key: cord-0293866-lpjv2exc
authors: Fedeson, Derek T.; Saake, Pia; Calero, Patricia; Nikel, Pablo Iván; Ducat, Daniel C.
title: Biotransformation of 2,4-dinitrotoluene in a phototrophic co-culture of engineered Synechococcus elongatus and Pseudomonas putida
date: 2018-08-31
journal: bioRxiv
DOI: 10.1101/404988
sha: ad7ad52652281146a05d72876a5a4adff847e157
doc_id: 293866
cord_uid: lpjv2exc

In contrast to the current paradigm of using microbial monocultures in most biotechnological applications, increasing efforts are being directed towards engineering mixed-species consortia to perform functions that are difficult to program into individual strains. Additionally, the division of labor between specialist species found in natural consortia can lead to increased catalytic efficiency and stability relative to a monoculture or a community composed of generalists. In this work, we have designed a synthetic co-culture for phototrophic degradation of xenobiotics, composed of a cyanobacterium, (Synechococcus elongatus PCC 7942) and a heterotrophic bacterium (Pseudomonas putida EM173). Cyanobacteria fix CO2 through photosynthetic metabolism and secrete sufficient carbohydrates to support the growth and active metabolism of P. putida, which has been engineered to consume sucrose as the only carbon source and to degrade the environmental pollutant 2,4-dinitrotoluene (2,4-DNT). The synthetic consortium is able to degrade 2,4-DNT with only light and CO2 as inputs for the system, and it was stable over time through repeated backdilutions. Furthermore, cycling this consortium through low nitrogen medium promoted the accumulation of polyhydroxyalkanoate (PHA)–an added-value biopolymer–in P. putida, thus highlighting the versatility of this production platform. Altogether, the synthetic consortium allows for simultaneous bioproduction of PHA and remediation of the industrial pollutant 2,4-DNT, using light and CO2 as inputs. Importance In this study, we have created an artificial consortium composed of two bacterial species that enables the degradation of the industrially-produced environmental pollutant 2,4-DNT while simultaneously producing PHA bioplastic. In these co-cultures, the photosynthetic cyanobacteria fuel an engineered P. putida strain programmed both to use sucrose as a carbon source and to perform the biotransformation of 2,4-DNT. The division of labor in this synthetic co-culture is reminiscent of that commonly observed in microbial communities and represents a proof-of-principle example of how artificial consortia can be employed for bioremediation purposes. Furthermore, this co-culture system enabled the utilization of freshwater sources that could not be utilized in classical agriculture settings, reducing the potential competition of this alternative method of bioproduction with current agricultural practices, as well as remediation of contaminated water streams.

culture for phototrophic degradation of xenobiotics, composed of a cyanobacterium, 23 (Synechococcus elongatus PCC 7942) and a heterotrophic bacterium (Pseudomonas putida 24 EM173). Cyanobacteria fix CO2 through photosynthetic metabolism and secrete sufficient 25 carbohydrates to support the growth and active metabolism of P. putida, which has been 26 engineered to consume sucrose as the only carbon source and to degrade the 27 environmental pollutant 2,4-dinitrotoluene (2,4-DNT). The synthetic consortium is able to 28 degrade 2,4-DNT with only light and CO2 as inputs for the system, and it was stable over 29 time through repeated backdilutions. Furthermore, cycling this consortium through low 30 nitrogen medium promoted the accumulation of polyhydroxyalkanoate (PHA)-an added-31 value biopolymer-in P. putida, thus highlighting the versatility of this production platform. 32

Altogether, the synthetic consortium allows for simultaneous bioproduction of PHA and 33 remediation of the industrial pollutant 2,4-DNT, using light and CO2 as inputs. 34 35 Importance: 36

In this study, we have created an artificial consortium composed of two bacterial 37 species that enables the degradation of the industrially-produced environmental pollutant 38 2,4-DNT while simultaneously producing PHA bioplastic. In these co-cultures, the 39 photosynthetic cyanobacteria fuel an engineered P. putida strain programmed both to use 40

In nature, bacteria typically co-exist in communities with hundreds to thousands of 52 other microorganisms, creating a complex web of inter-species metabolic reactions (1-4). 53

Most consortia exhibit a high degree of "division of labor," where individual species have 54 specialized metabolisms and exchange metabolites and signals with neighbors (5, 6). 55

