Interkingdom Cross-Feeding of Ammonium from Marine Methylamine-Degrading Bacteria to the Diatom Phaeodactylum tricornutum

ABSTRACT

Methylamines occur ubiquitously in the oceans and can serve as carbon, nitrogen, and energy sources for heterotrophic bacteria from different phylogenetic groups within the marine bacterioplankton. Diatoms, which constitute a large part of the marine phytoplankton, are believed to be incapable of using methylamines as a nitrogen source. As diatoms are typically associated with heterotrophic bacteria, the hypothesis came up that methylotrophic bacteria may provide ammonium to diatoms by degradation of methylamines. This hypothesis was investigated with the diatom Phaeodactylum tricornutum and monomethylamine (MMA) as the substrate. Bacteria supporting photoautotrophic growth of P. tricornutum with MMA as the sole nitrogen source could readily be isolated from seawater. Two strains, Donghicola sp. strain KarMa, which harbored genes for both monomethylamine dehydrogenase and the N methylglutamate pathway, and Methylophaga sp. strain M1, which catalyzed MMA oxidation by MMA dehydrogenase, were selected for further characterization. While strain M1 grew with MMA as the sole substrate, strain KarMa could utilize MMA as a nitrogen source only when, e.g., glucose was provided as a carbon source. With both strains, release of ammonium was detected during MMA utilization. In coculture with P. tricornutum, strain KarMa supported photoautotrophic growth with 2 mM MMA to the same extent as with the equimolar amount of NH₄Cl. In coculture with strain M1, photoautotrophic growth of P. tricornutum was also supported, but to a much lower degree than by strain KarMa. This proof-of-principle study with a synthetic microbial community suggests that interkingdom cross-feeding of ammonium from methylamine-degrading bacteria is a contribution to phytoplankton growth which has been overlooked so far.

IMPORTANCE Interactions between diatoms and heterotrophic bacteria are important for marine carbon cycling. In this study, a novel interaction is described. Bacteria able to degrade monomethylamine, which is a ubiquitous organic nitrogen compound in marine environments, can provide ammonium to diatoms. This interkingdom metabolite transfer enables growth under photoautotrophic conditions in coculture, which would not be possible in the respective monocultures. This proof-of-principle study calls attention to a so far overlooked contribution to phytoplankton growth.

INTRODUCTION

Within the marine phytoplankton, diatoms (Bacillariophyceae) contribute significantly to the phototrophic primary production in the oceans (1). Typically, pelagic as well as benthic diatoms are associated with heterotrophic bacteria, leading to organismic interactions that range from commensal to antagonistic relationships (2). As the concentration of dissolved organic carbon is usually low in the water column of the oceans, heterotrophic bacteria can obviously profit from these interactions by using organic substrates released by the photoautotrophic diatoms. This mainly commensal relationship has been known for a long time and led to the definition of the phycosphere as a diffusive boundary layer around a phytoplankton cell, in which heterotrophic bacteria have access to organic substrates released by the photoautotrophic microalgae (3). Regarding antagonistic relationships, some bacteria can attack diatoms to utilize them as a nutrient source (4), and diatoms can use defense mechanisms against pathogenic bacteria, such as the formation of ω fatty acids as antibacterial compounds (5, 6). Interactions in which diatoms profit from heterotrophic bacteria are less obvious, and there are only a few clear examples and mechanisms known so far (2). The most prominent example is transfer of vitamin B₁₂ from bacteria to diatoms, which can be auxotrophic for this compound (7). Recently, diatom growth stimulation by a bacterium-produced substance was shown (8). Among the benefits that diatoms may receive from heterotrophic bacteria is the improved access to inorganic nutrients such as iron (9). As diatoms often are limited in their nitrogen supply in many marine habitats, they can also profit from ammonium provided by nitrogen-fixing cyanobacteria (10). Apart from inorganic nitrogen compounds, ocean waters also contain a considerable amount of organic nitrogen compounds, in particular methylamines (11). Methylamines such as monomethylamine (MMA) are derived from the degradation of proteins and other organic nitrogen compounds and occur ubiquitously in surface waters. Notably, methylamines are also formed during the degradation of osmolytes such as glycine betaine produced by marine organisms (12). Methylamines are C₁ compounds (i.e., compounds containing no carbon-carbon bond), which can be utilized as carbon and energy sources by methylotrophic bacteria (13, 14). Many bacteria can also use methylamines as a nitrogen source only (15). In contrast to bacteria, diatoms have never been shown to be able to use MMA or other methylamines as a nitrogen source (16–18).

In the oceans, methylotrophic bacteria appear to be very abundant (19). Methylamine-degrading bacteria particularly, including MMA-degrading strains, have also been shown to co-occur with phytoplankton blooms (20). There are three known pathways for the degradation of MMA (13, 21). In the first pathway, MMA is oxidatively deaminated to formaldehyde and ammonium by a periplasmic MMA dehydrogenase (MADH) encoded by the genes mauAB (22). The second pathway is a variant that employs a MMA oxidase for this reaction and occurs in Gram-positive bacteria, e.g., in a strain of Arthrobacter sp. (23), as well as in some enterobacteria (24, 25). In the third pathway, MMA is bound to glutamate, yielding N-methylglutamate (NMG) by N-methylglutamate synthase (NMGS) encoded by the genes mgsABC (26). NMG is then split into glutamate and formaldehyde catalyzed by NMG dehydrogenase (NMGDH) encoded by mgdABCD. Concomitant with the intermediate NMG, MMA can also be bound to glutamate in a different reaction, yielding γ-glutamylmethylamide catalyzed by γ-glutamylmethylamide synthetase (GMAS) encoded by the gene gmas. The intermediate γ-glutamylmethylamide is cleaved to formaldehyde, α-ketoglutarate, and ammonium by an unknown enzyme system. While the MMA dehydrogenase pathway is believed to be dominant among true methylotrophic bacteria, which can use MMA as a carbon and energy source, the NMG pathway is also found in bacteria that use MMA as a nitrogen source only (15).

