Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

High titer heterologous rhamnolipid production

High titer heterologous rhamnolipid production Heterologous mono-rhamnolipid production by Pseudomonas putida KT2440 pSynPro8oT_rhlAB using glucose as the single carbon source was characterized in fed-batch bioreactor cultivations. For the described experiments, a defined mineral salt medium was used, and a two phase glucose feeding profile was applied, which yielded a final rhamnolipid concentration of 14.9 g/L. Applying the feeding profile, glucose stayed almost constant until 28 h of cultivation and decreased afterwards to limiting levels. Until the end of cultivation 253.0 ± 0.1 g glucose was added to the bioreactor of which a total of 252.0 ± 0.6 g glucose was metabolized. By modeling the fed-batch bioreactor cultivations the time courses of generated biomass, rhamnolipid and consumed glucose were described. The model was furthermore used to derive key process parameters from the collected data. The obtained values for the specific product formation rates (q ) reached 18 mg/(g h) and yield coefficients (Y ) 10 mg/g respectively. RL RL/S Keywords: Heterologous rhamnolipid biosynthesis, Pseudomonas putida, Biosurfactant, Fed batch cultivation carbon source that avoids the difficulties in using vegeta - Introduction ble oil used as a carbon source for P. aeruginosa (Müller Low molecular weight amphiphilic compounds secreted and Hausmann 2011). However, reported heterologous by several microorganisms are often termed biosur- rhamnolipid production with maximal product concen- factants. These microbial surfactants exhibit diverse tration of 7.3 g/L (Cha et al. 2008) is by far not convinc- structures, are generally assumed to be easily biodegrad- ing in comparison to that reported from P. aeruginosa. As able, to display a non-toxic character and often show bio- stated before, heterologous rhamnolipid production in active properties (Hausmann and Syldatk 2014). Despite Pseudomonas putida KT2440 is not naturally regulated these advantages microbial surfactants are so far not by quorum sensing leading to constitutive rhamnolipid used in large scale industrial production. Several special- biosynthesis. However, the actual handling of P. putida ized bioreactor concepts for the biosurfactant production rhamnolipid producing strains remains challenging. have been reviewed by Beuker et  al. (2014). The gly - This highly complex bacterial process exhibits multilay - colipid rhamnolipid is one of the most studied microbial ered interactions between glucose metabolization, bio- surfactants (Abdel-Mawgoud et al. 2011). mass growth and product formation. These interactions The heterologous and glucose-based rhamnolipid pro - impose difficulties to determine influences of changes in duction brings several advantages and has been com- glucose feed on the cultivation. To encounter these dif- paratively extensively studied since the 1990s. Firstly the ficulties model based process optimization with several pathogenicity and the complex biosynthesis regulation iteration steps has been successfully employed in the past of the natural producer Pseudomonas aeruginosa can (Kovárová-Kovar et al. 2000). be avoided. Secondly glucose represents a convenient Wittgens et  al. (2011) examined heterologous rhamnolipid production using P.  putida KT42C1 *Correspondence: Marius.Henkel@uni-hohenheim.de pVLT31_rhlAB in glucose enriched LB medium in a Department of Bioprocess Engineering (150k), Institute of Food Science and Biotechnology, University of Hohenheim, Fruwirthstr. 12, baffled shake flask experiment. In their study Witt - 70599 Stuttgart, Germany gens et  al. (2011) described biomass growth using a Full list of author information is available at the end of the article © The Author(s) 2016. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Beuker et al. AMB Expr (2016) 6:124 Page 2 of 7 logistic equation and glucose metabolization as well as contained 100  mL SupM medium in a 1  L baffled shake rhamnolipid formation using biomass dependent ordi- flask inoculated with 1 mL from the 24 h LB culture and nary differential equations (ODEs). In their approach incubated for 12 h. rhamnolipid productivity as well as glucose metaboli- zation exhibited constant values and were not growth Bioreactor cultivations rate dependent. P. putida KT42C1 pVLT31_rhlAB was All bioreactor cultivations were carried out as dupli- reported to produce up to 1.5 g/L of rhamnolipid. cates. The bioreactor (Minifors, HT Infors, Bottmingen, Switzerland) was equipped with an integrated pH, tem- Methods perature and aeration control system. During bioreac- Methods were adapted from foam fractionation pro- tor cultivation aeration was set to 0.133  vvm and pO cesses as described in Beuker et al. (2016). was controlled at 13% via stirring rate starting with a minimum of 300 rpm. Temperature was held constant at Chemicals 30 °C and pH was controlled to 6.8 via 1 M H SO or 19% 2 4 All chemicals used in the current study were purchased NH OH. Generated foam was detected by an antifoam from Carl Roth GmbH (Karlsruhe, Germany) if not probe and antifoaming agent Tego KS 53 (Evonik Indus- stated otherwise. tries) was added if needed. Bioreactors were inoculated with 12 h SupM seed culture to a final OD of 0.5. Initially Microorganism and plasmid 6 g glucose was provided in the culture vessel. After 5  h P. putida KT2440 with plasmid pSynPro8oT_rhlAB pro- of batch fermentation a dual phase feeding profile was ducing mono-rhamnolipids was used as described in started as displayed in Fig. 2b. Beuker et al. (2016). Analytical methods Culture conditions Sampling and processing Media Taken bioreactor samples were mixed with equal vol- Tetracycline was added to all media to an end concentra- umes of hexane and centrifuged (4700 rpm, 15 min, 4 °C) tion of 20  mg/L. For the first culture LB medium (5  g/L to remove antifoaming agent. The hexane phase was dis - yeast extract (BD, Heidelberg, Germany), 10  g/L tryp- carded and the cell free aqueous supernatant was used tone (BD), 5  g/L NaCl; pH 7.0) was utilized. For seed for rhamnolipid, glucose and ammonium detection. The culture SupM medium (4.4 g/L Na HPO ∙2 H O, 1.5 g/L cell pellet was washed with 1:1 (v/v) 9 g/L NaCl solution, 2 4 2 KH PO , 1  g/L NH Cl, 0.2  g/L MgSO ∙7 H O, 0.02  g/L centrifuged (4700 rpm, 15 min, 4 °C), dried at 100 °C and 2 4 4 4 2 CaCl ∙2 H O, 0.006 g/L FeCl , 30 g/L glucose, 10 g/L yeast used for gravimetrical cell dry weight determination. 2 2 3 extract, 1  mL/L trace element solution 2, pH 6.8; trace Rhamnolipid detection was performed as described element solution 2: 0.3 g/L H BO , 0.2 g/L CoCl ∙6 H O, by Schenk et  al. (1995) with minor adjustments. Part 3 3 2 2 0.1  g/L ZnSO ∙7 H O, 0.03  g/L MnCl ∙4 H O, 0.01  g/L of the liquid phase was acidified with 1:100 (v/v) phos - 4 2 2 2 CuCl ∙2 H O, 0.03 g/L Na MoO ∙2 H O, 0.02 g/L NiCl ∙6 phoric acid and rhamnolipids were extracted twice with 2 2 2 4 2 2 H O) was applied. In the bioreactor cultivation ModR 1.25:1 (v/v) ethyl acetate. For rhamnolipid measure- medium (22  g/L KH PO , 2.6  g/L (NH ) HPO , 1.4  g/L ment appropriate amount of this ethyl acetate extract 2 4 4 2 4 MgSO ∙7 H O, 0.87 g/L citric acid, 0.01 g/L FeSO ∙7 H O, was evaporated. Rhamnolipids were resolved in ace- 4 2 4 2 5 g/L glucose, 10 mL/L trace element solution 2, pH 6.8) tonitrile and derivatized for 90  min at 1400  rpm and was used. 15  g/L antifoaming agent Tego KS 53 (Evonik 60  °C using a 1:1 mixture of 40  mM bromphenacylbro- Industries, Essen, Germany) was added to the bioreac- mid and 20  mM tri-ethyl-ammonium/-amin. Detection