Interactions range from the cooperative degradation of toluene (7, 8) to the consumption 56 of metabolic waste products (9). Compartmentalization of metabolism across distinct 57 species can confer metabolic capabilities on a consortium that may be difficult to engineer 58 within any one individual. Additionally, natural microbial consortia frequently exhibit a 59 high degree of robustness in the face of dynamic environmental conditions and are resilient 60 to invasive microbes (10, 11). Thus, there has been increasing interest in rationally 61 engineering microbial consortia for desired outputs by dividing metabolic pathways across 62 species (12). 63 7 PHA as a secondary function. This work is proof of principle for the use of synthetic 139 cyanobacteria/heterotroph consortia for combined bioremediation and bioproduction 140 applications. 141

Alginate-encapsulated S. elongatus CscB can tolerate 2,4-DNT at higher 143 concentrations than planktonic cultures 144

As 2,4-DNT is known to be highly toxic to a range of biological organisms (42, 43), Figure 2A ). The toxicity of this compound at these concentrations is highly relevant as 152 leachates from soil contaminated with TNT and its breakdown products have been 153 recorded as high as 98 µM (44). Chl a concentration is commonly measured as a proxy for 154 physiological stress in cyanobacteria as chlorophyll a concentrations are strongly 155 influenced by a variety of stress conditions (45) (46) (47) . Cultures at concentrations of 2,4-DNT 156 ranging from 31 μM to 125 μM exhibited an overall loss of Chl a ( Figure 2B ). This 157 progressive loss of Chl a aligned with visual chlorosis and bleaching of these cultures; this 158 observation was consistent at multiple 2,4-DNT concentrations and at higher starting 159 culture densities ( Figure S1 in the Supplemental Materials). These preliminary experiments 160 indicated that planktonic cyanobacteria are highly sensitive to even low concentrations of 161 8 2,4-DNT, which would complicate their ability to be engineered to directly degrade this 162 nitroaromatic. Further, the viability loss of planktonic S. elongatus CscB would need to be 163 mitigated for any co-culture applications targeting the remediation of these compounds. 164

In previous work (21), we utilized alginate hydrogel encapsulation of S. elongatus 165 CscB to stabilize a co-culture with Halomonas boliviensis under prolonged nitrogen stress 166 conditions. Encapsulation has been used for the immobilization of a variety of cell types, 167 and has often led to increased stress tolerance, cell longevity, and metabolic flux toward 168 target end products (48-53). Alginate-encapsulated S. elongatus CscB cells did not exhibit 169 chlorosis in the presence of 2,4-DNT, even when the concentration was raised to 250 μM, 170 near the solubility limit for this compound ( Figure S2 in the Supplemental Materials). We 171 then exposed encapsulated S. elongatus CscB cells to 2,4-DNT at 250 μM for 7 days while 172 simultaneously inducing expression of the CscB exporter. Chl a was extracted from the 173 beads and measured via spectrophotometry ( Figure 2C ). The data show that while the 174 induction of cscB expression leads to a slight decrease in relative Chl a levels, the addition 175 of 2,4-DNT to alginate-encapsulated cyanobacteria did not further decrease Chl a 176 concentration. Similarly, the Chl a concentration per cell was maintained at a level similar 177 to that of planktonic cells grown under our laboratory conditions ( Figure S3 in the 178 Supplemental Materials). Finally, we measured sucrose export rates from IPTG-induced, 179 encapsulated S. elongatus CscB in the presence of increasing 2,4-DNT concentrations 180 ( Figure 2D ). Sucrose export was maintained for multiple days despite exposure to 125 or 181 250 µM 2,4-DNT. Altogether, alginate encapsulation appears to stabilize S. elongatus when 182 exposed to high levels of 2,4-DNT over long time periods.