The co-occurrence of methylamines and methylotrophic bacteria in the marine water column together with the association of heterotrophic bacteria and diatoms raises the hypothesis that bacteria may provide a nitrogen source to diatoms by methylamine degradation. Such a potential cross-feeding from methylamine-degrading bacteria to diatoms has, to our knowledge, never been addressed, although methylamines are recommended substrates for testing the axenity of diatom cultures (National Center for Marine Algae and Microbiota [https://ncma.bigelow.org/]). Thus, the goal of this study was to investigate whether this cross-feeding is possible and allows photoautotrophic growth. As a model organism for this proof-of-principle study, we chose the pennate diatom Phaeodactylum tricornutum, which has been shown to be unable to use methylamines (27). In addition, it can grow axenically, which excludes any additional requirement for bacteria. Our study was initiated by the attempt to isolate bacteria that support photoautotrophic growth of P. tricornutum in enrichment cultures with MMA as the sole nitrogen source.

MATERIALS AND METHODS

Organisms and growth media.

The diatom Phaeodactylum tricornutum (strain UTEX 646) was kindly provided by Peter Kroth (Constance, Germany) and verified as an axenic culture as described previously (28). All bacterial strains used for cocultivation with P. tricornutum were isolated in this study. All microorganisms were cultivated in the marine mineral medium SW-f/2, which is based on the commercial Tropic marine salt mixture (Tropic Marin, Germany) in a final concentration of 3.32% (wt/vol) and adjusted to pH 7.0 with HCl. After autoclaving, the salt solution was cooled and supplemented with (final concentrations in parentheses) NaNO₃ (880 μM), NaH₂PO₄ (36.2 μM), NaSiO₃ (35 μM), and Tris-HCl (2 mM [pH 8.2]). Finally, the medium was supplemented with f/2 vitamin (0.05%) and f/2 trace element (0.1%) solutions (29). For diatom precultures, the SW-f/2 medium was prepared with a diluted Tropic marine salt solution (1.66% [wt/vol]). For monocultures of bacteria as well as for cocultures, nitrate was omitted from the medium; instead MMA (Sigma-Aldrich, Germany) or NH₄Cl as the nitrogen source and glucose or other carbon and energy sources were added at concentrations indicated in Results. Solid media were prepared by adding 1.5% agar to the salt solution prior to autoclaving.

Isolation of MMA-degrading bacteria.

Bacteria able to degrade MMA were enriched by cocultivation with axenic diatom cultures of P. tricornutum as described previously (28). Samples from the North Sea (seawater samples from Baltrum and from Bremerhaven, Germany) were used for inoculation of SW-f/2 medium with 0.88 mM NaNO₃ or 2 mM MMA-containing axenic diatom cultures at a chlorophyll concentration of 0.2 μg ml⁻¹. Photoautotrophic growth in enrichment cultures was measured with a ChemiDoc MP imaging device (Bio-Rad) as described previously (28). After 14 days of incubation, 50 μl of enrichment cultures showing photoautotrophic growth was plated on SW-f/2 solid medium containing 10 mM MMA and incubated at 30°C in the dark. From single colonies occurring on these agar plates, bacterial strains were isolated by repeated restreaking of single colonies on fresh solid medium until cultures with a homogeneous macroscopic and microscopic appearance were obtained. Bacterial strains showing different colony morphologies on solid media were chosen for further identification by analysis of their 16S rRNA genes (described below).

Growth experiments in monocultures.

Growth experiments with bacterial strains in monocultures were performed in 10-ml test tubes containing 3 ml SW-f/2 medium that were incubated at 30°C on a rotatory shaker at 200 rpm (Orbital Shaker 3015; GFL, Germany). Growth was measured by optical density at a wavelength of 600 nm (OD₆₀₀) in a photometer (spectrophotometer M107; Camspec, United Kingdom) or by CFU as described previously (30) by decimally diluting cultures in SW-f/2 medium without carbon and nitrogen sources and plating appropriate dilution steps on SW-f/2 plates containing 10 mM glucose and 2 mM MMA. Precultures of both strains were inoculated from solid medium; main cultures were inoculated from precultures to an OD₆₀₀ of 0.01. Prior to inoculation of main cultures, appropriate volumes of the precultures were washed twice by centrifugation at 8,000 × g for 5 min at room temperature. Precultures and, unless otherwise stated, main cultures of Donghicola sp. strain KarMa were grown with 10 mM glucose or 10 mM succinate and 2 mM MMA. For Methylophaga sp. strain M1, precultures and, unless otherwise stated, main cultures were grown with 20 mM MMA.

