Biotechnology Institute, University of Minnesota, St. Paul, MN, USA
Department of Bioproducts and Biosystems Engineering, University of Minnesota, St. Paul, MN, USA
Correspondence: Brett M. Barney, Department of Bioproducts and Biosystems Engineering, University of Minnesota, 1390 Eckles Avenue, St. Paul, MN 55108-6130, USA. Tel.: +1 612 626 8751; fax: +1 612 625 6286; e-mail: firstname.lastname@example.org
Microalgae are viewed as a potential future agricultural and biofuel feedstock and also provide an ideal biological means of carbon sequestration based on rapid growth rates and high biomass yields. Any potential improvement using high-yield microalgae to fix carbon will require additional fertilizer inputs to provide the necessary nitrogen required for protein and nucleotide biosynthesis. The free-living diazotroph Azotobacter vinelandii can fix nitrogen under aerobic conditions in the presence of reduced carbon sources such as sucrose or glycerol and is also known to produce a variety of siderophores to scavenge different metals from the environment. In this study, we identified two strains of green algae, Neochloris oleoabundans and Scenedesmus sp. BA032, that are able to utilize the A. vinelandii siderophore azotobactin as a source of nitrogen to support growth. When grown in a co-culture, S. sp. BA032 and N. oleoabundans obtained the nitrogen required for growth through the association with A. vinelandii. These results, indicating a commensalistic relationship, provide a proof of concept for developing a mutualistic or symbiotic relationship between these two species using siderophores as a nitrogen shuttle and might further indicate an additional fate of siderophores in the environment.
Nitrogen is an important component of living systems. In agriculture, conventional crops such as soybeans, alfalfa, and clover meet their nitrogen requirements by forming symbiotic relationships with diazotrophic soil bacteria (Jones et al., 2007; Ikeda et al., 2010), while many other agricultural crops such as corn require substantial nitrogen inputs from Haber-Bosch-based industrial processes. Potential next-generation crops such as single-celled microalgae provide possible benefits in terms of total biomass yield per square area, high density, and rapid growth vs. current conventional agriculture crops, but would also require substantial increases in specific nutrient inputs such as nitrogen. Thus, a key concern in pursuing sustainable next-generation oil crops should include specific consideration of the sustainability of the nitrogen inputs.
Azotobacter vinelandii is a common Gram-negative soil bacterium that can fix atmospheric nitrogen under aerobic conditions. This characteristic differentiates it from many other nitrogen-fixing bacteria that require anaerobic or microaerobic conditions to protect the oxygen-sensitive nitrogenase (Setubal et al., 2009). This feature also makes it an ideal candidate strain for potential co-culture with oxygen-producing phototrophs such as microalgae (Ortiz-Marquez et al., 2012). Additionally, A. vinelandii is considered as an ideal strain in potential biotechnology applications for the production of higher-value bioproducts such as polyhydroxyalkanoates, which could serve as potential bioplastics (Setubal et al., 2009).
Azotobacter strains produce a range of nitrogen compounds which may be released into the extracellular space under certain conditions. Extensive ammonia release was reported two decades ago from A. vinelandii based on a modification resulting in differential regulation of nitrogen fixation genes, resulting in high concentrations of ammonia accumulating in the growth media (Bali et al., 1992; Brewin et al., 1999). This feature was recently utilized to demonstrate the potential to apply ammonia production to the co-culture of various algae strains (Ortiz-Marquez et al., 2012). A. vinelandii has also been reported to excrete a range of additional nitrogen compounds to serve various functions, including proteins involved in the production of alginate (Gimmestad et al., 2006).
As a diazotroph, A. vinelandii requires substantial quantities of iron to grow under optimal conditions. Iron is important for central metabolism and is also essential as a component of the two proteins that make up the nitrogenase complex (Barney et al., 2006; Wichard et al., 2009). A. vinelandii contains multiple siderophore biosynthetic pathways and diverts a portion of the nitrogen obtained through biological nitrogen fixation to the specialized nitrogen-rich pyoverdine and catechol siderophores (Fig. 1), which are excreted into the extracellular space, to assist in the uptake of required metal ions (Tindale et al., 2000; Yoneyama et al., 2011). Siderophore systems enable this bacterium to adapt to metal limitation in the environment (Wichard et al., 2009). The siderophores bind various metals with a high affinity and are then taken up by the cell through membrane-bound transport systems (Cornish & Page, 1995; Palanché et al., 2004; Wichard et al., 2009). It is this latter class of siderophore compounds that were of particular interest in this study to determine the potential of these compounds to serve as a nitrogen source to various algal strains.