tor cultivation. Feed medium contained 400 g/L glucose, of rhamnolipids was performed using a HPLC device 22 g/L KH PO , 3.43 g/L KH PO , 0.01 g/L FeSO ∙7 H O (Agilent 1100 Series, Agilent, Waldbronn, Germany) 2 4 2 4 4 2 −3 and 0.1∙10  g/L CuCl ∙2 H O. equipped with a 15  cm reversed phase column (Supel- 2 2 cosil LC-18, Supelco, Deisenhofen, Germany) at 30 °C. Preparation of seed culture The mobile phase was composed of 100% methanol and All shake flasks were inoculated in a shake incuba - ultrapure water. For rhamnolipid detection a gradient tor chamber (Multitron II, HT Infors, Bottmingen, was applied. During the first 17  min methanol concen - Switzerland) at 30  °C and 120  rpm. First 25  mL LB in a tration was increased starting at 80 to 100%. This meth - 100  mL baffled shake flask were inoculated with 50  µL anol concentration was held for 8 min and decreased to from a glycerol stock solution of P. putida KT2440 80% during the next 5 min. Rhamnolipids were detected pSynPro8oT_rhlAB and incubated for 24 h. Seed cultures at a wave length of 254  nm at 30  °C. For calibration Beuker et al. AMB Expr (2016) 6:124 Page 3 of 7 standard solutions of rhamnolipid in acetonitrile were Biomass used. Biomass mass growth was described via an autocatalytic The concentration of glucose and ammonium were process and a jump function to model lag times (Eq.  1). detected from the aqueous phase of samples using glu- The jump function depended on the lag time (τ) and the cose (R-Biopharm AG, Darmstadt, Germany) and autocatalytic process on existing biomass (X) and specific ammonium (Merck KGaA, Darmstadt, Germany) growth rate (µ). assay kits, respectively, according to the manufacturers’ dX 1 instructions. = ∗ μ ∗ X (1) 100∗(−t+τ ) dt 1 + e Software for modeling and parameter optimization The specific growth rate was represented by a combi - Modeling was performed in a mathematical/numeri- nation of Monod and Luong kinetic (Eq.  2). Unhindered cal computing environment (Matlab, The MathWorks, growth was described by Monod kinetic depending on Natick, MA, USA). To solve the ordinary differential the concentration of the growth limiting substrate (c ), Glu equations the numeric solver “ode45” embedded in Mat- a Monod constant (K ) and a maximal growth rate (µ ) S max lab environment was used based on an explicit Runge– (Monod 1949). Inhibiting effects of high glucose concen - Kutta formula. Parameter optimization was performed tration were integrated using Luong kinetic including the using the Matlab embedded function “lsqnonlin” apply- concentration of the growth limiting substrate, an inhibi- ing a trust-region-reflective algorithm, which is based on tory constant (K ) and a shape factor (n) (Luong 1987). the interior reflective Newton method. The parameter c c Glu Glu optimization was executed by minimizing the sum of μ = μ ∗ ∗ 1 − max (2) c + K K S i Glu squares of the errors between simulation data and meas- ured data considering biomass, rhamnolipid and glucose. Therefore, biomass growth was modeled using Eq. 3. Equations for modeling dX 1 Glu/V BR The general model setup is depicted in Fig.  1 outlining = ∗ μ ∗ max 100∗(−t+τ ) dt 1 + e Glu/V + K BR s the interactions between glucose consumption, biomass Glu/V BR (3) growth, rhamnolipid production, feed and volume of the ∗ 1 − ∗ X cultivation broth. All parameters used for modeling are K summarized in Table 1. Fig. 1 Dependencies between volume, feed, glucose, biomass and rhamnolipid; dependent parameter are depicted in framed boxes, dashed boxes depict type of dependencies, arrows show the source(s) and sink of dependencies Beuker et al. AMB Expr (2016) 6:124 Page 4 of 7 Table 1 Parameters for modeling Bioreactor volume For changes in the overall bioreactor volume just changes Parameter Meaning Unit due to the feed medium were