Photoautrotrophic growth of the diatom P. tricornutum in monocultures was measured by chlorophyll fluorescence as described above. Precultures and main cultures were incubated in Erlenmeyer flasks at 21°C at 100 rpm (Orbital Shaker 3015; GFL, Germany) and a photon flux density of 80 microeinsteins m⁻² s⁻¹ (light source, Lumilux Warmwhite; Osram, Germany) with a 14-h/10-h light/dark cycle in a light incubator. Precultures of diatoms were inoculated in 10 ml in 50%-diluted SW-f/2 in 50-ml Erlenmeyer flasks from solid medium and incubated for 7 days. Main cultures were inoculated from precultures to a chlorophyll concentration of 0.02 μg ml⁻¹ in 30 ml SW-f/2 in 100-ml Erlenmeyer flasks or in 100 ml SW-f/2 in 250-ml Erlenmeyer flasks.

Growth experiments in cocultures.

Cocultures of P. tricornutum with strains KarMa and M1, respectively, were set up in 100-ml Erlenmeyer flasks containing 50 ml SW-f/2 medium containing MMA or NH₄Cl (each 2 mM) instead of NaNO₃ as the nitrogen source. Cocultures were incubated in the same way described above for monocultures of P. tricornutum. For inoculation of cocultures, main cultures of the individual microorganisms were set up as described above. Appropriate volumes of the bacterial main cultures were harvested by centrifugation (8,000 × g, 5 min, room temperature) and added to the coculture medium to an OD₆₀₀ of 0.01. In parallel, appropriate volumes of the diatom main culture were harvested by centrifugation (16,000 × g, 10 min, room temperature) and added to the coculture medium to a chlorophyll concentration of 0.02 μg ml⁻¹. Cocultures containing Donghicola sp. strain KarMa tagged with the gene for the green fluorescent protein (GFP) were also supplemented with 20 μg ml⁻¹ gentamicin. Growth of P. tricornutum in cocultures was measured by chlorophyll fluorescence and by microscopic cell counts using aliquots taken at regular intervals from the growing cultures. To measure chlorophyll fluorescence, 200-μl aliquots were transferred to a microtiter plate, which was placed into an imaging device as described previously (28). For counting of cell numbers, aliquots were transferred to a microscopic Thoma cell counting chamber; each sample was evaluated three times in six-count quadrats. Bacterial growth in coculture was determined as described above for monocultures.

Experiments with resting cells.

Main cultures of Donghicola sp. strain KarMa were cultivated as described above. At the beginning of stationary growth phase, cells were harvested by centrifugation (8,000 × g, 5 min, room temperature), washed with SW-f/2 medium without any nitrogen source followed by centrifugation, suspended to an OD₆₀₀ of 0.7 in SW-f/2 medium with 2 mM MMA in 250-ml Erlenmeyer flasks, and incubated for 1 day at 30°C and 200 rpm (Orbital Shaker 3015; GFL, Germany). Aliquots from cell suspensions were removed for determination of optical density and concentrations of MMA and ammonium.

Microscopy and staining.

For cell counting and morphological analysis, as well as alcian blue staining, a light microscope (Nikon Eclipse C1; Nikon, Japan) was used. For alcian blue staining, 20 μl of each coculture was heat fixated on glass slides. The fixated sample was covered with 1 ml 1% (wt/vol) alcian blue (Roth, Germany). For microscopy, the stain was dissolved in 3% [wt/vol] acetic acid at pH 2.36. Afterwards, the glass slide was washed with 2 ml 3% (vol/vol) acetic acid, before it was washed strongly with double-distilled water (31). Furthermore, a confocal laser scanning microscope (CLSM) (Leica TCS SP5/AOBS; Leica, Germany) with hybrid detection was used. Chlorophyll fluorescence was detected between 580 to 680 nm by an excitation wavelength of 488 nm from a krypton laser. GFP and a concanavalin A-fluorescein isothiocyanate (FITC) dye (Fischer scientific, Germany) were detected between 488 to 490 nm by an excitation of 510 nm. For concanavalin A-FITC staining, an established protocol (32) was modified. Cell suspensions was applied on glass slides, dried, and covered with 10 μg ml⁻¹ concanavalin A-FITC in SW-f/2 medium.

Determination of MMA and ammomium.

MMA and ammonium were determined by derivatization of the amino group with diethylethoxymethylenemalonate (DEEMM) followed by high-performance liquid chromatography (HPLC) analysis. For this, aliquots from culture supernatants were centrifuged at 16,000 × g for 10 min at room temperature to remove the cells. The supernatant was either stored at −20°C until analysis or directly derivatized with DEEMM as described previously (33). For verification of the derivatization reaction, samples were analyzed with HPLC coupled to mass spectrometry (LC-MS). The LC-MS system consisted of a Dionex Ultimate 3000 high-performance liquid chromatograph (ThermoFisher Scientific, Germany) with a UV/visible light diode array detector and an ion trap mass spectrometer (Amazon speed; Bruker, Bremen, Germany) with an electrospray ion source for electrospray ionization (ESI). A C₁₈ reversed-phase column (at 25°C, 150 by 3 mm; Eurosphere II 100-5 C₁₈ column [Knauer, Germany]) with ammonium-acetate (10 mM [pH 7]) and acetonitrile (HPLC grade) at a flow rate of 0.3 ml min⁻¹ was used to separate the samples. Ionization of samples was performed at alternating ionization mode with electrospray ionization using the following settings: capillary voltage, 4,000 V; plate offset, 500 V; nebulizer pressure, 22.5 lb/in²; dry gas flow, 12 liters min⁻¹; and dry gas temperature, 200°C. The mass spectrometer was operated in ultrascan mode with a scan range of 50 to 1,000 Da. For routine quantification of ammonium and MMA, the DEEMM-derivatized culture supernatant was separated by HPLC with a C₁₈ reversed-phase column at 30°C (Vertex Plus C₁₈ reversed-phase column; 250 by 3 mm [Knauer, Germany]) through a binary gradient with potassium acetate (10 mM [pH 7]) as eluent A and acetonitrile (HPLC grade) as eluent B. The gradient started with 10% eluent B at a flow rate of 0.5 ml min⁻¹. After 2 min, the flow rate was increased to 0.6 ml min⁻¹, and eluent B was increased to 45% within 20 min. Then, eluent B was increased to 70% within 2 min. After 1 min, eluent B was reduced back to 10% followed by an equilibration for 5 min. For HPLC analysis, a 1-ml sample each from the pure culture and coculture was taken. The linear detection range was determined to be between 50 μM and 1 mM.