Materials and methods
Reagents, strains and cell counting
Azotobacter vinelandii DJ (trans) was obtained from Dennis Dean (Virginia Tech), while Scenedesmus sp. BA032 is an environmental isolate collected from the Cache Valley in northern Utah. Neochloris oleoabundans was obtained from the UTEX culture collection of algae. All chemicals and reagents were obtained from Sigma Aldrich (St. Louis, MO) or Thermo Fisher Scientific (Pittsburgh, PA). Cell counts of S. sp. BA032 were measured using a hemocytometer following the directions of the manufacturer (Hausser Scientific, Horsham, PA).
Bacterial strains and growth conditions
Cultures of A. vinelandii DJ (trans) and gene substitution strains described below were grown aerobically in modified Burk's media for siderophore production (SPB media), which lacks added iron and includes zinc (58 mM sucrose, 0.34 mM CaCl∙2H2O, 0.41 mM MgSO4∙7H2O, 40 μM Na2MoO4∙2H2O, 40 μM ZnCl, 0.73 mM K2HPO4, and 2.32 mM KH2PO4) at 30 °C with agitation at 200 r.p.m. (Page & Sadoff, 1975; Huyer & Page, 1988; Page et al., 2003).
Construction of A. vinelandii gene substitution strains
The csbC gene of A. vinelandii codes for a key enzyme in the biosynthesis of catechol-based siderophores (Tindale et al., 2000). The csbC gene was cloned along with flanking regions of c. 500 bp into a pUC19-derived plasmid using primers BBP1301 (5′-CAGATAAGCTTGTCTG GTCGATCA GGATCGCC ATG-3′) and BBP1302 (5′-GACAGGTACCACGCTG TAGAAATA GTCGTCGT G-3′) with HindIII and KpnI sites used to insert the gene underlined. The csbC gene was then removed from the plasmid using PCR with primers BBP1331 (5′-GACAGGATCCTCTAGA TATGCATA TGGCCTCC TTACGGCT AGAGGACG AG-3′) and BBP1332 (5′-CTGAGGATCCGAGCAC CTCGACCC CGACCTCT TC-3′) with BamHI sites underlined. The spectinomycin resistance cassette from pHP45Ω (Prentki & Krisch, 1984) was introduced into the BamHI site to produce pPCRSIDK3. The strain A. vinelandii AZBB040 was constructed by transforming A. vinelandii DJ (trans) with pPCRSIDK3. Following a double homologous recombination event, colonies were selected for antibiotic resistance resulting in a strain A. vinelandii AZBB040, which contains the ΔcsbC::spectr substitution. Strains were confirmed to contain the genomic modification by PCR with primers BBP1419 (5′-GAACAG CACGAA GCTCAG CATCAG C-3′) and BBP1420 (5′-CGAACA CCTGTT GCAGCT TGCAGC-3′) lying outside of the region modified.
The gene coding for Avin_25580 of A. vinelandii encodes a key enzyme in the biosynthesis of azotobactin siderophore (Yoneyama et al., 2011). The gene for Avin_25580 was cloned along with flanking regions of c. 500 bp into a pUC19-derived plasmid using primers BBP1502 (5′-GACTAAGCTTGAAGCG TTCCCGGC TGAAGGTC-3′) and BBP1503 (5′-GTGACC CTGTTCAT GCTGCTGC TG-3′) with HindIII and EcoRI (downstream of primer) sites used to insert the gene. The gene for Avin_25580 was then removed from the plasmid using PCR with primers BBP1562 (5′-NNNGGATCCACCCAGG TCAACGAC CTGCTGCT G-3′) and BBP1563 (5′-NNNGGATCCGTCCTCG CCACCTGC TCCTCGAT CAACTG-3′) with BamHI sites. A tetracycline resistance cassette derived from pRK415 (Mather et al., 1995) was introduced into the BamHI site to produce pPCRSIDK9. A. vinelandii was transformed as described above. Colonies were selected for antibiotic resistance resulting in the strain A. vinelandii AZBB041, which contains the substitution to the gene coding for Avin_25580. Strains were confirmed to contain the genomic modification by PCR with primers BBP1623 (5′-CACTTG CTGGAC AAAGAG ACGGTC C-3′) and BBP1624 (5′-CTGCGG TTCAGC TCTCCG TAGCTC ATTTC-3′) lying outside of the region modified.