Engineering P. putida EM173 for sucrose consumption and evaluation of growth 184 parameters in the presence of alginate-encapsulated S. elongatus CscB 185

We next set to construct strains of P. putida strains (i) that can utilize sucrose as the 186 only carbon source and (ii) capable of degrading 2,4-DNT. P. putida does not normally 187 utilize sucrose as a carbon substrate. The specific strains we used are derivatives of the 188 genetically-tractable, prophage-less P. putida strain EM173 (54). To enable sucrose 189 consumption by P. putida EM173, we first transformed this strain with plasmid pSEVA221-190 cscRABY (55, 56), bearing the sucrose utilization genes from P. protegens Pf-5 ( Figure 3A) . declined in a near linear fashion from 20 g/L to < 3 g/L at 24 h. As sucrose was the only 209 added carbon source and we could clearly observe the consumption of sucrose over time, 210 these data definitively demonstrate that the pSEVA221-cscRABY vector enabled sucrose 211 utilization in these two strains. 212

We cultured our doubly-modified P. putida EM·DNT·S in the presence of a range of 213 sucrose concentrations to determine if this strain is capable of growing on a minimal 214 medium designed for cyanobacterial growth with sucrose as a sole carbon source. For this 215 purpose, we generated a phosphate buffered minimal medium derived from BG-11, herein 216 referred to as M3 medium (Table S1 in the Supplemental Material). To gauge the growth 217 capacity of P. putida EM·DNT·S under these conditions, we inoculated M3 medium that did 218 not contain a carbon source, and incubated cells overnight. This promoted acclimation to 219 the medium and depletion of internal carbon storage compounds that could confound the 220 accurate determination of growth in the M3 medium. These cells were washed with fresh 221 M3 medium before being inoculated into culture flasks with a range of sucrose 222 concentrations (0 g/L to 10 g/L) and growth was tracked for 54 h ( Figure 3C ). Bacterial 223 growth was evident at sucrose concentrations ranging from 1.25-10 g/L ( Figure 3C ), 224 though carbon may have been growth-limiting at concentrations lower than 2.5 g/L ( Figure  225 3C). These results confirm that the heterologous expression of the cscRABY genes from P. 226

protegens Pf-5 is sufficient to confer sucrose utilization on P. putida in our background 227 strain bearing the 2,4-DNT degradation gene cluster, and indicated that the engineered P. 228 putida strain can grow in the cyanobacterial M3 medium-setting the basis for conducting 229 co-cultures. 230

Growth of engineered P. putida strains is supported by sucrose-rich 231 cyanobacterial exudates in a synthetic consortium system 232 We next explored how the engineered P. putida strains (P. putida EM·S and P. putida 233 EM·DNT·S) behaved in co-culture. The P. putida strains, grown overnight in M3 medium 234 with 20 g/L sucrose, were inoculated at OD600 ~ 0.1 into culture flasks containing either 235 empty alginate beads or alginate beads with encapsulated S. elongatus CscB ( Figure 3D 15 demonstrated a similar color change in the culture supernatant to that of P. putida 329 EM·DNT·S monocultures, indicating that the exogenous pathway retained its function in the 330 consortium system. These results demonstrate that this system can provide a directed 331 method for the photosynthetically-driven degradation of 2,4-DNT. Subsequent LC-MS 332 testing of co-culture supernatants shortly (1 h) and 24 h after inoculation of the co-culture 333 with 2,4-DNT revealed that, as in previous monoculture experiments, the P. putida 334 EM·DNT·S co-cultures accumulated 4M5NC ( Figure 5C ). This observation, again, indicates 335 that 2,4-DNT is being degraded via the oxidative pathways in these co-cultures. 336

We initiated a longer duration co-culture in which we inoculated both strains of P. Supplemental Materials) and maintain consistent production of sucrose after culture back 367 dilution. Encapsulated cyanobacteria tolerated concentrations of 2,4-DNT 5 times higher 368 than could be tolerated by planktonic cyanobacterial cultures and did so without 369 significantly changing chlorophyll a content ( Figure 2BC ). While there is precedence for the 370 improved resilience of cells that are encapsulated within hydrogels (21, 59), the 371 mechanism by which this occurs has yet to be elucidated. Future work to investigate how 372 encapsulation modulates cyanobacterium stress response to nitroaromatic compounds 373 could yield genetic targets that could be modified to bolster the resilience of the 374 cyanobacteria without the need for mechanical encapsulation. 375 P. putida is a gram-negative soil bacterium that has recently gained significant 376 attention as a chassis for industrially-relevant synthetic biology. This is in part thanks to the 377 full sequencing (40) and subsequent generation of P. putida strains (e.g., P. putida EM173 378 used in this work) with reduced genomes that demonstrate enhanced expression of 379 heterologous proteins (54). P. putida's metabolic diversity and high tolerance of oxidative 380 stress make it an ideal model organism for studying toxin remediation as well as 381 bioproduction of added-value compounds (41) . The introduced oxidative 2,4-DNT 382 degradation pathway ( Figure 4A ) is chromosomally integrated in this strain. While this 383 pathway avoids generating highly reactive intermediates with hydroxylamino groups, the 384 proteins in this pathway are not yet fully optimized for this new substrate (38, 57), leading 385 to the production of oxidative stress in the Burkholderia sp. R34 from which the pathway 386 originates (38). This oxidative damage is thought to contribute to an increased rate of 387 mutation and fosters a more rapid evolution of this strain to combat oxidative stress (38). 388