MMA dehydrogenase assay.

Fifty milliliters of main culture of strains KarMa or M1 was grown in 250-ml Erlenmeyer flasks with 10 mM glucose plus 2 mM MMA or with 20 mM MMA, respectively, as described above. Cultures were harvested in the late exponential growth phase by centrifugation at 10,000 × g for 10 min at 4°C. Cells were washed once with SW-f/2 medium without vitamins and trace elements and concentrated in the same solution. Dense cell suspensions were sonicated with 0.5 cycle and 50% amplitude (UP200S; Hielscher, Germany), followed by centrifugation at 16,000 × g for 10 min at 4°C to remove cell debris. The supernatant was used as cell extract subsequently for MMA dehydrogenase assay largely following a previously described protocol (34). Assays contained SW medium with Tris-HCl (2 mM) as buffering solution (pH 7.5), 50 μM dichlorophenolindophenol (DCPIP), 300 μM phenazinemethosulfate (PMS), and cell extract (ca. 1 mg protein) and were started by the addition of 7 mM MMA. MMA dehydrogenase activity was determined by the MMA-dependent reduction rate of DCPIP at 600 nm and 30°C (ε = 21.5 mM⁻¹cm⁻¹). Protein concentrations in cell extracts were determined as described previously (35).

Analysis of 16S rRNA genes.

Amplification of 16S rRNA genes was performed as described previously (28). Sequence alignment was performed using the default setting of SeqMan Pro (DNAStar, Inc., Madison, WI). The RDP classifier (https://rdp.cme.msu.edu/classifier/classifier.jsp) was used for analysis of phylogenetic affiliation with default settings (36). The BLAST search tool from NCBI (http://www.ncbi.nlm.nih.gov/ [37]) was used with default settings, although “Models (XM/XP)” and “Uncultured/environmental sequences” were excluded.

Generation of fluorescence-tagged cells of Donghicola sp. strain KarMa.

Strain KarMa was tagged with a stable version of the green fluorescence protein (GFP). For this, the gfp gene with restriction sites for SacI and XbaI (gfp forward, 5′-TTTTTTTGAATTCATCCCCGGGTACCTAGAATTA-3′; gfp reverse, 5′-TAATCTAGACAGGGTTTTCCCAGTCACGA-3′) was amplified from the vector pUCP18::gfp, cloned into the restriction sites of the vector pBBR1MCS-5 (38), and transformed into chemically competent cells of Escherichia coli strain S17-1 (39). The generated E. coli strain, S17-1(pBBR1MCS-5::gfp), was used for conjugation of Donghicola sp. strain KarMa. The biparental conjugation protocol as described in Piekarski et al. (40) was modified. The donor strain was incubated in LB medium containing 20 μg ml⁻¹ gentamicin at 37°C overnight. A main culture of strain KarMa was incubated until an OD₆₀₀ of 1 was reached. Both cultures were mixed with an OD₆₀₀ donor-to-recipient ratio of 1:2 in 500 μl SW-f/2 medium, and 100 μl of suspension was plated as droplets on LBS agar plates (1% [wt/vol] tryptone, 0.5% [wt/vol] yeast extract, 2% [wt/vol] NaCl, 50 mM Tris-HCl, with pH adjusted to 7.5 prior to autoclaving). After incubation for 6 h at 30°C, colony material was transferred to 4 ml SW-f/2 medium and mixed vigorously. Suspensions (50, 100, and 200 μl) were plated on SW-f/2 medium agar plates with 10 mM glucose and 2 mM MMA containing 20 μg ml⁻¹ gentamicin. Cells of strain KarMa carrying the plasmid were identified by detection of GFP fluorescence with the imager. Strain KarMa(pBBR1MCS-5::gfp) showed a similar growth pattern to the untagged wild type in single culture (see Fig. S1A in the supplemental material).

Annotation of genes associated with the degradation of MMA in Donghicola sp. strain KarMa.

Genes putatively involved in degradation of MMA by strain KarMa were identified by a BLASTN search with an expectation value cutoff of 10 on the GeneDB platform (41) of known MADH and NMG pathway-associated genes. The identity of the genes was further analyzed by a BLASTP search tool with default settings from NCBI (37) with translated amino acid sequences.

RNA isolation from cultures of Donghicola sp. strain KarMa.