considered (Eq.  6). Vol- α* Growth dependent production term – ume addition by acid, base and antifoaming agent −1 β* Growth independent production term h were assumed to level out by volume reduction due to −1 µ Growth rate h sampling. −1 µ * Maximal growth rate h max dV m ˙ −1 BR Feed ρ Density g L (6) dt ρ τ* Lag time h BR Bioreactor – To describe the time course of the bioreactor cultiva- −1 c Concentration of index g L tion the 10 parameter shown in Table  1 were optimized Index Glu Glucose g using the developed software. Optimization was imple- −1 K * Inhibiting concentration of glucose g L mented using measuring data of the double determina- −1 K * Monod constant g L tion cultivation applying a dual phase feeding profile and −1 m ˙ Mass flux g h the set of ODEs (Eq. 3, 4, 5, 6). −1 −1 m * Maintenance coefficient g g L S/X n* Shape factor for Luong kinetic – Results RL Rhamnolipid g Time courses of overall biomass, rhamnolipid and glucose t Time h during bioreactor cultivation V Volume of index L Time courses of generated biomass, rhamnolipid and Index X Biomass g consumed glucose are depicted in Fig.  2. In bioreactor −1 Y * Rhamnolipid yield coefficient g g cultivations applying the dual phase feeding profile glu - P/S −1 Y * Biomass yield coefficient g g cose amount decreased in the beginning, stayed almost X/S constant at 0.2 g from 12 h till 28 h and decreased after- * Parameter determined via optimization wards to limiting levels. Biomass grew constantly with a leveling trend in the end reaching 35.3 ± 3.2 g in the end of cultivation. Rhamnolipid generation started after 24 h Glucose and increased steadily to a maximal value of 22.2 ± 4.3 g The time course of glucose was modeled taking various after 72 h, which equals a concentration of 14.7 ± 2.6 g/L. glucose sinks as well as sources into account (Eq. 4). Glu- Until the end of cultivation 253.0  ±  0.1  g glucose was cose was assumed to be utilized by biomass and rham- added to the bioreactor and in total 252.0 ± 0.6 g glucose nolipid generation as well as by maintenance of the was metabolized. biomass. Conversion efficiencies of glucose to biomass or rhamnolipid were determined by biomass and rham- Parameter optimization nolipid yield coefficients (Y and Y , respectively). X/S P/S To describe the time course of bioreactor cultivations Glucose amount used for maintenance was described by a parameter optimization was conducted using the set a maintenance coefficient (m ). Applied feed (m ˙ ) Feed S/X of ODEs (Eq.  3, 4, 5, 6) and the experimental data. The with its glucose feed concentration (c ) was deter- Glu in Feed results are depicted in Fig. 2 and Table 2. mined as glucose input. Biomass and glucose time course were satisfactorily dGlu dX 1 dRL described using the model with the optimized parameter =˙ m ∗ c − ∗ + Feed Glu in Feed with end values of 33.5 and 0.0 g, respectively. However, dt dt Y dt X/S rhamnolipid production was just described properly in (4) ∗ + m ∗ X S/X the end of cultivation with a final rhamnolipid mass of P/S 23.2  g which equals a concentration of 14.9  ±  0.5  g/L. The overall rhamnolipid production rate reached 0.3 g/h Rhamnolipid and the final production rate 0.4  g/h. The volume of the Rhamnolipid kinetic was described by a growth depend- bioreactor was underestimated after 14 h, because mod- ent (α) and growth independent (β) part of the specific eled volume just considered changes caused by feeding biomass productivity (Eq. 5). medium. dRL In Fig. 3 the model-derived time courses of the growth = (α ∗ μ + β) ∗ X (5) dt rate (µ), the specific rhamnolipid production rate (q ) RL Beuker et al. AMB Expr (2016) 6:124 Page 5 of 7 Fig. 2 Time courses of P. putida KT2440 pSynPro8oT_rhlAB bioreactor cultivation measurement data compared to modeled data. a experimentally determined (black