Optical characterization to confirm phenotype
Azotobacter vinelandii wild-type and gene substitution strains were grown for 72 h in SPB media, and cell density was determined by measuring the optical density (OD) at 600 nm to confirm similar rates of growth. Specific optical features of the siderophores were used to monitor the culture supernatant (Page & Huyer, 1984; Page et al., 1991; Rodríguez-López et al., 1991). The absorbance spectra were measured using a Varian Bio 50 UV/Vis spectrophotometer. Levels of catechols were determined using an approach similar to that used by Page and Huyer (Page & Huyer, 1984) and the reported extinction coefficient for dihydroxybenzoic acid (DHBA) at pH 3.0 (Rodríguez-López et al., 1991). The absorbance at 380 nm was used to estimate the levels of azotobactin in culture supernatants using the reported extinction coefficients [λmax = 380 nm at pH 4.0, ε = 23 500 M−1 cm−1 (Page et al., 1991)] and catechols [λmax = 310 nm at pH 3.0, ε = 9200 M−1 cm−1 (Rodríguez-López et al., 1991)].
Algae growth medium and conditions
Algae were maintained on a derivative of Bold's basal medium (Bold, 1949) containing 25 mg L−1 NaCl, 75 mg L−1 MgSO4·7H2O, 25 mg L−1 CaCl2·2H2O, 100 mg L−1 Na2SO4, 300 mg L−1 K2HPO4, 600 mg L−1 NaNO3, and 5 mg L−1 FeSO4·7H2O, adjusted to pH 7.6. The medium was further supplemented with 1 mL L−1 of trace metals solution containing 1.0 g L−1 boric acid, 1 g L−1 sodium EDTA, 200 mg L−1 MnCl2·4H2O, 20 mg L−1 ZnCl2, 15 mg L−1 CuCl2·2H2O, 15 mg L−1 Na2MoO4·2H2O, and 15 mg L−1 CoCl2·6H2O. Cultures of S. sp. BA032 and N. oleoabundans were maintained on agar plates of the modified medium listed above.
Siderophores isolation from A. vinelandii
Cultures of A. vinelandii wild-type cells were cultured on SPB medium. Following growth, cells were centrifuged at 7000 g for 10 min. Supernatant was separated from the cells, then filtered (0.22 μm pore size, Nalgene Filtration, Thermo Scientific), and collected. The filtered supernatant was added to a column containing Q-Sepharose (GE Healthcare). The material retained by the column was rinsed with distilled water and eluted with 500 mM NaCl in distilled water. The sample of crude concentrated siderophores was substituted for sodium nitrate in the medium described above based on Bold's basal medium (Bold, 1949) to screen algae strains for potential growth on agar plates.
Algae growth with siderophores
To determine the ability of specific algae strains to grow on A. vinelandii siderophores, A. vinelandii wild-type and gene substitution strains were first grown on SPB medium for 3 days, then cells were removed by centrifugation, and the obtained supernatant was filtered as described above into an autoclaved container. The supernatant was transferred aseptically to a clean Erlenmeyer flask, and an equivalent quantity of algal cells was inoculated into each flask. Cells were grown under a bank of fluorescent lights (c. 200 µmols s−1 m−2) with constant agitation for c. 8 days while counting the algal cells using a hemocytometer.
Results and discussion
Identification of algae capable of growing on siderophores as a nitrogen source
A primary objective of this work was to determine whether A. vinelandii-derived siderophores could serve as a suitable shuttle of nitrogen from A. vinelandii to nondiazotrophic photosynthetic species such as algae. It has been reported previously that certain species are known to take advantage of siderophore-producing bacteria by utilizing foreign siderophores for metal uptake (Amin et al., 2009; D'Onofrio et al., 2010). To determine whether this feature could also satisfy fixed nitrogen requirements for a phototroph, we first isolated siderophores from media of A. vinelandii using a simple ion-exchange resin to collect the siderophores from spent media as described in the methods. Then, 18 green algae strains were screened on a simple Bold's basal media where sodium nitrate was replaced with the isolated siderophores. Two strains were able to replicate on the siderophore-containing media. The first was a strain of Scenedesmus sp. BA032 isolated from the Cache Valley in northern Utah, and the second was Neochloris oleoabundans, a strain of interest for potential neutral lipid production (Li et al., 2008; Gouveia et al., 2009). These two strains were selected for further siderophore utilization studies as described below.