For isolation of RNA, precultures of strain KarMa were grown in SW-f/2 medium containing 10 mM glucose and 1 mM MMA or 1 mM NH₄Cl for 16 h at 30°C. Cells were washed and used to inoculate main cultures in SW-f/2 medium containing 10 mM glucose and 1 mM MMA or 1 mM NH₄Cl in triplicates with an OD₆₀₀ of 0.05. After incubation for 12 h at 30°C, the respective triplicates were combined, harvested by centrifugation (10,000 × g, 5 min, room temperature), and resuspended in SW-f/2 medium. Cell suspensions (10⁹ cells) were resuspended in 100 μl TE buffer (10 mM Tris, 1 mM EDTA [pH 8]) with RiboLock RNase inhibitor (Thermo scientific) and incubated with 1 mg ml⁻¹ lysozyme for 10 min at room temperature. Cells were broken with acid-washed glass beads (150 to 212 μm, 70 to 100 U.S. sieve; Sigma-Aldrich) with Mikro-Dismembrator S (Sartorius, Germany) at 3,000 rpm for 1 min. From these homogenates, RNA was extracted with the innuPREP RNA minikit from Analytik Jena (Jena, Germany) according to the manufacturer's instructions. Residual DNA was removed by digestion with 10 U RNase-free DNase I (Thermo scientific) for 1 h in the presence of RiboLock RNase inhibitor (Thermo scientific). After DNA digestion, the RNA was again purified with the innuPREP RNA minikit from Analytik Jena (Jena, Germany). RNA was checked for DNA contamination by amplification with the primer used for RNA transcript analysis.

RNA transcript analysis by semiquantitative RT-PCR.

For RNA transcript analysis, 20 ng of RNA was transcribed to cDNA with the ProtoScript II first-strand cDNA synthesis kit (New England BioLabs) according to the manufacturer's description for the standard protocol with d(T)₂₃ VN primer mix. The cDNA was used for amplification of the putative genes involved in degradation of MMA using the Q5 high-fidelity DNA polymerase (New England BioLabs) according to manufacturer's instructions. The following primers for amplification were used: mauB_Chr (forward, 5′-TTATTGATCCAGCTTGATC-3′; reverse, 5′-TGAAATACAGTACCGCC-3′), mgsB_PlA (forward, 5′-AGGCTCGACCGGTTACTACT-3′; reverse, 5′-AGTCGGCACCCAGAGATTTG-3′), gmas_PlA (forward, 5′-CTTGGAAACCCGATGTTGCG-3′; reverse, 5′-ATGTCGATGTCCCAACGCTT-3′), mgdC_PlA (forward, 5′-TCAGTCACGGAGACGTAC-3′; reverse, 5′-GACAACTTTGGGGATGGCAA-3′), mgdC_Chr (forward, 5′-TCAGACATTCTGCTTTTCCCC-3′; reverse, 5′-AAGAGTGGTTGCAGACCGAG-3′). Amplified PCR products were verified by sequencing.

Accession number(s).

Sequences of the aligned 16S rRNA genes for each isolate have been deposited in GenBank. The following accession numbers were assigned for the bacterial strains isolated in this study: Donghicola sp. strain KarMa, KU524453.1; Methylophaga sp. strain M1, ku524454.1; Saccharospirillum sp. strain M2, ku524455.1; Thalassospira sp. strain M3, ku524456.1; Mameliella sp. strain M4, ku524457.1; Sulfitobacter sp. strain M5, ku524458.1; and Arenibacter sp. strain M6, KU524459.1.

RESULTS

Isolation of MMA-degrading bacteria from cocultures with P. tricornutum.

For investigation of whether MMA could be rendered available as a nitrogen source for the diatom P. tricornutum, enrichment cultures were set up in which an axenic diatom culture was seeded with seawater samples from the North Sea as described previously (28). From enrichment cultures that showed photoautotrophic growth, bacteria were isolated on agar plates with MMA as the carbon, energy, and nitrogen source. Bacterial isolates showing growth on the respective agar plates were retested in liquid cocultures with P. tricornutum, and those strains that clearly supported growth of the diatom were identified by sequencing their 16S rRNA genes. These results showed that bacteria from the different phylogenetic groups (Bacteroidetes [Arenibacter sp. strain M6], Alphaproteobacteria [Donghicola sp. strain KarMa, Thalassospira sp. strain M3, Mameliella sp. strain M4, and Sulfitobacter sp. strain M5], and Gammaproteobacteria [Methylophaga sp. strain M1 and Saccharospirillum sp. strain M2]) are obviously able to support growth of P. tricornutum when MMA is the only nitrogen source (see Table S1 in the supplemental material). Before cocultures for investigating this possibility in detail were set up, two selected strains, Donghicola sp. strain KarMa and Methylophaga sp. strain M1, were chosen for further analysis of their MMA metabolism in monoculture. Strain KarMa was chosen for its strong growth support of P. tricornutum, while strain M1 was chosen because Methylophaga species are well-described methylotrophs (42).

Growth of strains KarMa and M1 with MMA.

Strain KarMa could not grow with MMA as the carbon and energy source (Fig. 1A; see Fig. S2A in the supplemental material) while it grew with MMA as a nitrogen source when glucose was supplied as a carbon source (Fig. 1B; see Fig. S2B). During growth with glucose, MMA rapidly disappeared from the culture supernatants and was not detectable anymore at the beginning of the exponential growth phase. Simultaneously, ammonium was detectable in the culture supernatant at low concentrations (Fig. 1B). When equimolar amounts of ammonium were supplied as the nitrogen source instead of MMA, the cells grew with similar rates and reached similar final optical densities (not shown). When a suspension of resting cells of strain KarMa at OD₆₀₀ of about 0.7 were supplied with MMA, it was rapidly consumed, and stoichiometric amounts of ammonium accumulated in the culture supernatant (see Fig. S3 in the supplemental material).