While P. putida has also been shown to experience increased oxidative stress when actively 389 utilizing this pathway, this species does not exhibit the same rate of DNA damage and 390 mutation (37). Thus, the physiological properties of P. putida allow for more efficient use of 391 an imperfect 2,4-DNT degradation pathway. Evolving or engineering this pathway toward 392 increased specificity for 2,4-DNT could allow for improved kinetics and reduced ROS 393 generation. Conversely, the lower substrate specificity might allow this pathway to be 394 wastewater. Here, the full media exchanges associated with long-term flask-based co-418 cultures allowed us to demonstrate the ability of these co-cultures not only to remediate 419 this compound but also produce the bioplastic precursor polyhydroxyalkanoate (PHA). 420

While sustainable production of PHA has been pursued in other contexts, this is, to 421 our knowledge, the first report in which PHA formation has been concurrent with the 422 degradation of a xenobiotic compound. In comparison to an independent report that aimed 423 to optimize PHA production from batch cultures of S. elongatus-P. putida (17) , we achieved 424 a lower specific productivity in cultures simultaneously degrading 2,4-DNT. Increased 425 productivities of PHA might be achieved by modifying total nitrogen supplied, duration of 426 nitrogen starvation, or concentration of 2,4-DNT. Of note, we observed no appreciable 427 difference in P. putida growth rates between co-cultures in nitrogen replete or in low 428 nitrogen (2mM nitrate). This raises the possibility that P. putida may be able to utilize an 429 unknown cyanobacterial by-product as a nitrogen source, and additional optimization may 430 be required to fully activate PHA production pathways. 431

One question that arose as part of this work was whether the presence of P. putida 432 EM·DNT·S provides a protective effect to S. elongatus CscB in co-cultures fed with 2,4-DNT 433 contaminated medium. If this were the case, it would shift this relationship from a 434 commensal to a more mutualistic relationship where each species benefits from the 435 presence of the other. Pursuing longer-term cultures with even more rigorous exposure to 436 2,4-DNT could reveal whether P. putida EM·DNT·S might provide such a protective effect. 437

We show that these artificial co-cultures are not only capable of utilizing media 438 contaminated with a toxic xenobiotic, but also of producing the bioplastic PHA. A scaled 439 version of this system could hypothetically take wastewater effluent from industrial 440 sources contaminated with 2,4-DNT, remediate the water allowing it to be utilized for other 441 functions, and provide a mechanism by which PHA could be produced. The more 442 immediately relevant take-away from this work is that photosynthetic co-cultures utilizing 443 Table S1 in the 460 Supplemental Material) with either 20 g/L or 2 g/L sucrose, as indicated. These cultures 461 were then used as inoculum for subsequent experiments. Antibiotic selection was omitted 462 for all co-culture experiments. 463

Alginate encapsulation was performed as previously described (21) 

Rules of 557 Engagement: Interspecies Interactions that Regulate Microbial Communities

Diversity of Industrially Relevant Microbes

Synthetic microbial ecology 562 and the dynamic interplay between microbial genotypes

Division of Labor in Biofilms : the 568 Ecology of Cell Differentiation

Unraveling interactions in microbial 570

Photoautotrophic production of polyhydroxyalkanoates in a synthetic mixed 597 culture of Synechococcus elongatus cscB and Pseudomonas putida cscAB

A Designed A. vinelandii−S. elongatus Coculture 600 for Chemical Photoproduction from Air, Water, Phosphate, and Trace Metals

Improving metabolite production in microbial co-603 cultures using a spatially constrained hydrogel

Mimicking lichens: Incorporation of yeast strains together with sucrose-607 secreting cyanobacteria improves survival, growth, ROS removal, and lipid 608 production in a stable mutualistic co-culture production platform