circles) and modeled total biomass time courses (black line), b experimentally determined (grey triangles), modeled glucose time courses (black line) and total fed glucose amount (grey dashed line), c measured rhamnolipid (empty squares) and modeled rhamnolipid time course (black line), d calculated bioreactor volume at each sampling point (black circles) and modeled bioreactor volume time course (black line) applying parameters depicted in Table 2 Table 2 Optimized model parameters 1/h were computed. Because rhamnolipid amounts in the fermentation broth accumulate only in the fed-batch Parameter Optimized value Unit phase after 14  h the values of the specific rhamnolipid α 0.0477 – production rate and the yield coefficient are only mean - −1 β 0.0125 h ingful after that time. The highest value for the specific −1 µ 0.4386 h max productivity (18  mg/(g  h)) was registered immediately τ 0.4686 h after switching to the feeding of glucose. Thereafter it −1 K 107.3604 g L i steadily reduced its value until reaching 10  mg/(g  h) at −1 K 0.2034 g L S the end of the process. The product yield increased from −1 −1 m 0.0599 g g L S/X 0.075 to 0.104  g/g during the fed-batch phase. This is n 0.6167 – an expected finding as the growth rates in the fed-batch −1 Y 0.2637 g g phase decreased steadily from 0.052 to 0.002 1/h. It may P/S −1 Y 0.3253 g g be assumed that a reduced growth rate leads to a reduced X/S glucose demand derived for cell growth. This may led to a better yield of product formation. and the yield coefficient (Y ) are shown. The model- P/S Discussion derived values for the growth rate show three distinct In this study a bioreactor fed-batch cultivation for het- phases, reflecting the batch phase before 14  h and the erologous rhamnolipid production in a defined medium fed-batch phase thereafter. The highest value for the and a process model was established. As a result up to growth rates in the batch phase were between 0.38 and 14.9  g/L rhamnolipid were obtained. Utilizing a dual 0.19 1/h while in the fed-batch phase values below 0.052 Beuker et al. AMB Expr (2016) 6:124 Page 6 of 7 metabolic capacity of the construct P. putida KT2440 pSynPro8oT_rhlAB. Differences in rhamnolipid amounts in experimental and modeled data were mainly caused by a lag time in rhamnolipid production in the beginning of cultivation. This lag time did not occur in the model. This lag time in rhamnolipid production could be caused by several rea- sons. One explanation could be the availability of one or both rhamnolipid precursors, which may be impaired or shifted at high growth rates thus resulting in a reduced specific production rate. It is noticeable that rhamnolipid production started when glucose started to be limit- ing corresponding to the onset of the glucose feeding. The recorded value of 18  mg/(g  h) is comparable to the maximum value for the P. putida KT42C1 pVLT31_rhlAB reported by Wittgens et al. (2011) and about 38% for the very similar construct P. putida KT2440 pSynPro8_rhlAB grown on a complex rich medium containing additional glucose (Tiso et  al. 2016). However, the comparison to experimental data reported on complex rich medium must be considered with caution. The established model could be used to describe time courses of biomass and glucose during the overall bio- reactor cultivation and of rhamnolipid in the end of bio- reactor cultivation. Additionally the time courses of the growth rate, the specific productivity and the yield coeffi - cient could be derived. To determine which factors influ - ences biomass as well as rhamnolipid production further experiments should be conducted concerning medium components, e.g. the influence of iron and copper. Addi - tionally, further investigations should be made to deter- mine if other essential components should be added to the