FIG 1.

Growth measured as OD₆₀₀ (■) of Donghicola sp. strain KarMa and Methylophaga sp. strain M1 with MMA (●) and formation of ammonium (▲). (A) Strain KarMa with MMA as the sole carbon and nitrogen source. (B) Strain KarMa with glucose (10 mM) as the carbon source and MMA as the nitrogen source. (C) Strain M1 with MMA as the sole carbon and nitrogen source. (D) Strain M1 with glucose (10 mM) and MMA. Error bars indicate standard deviations (n = 3).

Strain M1 could grow with MMA as a carbon, energy, and nitrogen source and consumed it completely (Fig. 1C; see Fig. S2C in the supplemental material). When larger amounts of MMA were added, the strain degraded MMA completely and reached higher final optical densities (see Fig. S2C). When MMA was supplied as the sole nitrogen source with glucose as the carbon and energy source, strain M1 grew and rapidly degraded MMA at the beginning of exponential growth (Fig. 1D; see Fig. S2D). The accumulating ammonium was partially consumed during further growth.

These results showed that strain M1 is a true methylotrophic bacterium, while strain KarMa is restricted to MMA utilization as a nitrogen source. Faint growth of strain KarMa on agar plates during its isolation with MMA as the sole carbon and nitrogen source was therefore likely dependent on the use of utilizable constituents from the autoclaved agar. MMA dehydrogenase activity was assayed in cell extracts of both strains. In extracts of strain M1 cells grown with MMA as the carbon, energy, and nitrogen source, MMA dehydrogenase activity was present (137 ± 5 mU mg protein⁻¹). In extracts of strain M1 cells grown with MMA as the nitrogen source and glucose as the carbon source, the specific activity of MMA dehydrogenase was in a similar range (126 ± 8 mU mg protein⁻¹). In extracts of cells grown with NH₄Cl as the nitrogen source and glucose as the carbon source, MMA dehydrogenase activity could not be detected. In contrast to strain M1, MMA dehydrogenase activity could not be detected in extracts of strain KarMa grown with MMA as the nitrogen source and glucose as the carbon source.

To identify the enzymes used for MMA utilization in strain KarMa in silico, candidate genes for the MADH and the NMG pathways were annotated in the genome (see Table S2 and Fig. S4 in the supplemental material [genome unpublished]). cDNA derived from cells grown with MMA or NH₄Cl as the N source, respectively, was analyzed by PCR using primers for the respective candidate genes. This analysis showed that a gene annotated as mauB, coding for the catalytic subunit of MADH, was transcribed during growth with MMA as well as NH₄Cl (see Fig. S5 in the supplemental material). Furthermore, genes putatively encoding NMGS (mgsB_PlA), GMAS (gmas_PlA), and NMGDH (mgdC_Chr) were transcribed during growth on either nitrogen source (see Fig. S5). This transcriptional analysis indicated that the MADH as well as the NMG pathway could be active in strain KarMa.

Cocultures with P. tricornutum and strain KarMa or M1.

In the next step, it was investigated how the MMA-degrading bacterial strains KarMa and M1 supported photoautotrophic growth of P. tricornutum. For strain KarMa, the GFP-tagged strain was used; these gentamicin-containing cocultures did not differ from cocultures with strain KarMa without antibiotics (see Fig. S1B in the supplemental material). With MMA as the sole nitrogen source, P. tricornutum could grow when KarMa was present, while it could not grow in the absence of the bacterium (Fig. 2A and B). The final cell numbers as well as the final chlorophyll fluorescence values did not significantly differ when equimolar amounts of NH₄Cl were provided as the nitrogen source. Control cocultures without any nitrogen source showed no growth of P. tricornutum, indicating that diatom growth was not caused by nitrogen compounds that may have leaked from lysed bacterial cells (see Fig. S6 in the supplemental material). Growth of strain KarMa measured by CFU showed similar courses when MMA or NH₄Cl was provided as the nitrogen source (Fig. 2C and D). HPLC analysis revealed that in cocultures, complete degradation of MMA after 3 days coincided with an accumulation peak of about 1 mM (Fig. 2E) and that the following ammonium depletion coincided with the fastest growth phase of P. tricornutum (Fig. 2A and B). In contrast, MMA was not degraded in monocultures of the diatom (Fig. 2E). In cocultures with NH₄Cl as the nitrogen source, ammonium was depleted at the same rate as in monocultures of P. tricornutum (Fig. 2F). Taken together, these results clearly showed that strain KarMa could support growth of the diatom by catalyzing the release of ammonium from MMA.

FIG 2.

Coculture of P. tricornutum and Donghicola sp. strain KarMa(pBBR1MCS-5::gfp). (A) Cell count and (B) chlorophyll fluorescence of P. tricornutum in monoculture with MMA (■) and with NH₄Cl (□) and in coculture with strain KarMa with MMA (●) and with NH₄Cl (○). (C) CFU of strain KarMa in coculture with P. tricornutum and MMA. (D) CFU of strain KarMa in coculture with P. tricornutum and NH₄Cl. (E) MMA (●) and ammonium concentrations (▲) of the coculture with MMA as well as the MMA concentration of the P. tricornutum monoculture (■). (F) Ammonium concentration of P. tricornutum monoculture (□) and coculture (○). Error bars indicate standard deviations (n = 3).

