioconductor (please refer to Material and methods) where the following pair-wise comparisons were performed between the column fractions with different benzoate concentrations: “medium” vs “high,” “low” vs “high,” and “below detection” vs “high.”
Q39WV0 Acetate kinase | |||
Q39TP8 Metal-dependent hydrolase, putative | |||
Q39TV9 putative benzoyl-CoA reductase electron transfer protein | |||
Q39UP3 Benzoyl-CoA reductase, putative | |||
Q39TX3 3-hydroxyacyl-CoA dehydrogenase | 1.4 | ||
Q39TV7 6-oxocyclohex-1-ene-1-carbonyl-CoA hydrolase | |||
Q39TP3 Electron transfer flavoprotein, alpha subunit | |||
Q39TP2 Electron transfer flavoprotein, beta subunit | |||
Q39QM3 Ferritin-like domain protein | 1.1 | 0.8 | |
Q39XP3 Sodium/solute symporter family protein | 2.6 | ||
Q39TP4 6-hydroxycyclohex-1-ene-1-carbonyl-CoA dehydrogenase | 1.5 | ||
Q39TP6 Lipoprotein, putative | |||
Q39TZ1 Succinyl:benzoate coenzyme A transferase | |||
Q39TW1 Polyferredoxin, putative benzoyl-CoA reductase | |||
Q39ZG7 ATPase, AAA | |||
Q39TP5 Cyclohexa-1,5-dienecarbonyl-CoA hydratase | 1.3 | 1.0 | |
Q39TV8 Benzoyl-CoA reductase, putative | 1.4 | 1.4 | |
Q39TW5 Benzoyl-CoA reductase electron transfer protein | |||
Q39TW0 Iron-sulfur cluster-binding oxidoreductase | |||
Q39TY0 Electron transfer flavoprotein-associated cytochrome b | 1.1 | ||
Q39TQ2 Benzoate–coenzyme A ligase | 1.1 | ||
Q39VH3 Lipoprotein release ABC transporter, membrane protein | 0.1 | ||
Q39TU7 Phosphotransbutyrylase | 1.5 | ||
Q39TX4 Enoyl-CoA hydratase/isomerase | 1.4 | ||
Q39UY1 Electron transfer flavoprotein, alpha subunit | 0.9 | 2.6 | |
Q39UY5 Acyl-CoA–carboxylate coenzyme A transferase, beta subunit | 0.7 | 0.5 | |
Q39WT9 Aldehyde:ferredoxin oxidoreductase, tungsten-containing | 0.6 | 0.4 | |
Q39WT8 Ethanol dehydrogenase, putative | 0.6 | 0.2 | |
Q39Z23 Acyl-CoA synthetase, AMP-forming | |||
Q39TX1 Thiolase | |||
Q39V92 Malonyl CoA-acyl carrier protein transacylase | |||
Q39T45 3-hydroxyacyl-[acyl-carrier-protein] dehydratase | |||
Q39V89 3-oxoacyl-[acyl-carrier-protein] synthase | |||
Q39TX0 Glutaryl-CoA dehydrogenase | |||
Q39TY6 Oxidoreductase, short-chain dehydrogenase/reductase family | 1.2 | ||
Q39QI9 3-oxoacyl-[acyl-carrier-protein] synthase | 1.8 | ||
Q39SJ7 Enoyl-[acyl-carrier-protein] reductase [NADH] | 0.8 | 0.4 | |
Q39V93 3-oxoacyl-[acyl-carrier-protein] synthase | 0.8 | 0.6 | |
Q39TQ0 Sigma-54-dependent transcriptional response regulator | |||
Q39TQ9 4-hydroxybenzoyl-CoA reductase subunit | 1.0 | 2.2 | |
Q39TQ1 Uncharacterized protein | 0.7 | 1.5 | |
Q39TQ4 Zinc-dependent hydrolase | 0.1 | ||
Q39TW6 Iron-sulfur cluster-binding protein | |||
Q39TU3 Phenylphosphate carboxylase, beta subunit | 1.0 | ||
Q39QL0 (R)-methylmalonyl-CoA mutase, isobutyryl-CoA mutase-like | 1.9 | 1.3 | |
Q39VB8 (R)-methylmalonyl-CoA mutase | 1.1 | 0.4 | |
Q39QK8 Methylmalonyl-CoA epimerase | |||
Q39UG6 Citramalate synthase | |||
Q39WZ8 Acetyl-CoA carboxylase | |||
Q39WZ9 Acetyl-CoA carboxylase | 2.2 | ||
Q39WC1 Acetyl-coenzyme A carboxylase subunit alpha | 1.4 | ||
Q39SS3 Acetyl-coenzyme A carboxylase subunit beta | 0.3 | 0.7 | |
Q39R65 Succinyl:acetate coenzyme A transferase | 1.0 | 1.4 | |
Q39QU2 Phosphoenolpyruvate carboxykinase | |||
Q39RG6 Pyruvate, phosphate dikinase | |||
Q39S03 Pyruvate dehydrogenase complex, E1 protein, beta subunit | 2.1 | ||
Q39V56 Malate oxidoreductase, NADP-dependent | 1.8 | ||
Q39ZF3 Pyruvate kinase | 1.4 | ||
Q39Q43 Pyruvate-flavodoxin oxidoreductase | 1.4 | ||
Q39W73 2-isopropylmalate synthase | 1.3 | ||
Q39XG6 Pyruvate carboxylase | 1.1 | ||
Q39SB7 Phosphoenolpyruvate carboxykinase [GTP] | 0.8 | 1.4 | |
Q39RZ6 Pyruvate dehydrogenase E1 component subunit alpha | |||
Q39RZ2 Dihydrolipoyl dehydrogenase | 3.6 | 0.6 | |
Q39S67 Citrate synthase | |||
Q39VX9 2-oxoglutarate:ferredoxin oxidoreductase, ferredoxin subunit | 3.0 | 0.7 | |
Q39TX7 Succinate–CoA ligase [ADP-forming] subunit alpha | |||
Q39VX8 2-oxoglutarate:ferredoxin oxidoreductase, alpha subunit | |||
Q39VX7 2-oxoglutarate:ferredoxin oxidoreductase | |||
Q39VX6 2-oxoglutarate:ferredoxin oxidoreductase, gamma subunit | 1.3 | 0.8 | |
Q39VY1 Isocitrate dehydrogenase, NADP-dependent | 1.5 | 0.9 | |
Q39T04 Succinate dehydrogenase/fumarate reductase | |||
Q39T03 Succinate dehydrogenase/fumarate reductase | |||
Q39WW6 Aconitate hydratase | 1.6 | 0.7 | |
Q39TX6 Succinate–CoA ligase [ADP-forming] subunit beta | 1.4 | ||
Q39T02 Succinate dehydrogenase/fumarate reductase | 0.7 | 2.7 | |
Q39XQ3 Succinate–CoA ligase [ADP-forming] subunit beta | 0.7 | ||