cultivation due to depletion by high biomass growth (e.g. sulphur, magnesium, calcium, sodium or trace elements). For future work, strain engineering should focus on enhancing the effective yield that reached a value of 10% in relation to the theoretical maximum value of 49% (Henkel et al. 2012). Therefore, the putative genera - tion of polysaccharides as byproducts should be mainly Fig. 3 Time courses of modeled-derived process parameters for P. addressed. Another approach will be the combination putida KT2440 pSynPro8oT_rhlAB bioreactor cultivation a time course of the strategy presented here using fed-batch process- of growth rate, b time course of specific rhamnolipid production rate, ing, with in  situ product removal by foam fractionation c time course of product yield (Beuker et  al. 2016) and thus avoiding the necessity of chemical antifoaming agents. This could lead to a further enhanced semi-continuous process. phase feeding profile in bioreactor cultivations resulted Authors’ contributions in an overall rhamnolipid productivity of 0.3  g/h and JB planned and executed the experiments, collected and calculated all rhamnolipid titer of 14.9  g/L. So far, in literature the data, created the graphs and figures and drafted this manuscript. AW and FR generated the plasmid pSynPro8_rhlAB. AS generated the plasmid highest rhamnolipid titer reported using heterologeous pSynPro8oT_rhlAB. TB conducted her Bachelor thesis under the supervision production hosts was 7.3  g/L (Cha et  al. 2008). There - of JB and conducted experiments applying the two staged feeding profile. fore, the effectiveness of this bioreactor cultivation apply - MH significantly contributed to data evaluation and interpretation of the experiment as well as graphs and figures. RH substantially contributed to ing the feeding profile was demonstrated as well as the Beuker et al. AMB Expr (2016) 6:124 Page 7 of 7 conception and design of the conducted experiments and contributed in Cha M, Lee N, Kim MM, Kim MM, Lee S (2008) Heterologous production of manuscript writing. All authors read and approved the final manuscript. Pseudomonas aeruginosa EMS1 biosurfactant in Pseudomonas putida. Bioresour Technol 99:2192–2199. doi:10.1016/j.biortech.2007.05.035 Author details Hausmann R, Syldatk C (2014) Types and classification of microbial surfactants. Department of Bioprocess Engineering (150k), Institute of Food Science In: Biosurfactants. CRC Press, Boca Raton, pp 3–18 and Biotechnology, University of Hohenheim, Fruwirthstr. 12, 70599 Stuttgart, Henkel M, Müller MM, Kügler JH, Lovaglio RB, Contiero J, Syldatk C, Hausmann Germany. Ulm Center for Peptide Pharmaceuticals (U-PEP), Ulm University, R (2012) Rhamnolipids as biosurfactants from renewable resources: Albert-Einstein-Allee 11, 89081 Ulm, Germany. Present Address: Evonik Indus- concepts for next-generation rhamnolipid production. Process Biochem tries, Evonik Technology and Infrastructure GmbH, Rodenbacher Chaussee 4, 47:1207–1219. doi:10.1016/j.procbio.2012.04.018 63457 Hanau-Wolfgang, Germany. Kovárová-Kovar K, Gehlen S, Kunze A, Keller T, von Däniken R, Kolb M, van Loon APGM (2000) Application of model-predictive control based on artificial Competing interests neural networks to optimize the fed-batch process for riboflavin produc- The authors declare that they have no competing interests. This article does tion. J Biotechnol 79:39–52. doi:10.1016/S0168-1656(00)00211-X not contain any studies with human participants or animals performed by any Luong JH (1987) Generalization of monod kinetics for analysis of growth data of the authors. with substrate inhibition. Biotechnol Bioeng 29:242–248. doi:10.1002/ bit.260290215 Received: 8 November 2016 Accepted: 1 December 2016 Monod J (1949) The growth of bacterial cultures. Annu Rev Microbiol 3:371–394. doi:10.1146/annurev.mi.03.100149.002103 