In coculture with strain M1, P. tricornutum could also grow with MMA as the nitrogen source (Fig. 3A and B). However, the final cell numbers and chlorophyll fluorescence were clearly lower than for growth with equimolar amounts of NH₄Cl as the nitrogen source and, thus, also lower than in coculture with strain KarMa. Growth of strain M1 measured by CFU was about three times higher with MMA as the nitrogen source than with NH₄Cl (Fig. 3C and D). HPLC analysis revealed that MMA was rapidly depleted within 1 day (Fig. 3E). At this time point, about 1 mM ammonium accumulated, and it was slowly and only partly depleted within the remaining incubation time. In contrast, ammonium was almost completely depleted, concomitant with growth of P. tricornutum, in cocultures with NH₄Cl as the nitrogen source (Fig. 3F).

FIG 3.

Coculture of P. tricornutum and Methylophaga sp. strain M1. (A) Cell count and (B) chlorophyll fluorescence of P. tricornutum in coculture with strain M1 with MMA (●) and NH₄Cl (○). (C) CFU of strain M1 in coculture with P. tricornutum and MMA (●). (D) CFU of strain M1 in coculture with P. tricornutum and NH₄Cl (○). (E) MMA (●) and ammonium (▲) concentration of the coculture with MMA. (F) Ammonium concentration (△) of the coculture with NH₄Cl. Error bars indicate standard deviations (n = 3).

Microscopic analysis of cocultures.

During the microscopic determination of diatom cell numbers in growth experiments, we observed that the cell shape of P. tricornutum changed in coculture with strain KarMa with MMA as the nitrogen source. At the beginning of the incubation, most of the diatom cells had an oval shape (Fig. 4A; see Fig. S7B in the supplemental material), which changed to the fusiform shape during further incubation (Fig. 4B; see Fig. S7A and C). Notably, in cocultures with NH₄Cl as the nitrogen source, the morphotype of the diatoms was fusiform throughout the whole incubation (not shown). Also in coculture with strain M1 as well as in monoculture, the diatom cells had an unchanged fusiform morphotype. In cocultures with strain KarMa, the morphotype of the bacterial cells of strain KarMa also changed during the incubation, irrespective of the nitrogen source. At the beginning of the incubation, strain KarMa cells grew mainly in long chains of more than 10 cells (Fig. 4A; see Fig. S7B). Later, the cells formed cell aggregates containing mainly single cells and short chains consisting of less than 10 cells; to these bacterial aggregates also diatom cells were attached (Fig. 4B; see Fig. S7A and C). Analysis of the extracellular polymers revealed that these aggregates contained acidic polysaccharides that could be stained with alcian blue (see Fig. S7A). Furthermore, the polymers reacted with the fluorescent-labeled lectin concanavalin A, indicating the presence of mannose and glucose moieties (see Fig. S7B and C).

FIG 4.

Microscopic analysis of the coculture of P. tricornutum and strain KarMa(pBBR1MCS-5::gfp) with MMA. Shown are phase-contrast micrographs of the organisms in coculture at (A) day 1 and (B) day 5. Size bars, 10 μm.

DISCUSSION

In this study, we provide clear proof of principle that ammonium cross-feeding from methylamine-degrading bacteria enables photoautotrophic growth of a diatom. In particular, growth experiments with synthetic communities of the diatom P. tricornutum with two selected bacteria clearly showed that the photoautotrophic growth of the diatom relied on ammonium that was released by degradation of MMA. Thus, our study expands the list of known diatom-bacterium interactions (2) and reveals a situation under which diatom growth depends on bacterial metabolic activity. The rapid onset of photoautotrophic growth of P. tricornutum in enrichment cultures containing MMA as the sole nitrogen source suggested that this kind of interkingdom cross-feeding could readily be established with bacteria from marine samples and that two alternative explanations for diatom growth are unlikely. First, it was unlikely that nitrogen-fixing bacteria were responsible for growth of P. tricornutum as the isolation of bacteria showed that MMA-degrading bacteria were abundant in these enrichments. Second, it was unlikely that growth of algal contaminants were responsible for chlorophyll increase because we did not observe any other diatoms or microalgae in the enrichment cultures.

According to the classical categories of organismic interactions (43), the observed interactions between P. tricornutum and the MMA-degrading bacterial strains KarMa and M1 can be classified as commensalism. This classification originates from the fact that both bacterial strains produce excess ammonium from MMA when they grow under carbon limitation. With strain M1, ammonium accumulation is inevitable because to obtain carbon from MMA, it has to be deaminated first. For a bacterium using MMA as the only nitrogen source, this fast deamination would not be necessary. However, as strain KarMa had rapidly and completely degraded MMA to ammonium already in the early exponential growth phase in monoculture, its MMA utilization was obviously not tightly coupled to the ammonium requirement for growth. With strain KarMa, this commensal interaction had a mutualistic character because strain KarMa reached about 36-fold-higher CFU numbers in cocultures compared to monocultures with MMA as the sole carbon source (Fig. 2C; see Fig. S2A in the supplemental material). These CFU numbers might even be underestimated because the bacterial cells formed chains and aggregates, which would give rise to colony formation by more than one cell per colony. This result indicates that strain KarMa received organic carbon compounds from the diatoms. It is not known which carbon compounds released by the diatom can serve as the substrates for strain KarMa nor how much they can serve. However, the fact that the MMA-degrading activity enabled photoautotrophic growth of P. tricornutum to the same extent as equimolar amounts of ammonium provided as NH₄Cl to the monoculture of P. tricornutum indicated that the bacterium received only a small portion of the nitrogen. This suggests that the amount of organic compounds consumed by strain KarMa is also not very high.