Q39VG0 2-[hydroxy(Phenyl)methyl]-succinyl-CoA dehydrogenase subunit | 2.2 | ||
Q39VF2 (R)-benzylsuccinate synthase, beta subunit | 5.8 | 3.9 | |
Q39VG2 Benzoylsuccinyl-CoA thiolase subunit | |||
Q39VG1 Benzoylsuccinyl-CoA thiolase subunit | |||
Q39VG4 Electron transfer flavoprotein, alpha subunit | 1.7 | 1.5 | |
Q39VG8 Succinyl:(R)-benzylsuccinate coenzyme A transferase subunit | |||
Q39VG5 Electron transfer flavoprotein, beta subunit | 1.5 | 1.4 | |
Q39VG6 (E)-2-benzylidenesuccinyl-CoA hydratase | |||
Q39VF5 Aromatic hydrocarbon degradation outer membrane protein | 2.1 | ||
Q39VG9 Succinyl:(R)-benzylsuccinate coenzyme A transferase subunit | |||
Q39VF9 2-[hydroxy(Phenyl)methyl]-succinyl-CoA dehydrogenase subunit | 0.6 | ||
Q39VG7 (R)-benzylsuccinyl-CoA dehydrogenase | 2.0 |
Most of the proteins predicted for benzoate metabolism (21 proteins) were detected in all columns (
Twelve proteins of the
Four gene products from the
Twelve out of 15 proteins predicted for toluene degradation by
Out of six predicted proteins for phenol degradation only the iron-sulfur cluster binding protein BamI (Gmet_2079) and beta subunit of phenylphosphate carboxylase were detected in column experiments and had significantly higher abundances at medium, low, and non-detectable benzoate concentrations relative to high benzoate concentrations (Data Sheet S1). In retentostat experiments four proteins from phenol catabolism were detected, including protein BamI (Gmet_2079) and phenylphosphate carboxylase, beta subunit (Gmet_2102) which was detected only at low growth rates in retentostats (Marozava et al.,
Eight proteins in the genome of
Neither proteins from the
It is still unclear how microorganisms regulate the consumption of organic compounds under natural conditions. Due to the very low growth rates and substrate concentrations, chemostats have been considered as a more appropriate setup to study microbial physiology at oligotrophic conditions, in contrast to batch incubations with high substrate concentrations and maximal growth rates (Kovarova-Kovar and Egli,
Several studies investigated the physiology of
To elucidate the physiology of
Proteomic analysis of
Expression of the two benzoate-activating enzymes (succinyl:benzoate coenzyme A transferase and benzoate-coenzyme A ligase) might assist
Increased production of antitoxin proteins, efflux pumps, and other cell wall-related proteins suggests that
Even in the absence of toluene in the inflow of the columns,
Schematic representation of the proposed regulation of catabolic pathways by
Some regulators are known to be related to the control of aromatic degradation pathways. For example, the BgeR transcriptional regulator represses expression of the benzoate degradation pathway of
In general, carbon catabolite repression is probably relieved when concentrations of the preferred substrate drop below certain thresholds but the specific molecular mechanism needs further investigation.
Only few studies exist on the regulation of the peripheral pathway of toluene degradation in anaerobic bacteria. Recently, it has been shown that the nitrate-reducing bacterium
The metagenomes of the other taxa detected in our columns showed some aerobic benzoate degradation pathways (Data Sheet S1). However, since the columns were run under anoxic conditions these pathways were probably not relevant. The only molecular oxygen in the system might have been produced from nitrate reduction via nitric oxide dismutases, since it has been shown recently that these enzymes are abundant in different environments but have been strongly overlooked (Zhu et al.,
We conclude that low substrate concentrations in combination with low growth rates and probably some other factors present in the environment such as mixed communities and a sessile lifestyle, prepare cells of
Based on our previous data and the column experiments presented here, we suggest an updated model for regulation of catabolic pathways of
Thus, we propose that organisms such as
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/
SM and RM conceived the idea and wrote the manuscript. SM performed the column experiments and analyzed the data. HM conducted metagenomics. JM-P performed metaproteomics. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank Alexander Probst for providing computing power and support with bioinformatics.
The Supplementary Material for this article can be found online at:
All proteins and their abundances detected in all column fractions.