Müller MM, Hausmann R (2011) Regulatory and metabolic network of rhamnolipid biosynthesis: traditional and advanced engineering towards biotechnological production. Appl Microbiol Biotechnol 91(2):251–264 Schenk T, Schuphan I, Schmidt B (1995) High-performance liquid chromato- References graphic determination of the rhamnolipids produced by Pseudomonas Abdel-Mawgoud AM, Hausmann R, Lépine F, Müller MM, Dèziel E (2011) aeruginosa. J Chromatogr A 693:7–13. doi:10.1016/0021-9673(94)01127-Z Rhamnolipids: detection, analysis, biosynthesis, genetic regulation, and Tiso T, Sabelhaus P, Behrens B, Wittgens A, Rosenau F, Hayen H, Blank LM (2016) bioengineering of production. In: Soberón-Chavez G (ed) Biosurfactants: Creating metabolic demand as an engineering strategy in Pseudomonas from genes to applications. Springer, Berlin putida—Rhamnolipid synthesis as an example. Metab Eng Commun Beuker J, Syldatk C, Hausmann R (2014) Bioreactors for the Production of 3:234–244. doi:10.1016/j.meteno.2016.08.002 biosurfactants. In: Biosurfactants: production and utilization—processes, Wittgens A, Tiso T, Arndt TT, Wenk P, Hemmerich J, Müller C, Wichmann R, Küp- technologies, and economics. CRC Press, Boca Raton, p 117 per B, Zwick M, Wilhelm S, Hausmann R, Syldatk C, Rosenau F, Blank LM Beuker J, Steier A, Wittgens A, Rosenau F, Henkel M, Hausmann R (2016) (2011) Growth independent rhamnolipid production from glucose using Integrated foam fractionation for heterologous rhamnolipid production the non-pathogenic Pseudomonas putida KT2440. Microb Cell Fact 10:80. with recombinant Pseudomonas putida in a bioreactor. AMB Express 6:11. doi:10.1186/1475-2859-10-80 doi:10.1186/s13568-016-0183-2 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png AMB Express Springer Journals

High titer heterologous rhamnolipid production

Download PDF
7 pages

Loading next page...
 
/lp/springer-journals/high-titer-heterologous-rhamnolipid-production-VKovkBawqI

References (14)

Publisher
Springer Journals
Copyright
Copyright © 2016 by The Author(s)
Subject
Life Sciences; Microbiology; Microbial Genetics and Genomics; Biotechnology
eISSN
2191-0855
DOI
10.1186/s13568-016-0298-5
Publisher site
See Article on Publisher Site

Abstract

Heterologous mono-rhamnolipid production by Pseudomonas putida KT2440 pSynPro8oT_rhlAB using glucose as the single carbon source was characterized in fed-batch bioreactor cultivations. For the described experiments, a defined mineral salt medium was used, and a two phase glucose feeding profile was applied, which yielded a final rhamnolipid concentration of 14.9 g/L. Applying the feeding profile, glucose stayed almost constant until 28 h of cultivation and decreased afterwards to limiting levels. Until the end of cultivation 253.0 ± 0.1 g glucose was added to the bioreactor of which a total of 252.0 ± 0.6 g glucose was metabolized. By modeling the fed-batch bioreactor cultivations the time courses of generated biomass, rhamnolipid and consumed glucose were described. The model was furthermore used to derive key process parameters from the collected data. The obtained values for the specific product formation rates (q ) reached 18 mg/(g h) and yield coefficients (Y ) 10 mg/g respectively. RL RL/S Keywords: Heterologous rhamnolipid biosynthesis, Pseudomonas putida, Biosurfactant, Fed batch cultivation carbon source that avoids the difficulties in using vegeta - Introduction ble oil used as a carbon source for P. aeruginosa (Müller Low molecular weight amphiphilic compounds secreted and Hausmann 2011). However, reported heterologous by several microorganisms are often termed biosur- rhamnolipid production with maximal product concen- factants. These microbial surfactants exhibit diverse tration of 7.3 g/L (Cha et al. 2008) is by far not convinc- structures, are generally