The fate of the carbon resulting from MMA degradation is not known for strain KarMa so far. In many nonmethylotrophic bacteria, it is oxidized to CO₂ to avoid the accumulation of toxic formaldehyde (13). In coculture with strain KarMa, P. tricornutum might, therefore, receive additional CO₂ for carbon fixation. However, cocultures as well as monocultures of P. tricornutum with ammonium as the nitrogen source showed very similar growth rates and reached cell numbers and levels of chlorophyll fluorescence very similar to those in coculture with strain KarMa and MMA. Thus, the additional CO₂ potentially available in the coculture with MMA obviously did not cause a stimulation of growth. In agreement with this conclusion, it has been shown that growth parameters of P. tricornutum did not increase with increasing CO₂ levels at similar temperatures to those applied in our study (44).

In contrast, the true methylotrophic bacterium strain M1 supported diatom growth with MMA as the nitrogen source to a much smaller extent than in coculture with NH₄Cl as the nitrogen source. In agreement with its ability to assimilate carbon from MMA, strain M1 showed about 4 times higher CFU numbers than strain KarMa. Interestingly, in the coculture with MMA, a relatively large residual concentration of ammonium was detected, which was not observed in the corresponding coculture with NH₄Cl. The reduced growth of the diatom together with this residual ammonium suggests an antagonistic influence of strain M1 on the diatom when MMA served as the sole nitrogen source. The basis for this antagonistic influence is unknown, but it appears not to rely on competition for ammonium. In the cocultures of P. tricornutum with strain KarMa, we also obtained indications of an influence of the bacterium on the diatom when MMA served as the nitrogen source. There, the diatom cells exhibited an oval morphotype, which is indicative of stress effects, e.g., nitrogen stress (45–47). As a nitrogen source was apparently not limiting in the coculture, the stress response might rely on the presence of formaldehyde that was presumably released from strain KarMa in the course of oxidative deamination of MMA. However, when we exposed P. tricornutum to formaldehyde in concentrations that could be maximally expected in the coculture, the morphotype of the diatom was not influenced. This indicates that toxic effects potentially caused by this compound were not responsible for the occurrence of cells with an oval morphotype.

The current view is that the MMA dehydrogenase pathway is dominant among bacteria that use MMA also as a carbon source, while those using it as a nitrogen source have mainly the cytosolic NMG pathway (12, 15). According to the genome sequence of strain KarMa (unpublished), this bacterium harbors the mau gene cluster for MMA degradation as well as the genes for the NMG pathway. Regarding the NMG pathway, the genome of strain KarMa contains a gene cluster with high similarity to proven NMG gene clusters in other methylamine-utilizing bacteria (26). However, while we could clearly find inducible activity of the MMA dehydrogenase in cell extracts of strain M1, we could not detect this enzyme activity in cell extracts of strain KarMa despite some attempts to modify the reaction conditions (e.g., testing ferric cyanide as an alternative electron acceptor and omitting PMS). Furthermore, all attempts to detect NMG-dehydrogenase activity according to existing protocols (48, 49) in cell extracts of strain KarMa have not been successful so far (data not shown). However, transcriptional analysis showed that mauB as the large subunit of the MADH and candidate genes required for the NMG pathway were transcribed in cells grown with MMA as well as with ammonium. To our knowledge, constitutive expression of both pathways in one bacterium has not been described so far and might point at a frequent occurrence and a high importance of MMA in natural habitats of strain KarMa. However, it cannot be ruled out that either pathway is subject to posttranscriptional or -translational regulation in such a way that only one pathway is eventually active in growing cells.

Given the abundance of methylotrophic bacteria in the marine plankton (19) and the ubiquitous occurrence of methylamines in the marine environment, the ecological role of methylamine-based ammonium cross-feeding is potentially a very important and so far overlooked interkingdom metabolite flux. This possibility is supported by culture-independent studies that show the presence of bacterial genera in the phycosphere, which could potentially be capable of methylamine utilization (20). In this respect, the environmental concentrations of MMA and other methylated C₁ compound metabolites have certainly been underestimated because of the fast degradation of such compounds in natural habitats (12, 37, 49). As our results suggest that interactions of diatoms with MMA-degrading bacteria are profitable in marine environments, it appears to be an attractive hypothesis that diatoms may release molecules to attract methylamine-degrading bacteria. In agreement with this hypothesis, it has been shown that bacteria can employ chemotaxis to approach and get established within the phycosphere (8, 50–52). To our knowledge, molecules that specifically attract methylotrophic bacteria are not known so far. Nevertheless, based on our present results, it is not possible to conclude that interactions between strains KarMa and M1 with P. tricornutum involve chemotaxis or any other chemical communication. To this end, the physiology of the genus Donghicola has only been barely described so far, and nothing is known about its potential chemical interactions with other organisms (53, 54). Further systematic and mechanistic studies on interactions between methylamine-degrading bacteria and diatoms are on the way in our laboratory.