A synthetic , light-driven consortium of 611 cyanobacteria and heterotrophic bacteria enables stable polyhydroxybutyrate 612 production

Biodegradation of Nitroaromatic Compounds

Toxicity and Metabolism of Explosives

Transgenic phytoremediation blasts onto the 622 scene

Chemical and Toxicological 624 Characterization of Slurry Reactor Biotreatment of Explosives-Contaminated 625

Bacterial pathways for degradation of 627 nitroaromatics

Regulation of superoxide stress 634 in Pseudomonas putida KT2440 is different from the SoxR paradigm in 635

Mutagenicity of 637 nitroaromatic degradation compounds

Irreversible binding of 639 biologically reduced 2,4,6-trinitrotoluene to soil

Biodegradation 642 of 2,4-dinitrotoluene by a Pseudomonas sp

Biodegradation of 4-Methyl-5

Nitrocatechol by Pseudomonas sp. Strain DNT

Aerobic Degradation of Dinitrotoluenes and 647

Pathway for Bacterial Degradation of Aerobic Degradation of Dinitrotoluenes and 648

Pathway for Bacterial Degradation of 2 , 6-Dinitrotoluene

The 651 metabolic redox regime of Pseudomonas putida tunes its evolvability towards 652 novel xenobiotic substrates

Stress Caused by Faulty Oxidation Reactions Fosters Evolution of 2

Mechanisms of solvent 657 tolerance in gram-negative bacteria

Complete genome sequence and comparative analysis of the 665 metabolically versatile Pseudomonas putida KT2440

From dirt to industrial 668 applications: Pseudomonas putida as a Synthetic Biology chassis for hosting 669 harsh biochemical reactions

Dinitrotoluene in Arabidopsis thaliana : Toxicity , Fate , and Gene Expression 672 Studies in Vitro

Toxicity of 2 , 4-dinitrotoluene to terrestrial 675 plants in natural soils

Chemical characterization and toxicological testing of windrow 678 composts from explosives-contaminated sediments

Nitrogen 681 starvation-induced chlorosis in Synechococcus PCC 7942. Low-level 682 photosynthesis as a mechanism of long-term survival

Oxidative stress in cyanobacteria

Growth and phosphorus 689 removal by Synechococcus elongatus co-immobilized in alginate beads with 690

Azospirillum brasilense

Immobilization 692 of Nicotiana tabacum plant cell suspensions within calcium alginate gel beads for 693 the production of enhanced amounts of scopolin

Applied o . d Microbiology Biotechnology Growth and hydrocarbon 696 production of Botryococcus braunii immobilized in calcium alginate gel

Production With Immobilized Cells of Strep-tomyces marinensis Nuv-5 in Calcium 699

Growth of Chlamydomonas reinhardtii in acetate-free medium when co-cultured

Extended H2photoproduction by N2-fixing 705 cyanobacteria immobilized in thin alginate films

Freeing 708

Pseudomonas putida KT2440 of its proviral load strengthens endurance to 709 environmental stresses

Engineering sucrose 711 metabolism in Pseudomonas putida highlights the importance of porins

Metabolic engineering to expand the substrate spectrum of Pseudomonas putida 715 toward sucrose

Association of dnt genes of 717

DNT with the substrate-blind regulator DntR draws the 718 evolutionary itinerary of 2,4-dinitrotoluene biodegradation

Endogenous Stress Caused by Faulty Oxidation 721

Reactions Fosters Evolution of 2 , 4-Dinitrotoluene-Degrading Bacteria 9

The Use of Immobilization in Alginate Beads 723 for Long-Term Storage of Pseudanabaena galeata (Cyanobacteria) in the

Bioremediation of 2,4-dinitrotoluene (2,4-DNT) in 726 immobilized micro-organism biological filter

Reductive removal of 2,4-dinitrotoluene and 2,4-728 dichlorophenol with zero-valent iron-included biochar

Aerobic Transformation of 733 2,4-Dinitrotoluene by Escherichia coli and Its Implications for the Detection of 734

Figure 1: Conceptual design of the photosynthetic co-culture designed for 738 simultaneous biodegradation and bioproduction