Identification of specific class of siderophores supporting algae growth
The siderophores produced by A. vinelandii fall into two different classes called the catechol and pyoverdine siderophores (Fig. 1). Others have recently reported the successful disruption of siderophore biosynthetic genes in A. vinelandii (Yoneyama et al., 2011). Here, we utilized a similar approach to what was taken previously by targeting a key gene in each of the two pathways that produce either class of siderophore (Tindale et al., 2000; Yoneyama et al., 2011). Gene substitution strains were constructed and the phenotype (deficient in the production of either class of siderophore) was confirmed by measuring the absorbance of spent culture media (Fig. 2), confirming the results reported previously by others (Yoneyama et al., 2011). These gene substitution strains are important for several reasons. First, there are two classes of siderophores produced by A. vinelandii, and disruption of either class should allow us to determine whether algae are growing on either class specifically. Second, siderophores are not the only extracellular nitrogen compound produced by A. vinelandii. Azotobacter vinelandii also produces other compounds, including ammonia, urea, or extracellular proteins (Bali et al., 1992; Brewin et al., 1999; Gimmestad et al., 2006), each of which could serve as sources of nitrogen to support algae growth. Thus, if the algae strain were able to grow on both substitution strains, this might indicate that another compound was responsible for providing the nitrogen, as has been shown for alternative algae strains (Ortiz-Marquez et al., 2012).
Growth of S. sp. BA032 in liquid culture containing siderophore
To test whether a specific siderophore class was responsible for the growth of S. sp. BA032, in the first experiment, A. vinelandii wild-type strain and the two single-gene substitution strains were first grown in SPB media, which results in the production of siderophores (Fekete et al., 1983; Huyer & Page, 1988). Culture supernatant was separated from A. vinelandii cells as described in the methods, and an equivalent quantity of S. sp. BA032 cells was inoculated into each supernatant sample for culture. Figure 3 (top) shows the results of this experiment along with a positive control (SPB media supplemented with nitrate, but not subjected to A. vinelandii cells) and a negative control (SPB media without added nitrogen sources, not subjected to A. vinelandii cells). These controls were compared to spent SPB media from the specific A. vinelandii strains following several days of growth. The experiment demonstrated that supernatants of A. vinelandii wild-type strain and the catechol siderophore gene substitution strain (A. vinelandii AZBB040) were able to support the growth of S. sp. BA032 cells, while the supernatant from the strain containing a substitution for a pyoverdine (azotobactin) siderophore gene (A. vinelandii AZBB041) had limited growth under similar conditions. This result supports the proposal that azotobactin is providing the source of nitrogen to sustain growth in the media.
Co-culture of A. vinelandii and S. sp. BA032
In addition to using the spent media of A. vinelandii, we also wished to study the potential to grow A. vinelandii and S. sp. BA032 as a co-culture. In this experiment, cells of A. vinelandii wild-type or the individual gene substitution strains were inoculated along with S. sp. BA032 into sterile media and grown together. As shown in Fig. 3 (bottom), a similar result was found to that shown in Fig. 3 (top), where the strain containing the substitution for the gene involved in the biosynthetic pathway to produce azotobactin (A. vinelandii AZBB041) resulted in very minimal growth, while A. vinelandii wild-type cells reached a similar level of growth as was found when using filtered supernatants. The strain containing the gene substitution of catechol siderophores (A. vinelandii AZBB040) was able to support S. sp. BA032, although the levels of cells obtained in the culture were not as high as when only isolated supernatant was utilized.
Growth of N. oleoabundans in liquid culture containing siderophore
The second strain that grew well on plates supplemented with A. vinelandii siderophores as the primary nitrogen source was N. oleoabundans. As was found for S. sp. BA032, supernatants of the wild-type and catechol siderophore gene substitution strains (A. vinelandii AZBB040) were able to support the growth of N. oleoabundans to comparable cell densities, while the supernatant of the strain carrying the azotobactin gene substitution resulted in a limited growth (Fig. 4). These results further support the likelihood that azotobactin is responsible for meeting the nitrogen requirements in N. oleoabundans as well.