Alginate-encapsulated S. elongatus CscB is co-cultured with P. putida EM!DNT!S

DNT) while simultaneously producing the bioplastic polyhydroxyalkanoate (PHA)

elongatus CscB embedded within alginate beads fix carbon dioxide via the Calvin-Benson 743

Fixed carbon is converted into 744 sucrose that is exported through heterologous sucrose permease (CscB) in to the culture 745 medium

can convert and store sucrose as PHA, making this system multifunctional

Figure 2: Toxic effects of 2,4-DNT exposure on planktonic and alginate-encapsulated

A) Varied concentrations of 2,4-DNT were added to planktonic cultures of S. elongatus 753

Chlorophyll a 754 content of cultures in (a) were measured and normalized to control cultures with no added 755 2,4-DNT. (C) Chlorophyll a content of alginate-encapsulated S. elongatus CscB cells was 756 measured after incubation for 5 days in M3 media. IPTG and/or 250 µM 2,4-DNT were 757 added as indicated and chlorophyll a values were normalized to control cultures without 758 these additives. (D) Total sucrose concentration in the culture supernatant of induced (1 759 mM IPTG) alginate-encapsulated S

For the experiments shown in (A-C), the mean values for n = 3 are 761 indicated, and error bars represent standard deviations

the mean values for n = 3 with 3 technical replicates per condition are indicated, and 763 error bars represent standard deviations

Growth characterization of sucrose-consuming P. putida EM·DNT·S strains 766 (A) Schematic representations of the 2,4-DNT degradation gene cluster from Burkholderia 767

R34 (top) and the gene cassette for sucrose utilization in pSEVA221-cscRABY (bottom)

Growth of kinetics and sucrose utilization of the P. putida strains in 20 g/L of sucrose in 769 M9 minimal medium. n = 4; error bars represent standard deviations

M3 771 medium. (D) P. putida strains EM·S and P. putida EM·DNT·S were grown in M3 media with

OD600 ~ 0.1 into flasks containing alginate 773 beads with or without encapsulated S. elongatus CscB, all cultures contained 1 mM IPTG for 774 induction of sucrose export. OD600 measurements were taken over the course of 216 hours, 775 tracking the growth of the P. putida strains in the co-culture. At 96 h post-inoculation, all of 776 the M3 media was removed and replaced with fresh medium

For the 778 experiments shown in (C) and (D), the mean values for n = 3 are indicated, and error bars 779 represent standard deviations

Figure 4: 2,4-DNT biotransformation in monocultures of engineered P. putida 783 (A) Exogenous pathway for the oxidative degradation of 2,4-DNT. (B) (Left) Representative 784 culture images of P. putida EM!DNT!S and P. putida EM!S grown over 22 h in M3 medium 785 with 20 g/L sucrose in the presence of 250 μM 2,4-DNT. (Right) Zoom-in of cultures 786 demonstrating characteristic changes in pigmentation of culture supernatants

Averaged spectral signatures (n=3) of supernatants from P. putida EM!DNT!S cultures 788 grown as in M3 medium with 2 g/L sucrose in the presence of 250 μM 2,4-DNT. Samples 789 were monitored via scanning spectrophotometry at the indicated time points

Representative LC/MS elution profile of P. putida EM!DNT!S (top) and P. putida EM!S 791 (bottom) supernatant after 4 hours of incubation M3

μM 2,4-DNT. 4M5NC elutes from the column after 2.3 min in negative ion mode

Quantification of 4M5NC in supernatants of both P. putida EM!S and P. putida EM!DNT!S 794 monocultures at the 4 h via LC/MS. P. putida EM!S strain did not generate any detectable 795 amount of 4M5NC, while the P. putida EM!DNT!S strain accumulated 4M5NC

Holliston, MA) was used to dispense the solution dropwise through a 30 G needle into a 472 ≥20-fold larger volume of 20 mM BaCl2. The drops traveled ~35 cm from needle to the 473 slowly stirred BaCl2 solution and were allowed to cure in the solution for at least 20 min. 474Solidified beads were rinsed once with BG-11 medium and incubated overnight in M3 475 medium (minus the 100 mM NaCl). To acclimate cells within the alginate and precipitate 476 residual Ba 2+ , beads were transferred through a series of media washes in following days. 477The first day post-encapsulation, beads were rinsed and resuspended in fresh M3 medium 478 (without NaCl), transferred to 250-mL baffled Erlenmeyer flasks, and placed into a 479