Quantities of siderophore available to culture algae
The production of siderophores in A. vinelandii has been studied for many years and is well characterized, including the enzymatic pathways and genes involved in their biosynthesis (Tindale et al., 2000; Page et al., 2003; Wichard et al., 2009; Yoneyama et al., 2011). In many reports, the production of specific siderophores under laboratory culture conditions has been found to be very high [40–60 mg L−1 of various azotobactin forms (Demange et al., 1988)]. Based on the absorbance values obtained in these studies (See Fig. 2 as an example) and a molecular weight of c. 1370 g mol−1, our yields of azotobactin were calculated to be around 25–50 mg L−1 in various cultures, which agrees well with previous reports (Demange et al., 1988). Based on the molecular formula of azotobactin 87 (C55H79N14O27), this constitutes a considerable number of nitrogen equivalents. From the starting quantities of azotobactin in supernatants used to grow S. sp. BA032 (c. 25 µM for A. vinelandii wild-type strain and AZBB040 gene substitution strain, Fig. 3 top) and the positive nitrogen control containing 700 µM of sodium nitrate, the number of cells obtained with S. sp. BA032 correlated well with the amount of nitrogen provided from either source, indicating that S. sp. BA032 was able to utilize essentially all of the provided azotobactin to support growth. Thus, siderophores constitute a source of nitrogen for the proliferation of nondiazotrophic microalgae organisms when excreted into the extracellular space. Azotobactin is composed of a small peptide chain (Fig. 1) synthesized via a nonribosomal peptide synthetase (Yoneyama et al., 2011). Due to the extensive number of nitrogen atoms found in an azotobactin molecule, even smaller molar concentrations of this compound can account for a significant mass amount of available nitrogen when released to the environment. Based on the potential for diffusion and previous reports of other organisms hijacking foreign siderophores for their own benefit (Amin et al., 2009; D'Onofrio et al., 2010), it was of interest to demonstrate whether siderophores from Azotobacter would also serve to satisfy nitrogen requirements to support other organisms in the environment.
Azotobacter is a bacterium predominantly associated with soils, while Scenedesmus is a ubiquitous green algae found in freshwater environments. Thus, there are some differences between the conditions where these strains are found. The light utilized to culture the algae strains could also potentially lead to photodegradation of the siderophores. Even if this were the case, the final fate of the nitrogen appears to be suitable for the culture of the algae strains selected here, which represent only a small sampling of potential strains of phototrophs from the environment that might make use of the excreted nitrogen siderophore compounds provided by a diazotroph.
Implications of findings
Under the current experimental design, the cultures tested here are best defined as commensal co-cultures and not symbiotic, as A. vinelandii is providing a nitrogen source, but is not getting anything in return from the algal strain. Azotobacter vinelandii grows on simple media requiring only sugar and minor amounts of additional minerals. Because the sugar provided to support A. vinelandii is an external source, this relationship is currently an open-loop system. However, many strains of algae and cyanobacteria are known to produce extracellular reduced carbon compounds that could support the growth of A. vinelandii (Brechignac & Schiller, 1992; Wolf, 1997; De Philippis & Vincenzini, 1998; Corzo et al., 2000; Bar-Zeev et al., 2009) and could be used as a secondary selection pressure. Under such conditions, the two strains could be grown in such a manner that a symbiotic relationship might evolve. Examples of phototrophs producing extracellular carbon that is released to the environment in the form of polysaccharides and simple sugars are common, although the exact reason for the energetically wasteful excretion of extracellular reduced carbon is not clear (Passow, 2002; Bar-Zeev et al., 2009). This research provides clear evidence that the nitrogen released to the environment as siderophores can be utilized by other organisms not only for obtaining metals from the environment (D'Onofrio et al., 2010), but also to provide requisite nitrogen required for growth and replication in some strains of green algae.
This work is supported by a grant (RC-0007-12) from the Initiative for Renewable Energy & the Environment (Institute on the Environment) to B.M.B and the Biotechnology Institute at the University of Minnesota for fellowship funding to J.A.V. We thank Jiashi Wei for assistance in early isolation of siderophores and preliminary algal culture studies. We thank the kind suggestions of the anonymous reviewers.