assumed to be easily biodegrad- ing in comparison to that reported from P. aeruginosa. As able, to display a non-toxic character and often show bio- stated before, heterologous rhamnolipid production in active properties (Hausmann and Syldatk 2014). Despite Pseudomonas putida KT2440 is not naturally regulated these advantages microbial surfactants are so far not by quorum sensing leading to constitutive rhamnolipid used in large scale industrial production. Several special- biosynthesis. However, the actual handling of P. putida ized bioreactor concepts for the biosurfactant production rhamnolipid producing strains remains challenging. have been reviewed by Beuker et  al. (2014). The gly - This highly complex bacterial process exhibits multilay - colipid rhamnolipid is one of the most studied microbial ered interactions between glucose metabolization, bio- surfactants (Abdel-Mawgoud et al. 2011). mass growth and product formation. These interactions The heterologous and glucose-based rhamnolipid pro - impose difficulties to determine influences of changes in duction brings several advantages and has been com- glucose feed on the cultivation. To encounter these dif- paratively extensively studied since the 1990s. Firstly the ficulties model based process optimization with several pathogenicity and the complex biosynthesis regulation iteration steps has been successfully employed in the past of the natural producer Pseudomonas aeruginosa can (Kovárová-Kovar et al. 2000). be avoided. Secondly glucose represents a convenient Wittgens et  al. (2011) examined heterologous rhamnolipid production using P.  putida KT42C1 *Correspondence: Marius.Henkel@uni-hohenheim.de pVLT31_rhlAB in glucose enriched LB medium in a Department of Bioprocess Engineering (150k), Institute of Food Science and Biotechnology, University of Hohenheim, Fruwirthstr. 12, baffled shake flask experiment. In their study Witt - 70599 Stuttgart, Germany gens et  al. (2011) described biomass growth using a Full list of author information is available at the end of the article © The Author(s) 2016. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Beuker et al. AMB Expr (2016) 6:124 Page 2 of 7 logistic equation and glucose metabolization as well as contained 100  mL SupM medium in a 1  L baffled shake rhamnolipid formation using biomass dependent ordi- flask inoculated with 1 mL from the 24 h LB culture and nary differential equations (ODEs). In their approach incubated for 12 h. rhamnolipid productivity as well as glucose metaboli- zation exhibited constant values and were not growth Bioreactor cultivations rate dependent. P. putida KT42C1 pVLT31_rhlAB was All bioreactor cultivations were carried out as dupli- reported to produce up to 1.5 g/L of rhamnolipid. cates. The bioreactor (Minifors, HT Infors, Bottmingen, Switzerland) was equipped with an integrated pH, tem- Methods perature and aeration control system. During bioreac- Methods were adapted from foam fractionation pro- tor cultivation aeration was set to 0.133  vvm and pO cesses as described in Beuker et al. (2016). was controlled at 13% via stirring rate starting with a minimum of 300 rpm. Temperature was held constant at Chemicals 30 °C and pH was controlled to 6.8 via 1 M H SO or 19% 2 4 All chemicals used in the current study were purchased NH OH. Generated foam was detected by an antifoam from Carl Roth GmbH (Karlsruhe, Germany) if not probe and antifoaming agent Tego KS 53 (Evonik Indus- stated otherwise. tries) was added if needed. Bioreactors were inoculated with 12 h SupM seed culture to a final OD of 0.5. Initially Microorganism and plasmid 6 g glucose was provided in the culture vessel. After 5  h P. putida KT2440 with plasmid pSynPro8oT_rhlAB pro- of batch fermentation a dual phase feeding profile was ducing mono-rhamnolipids was used as described in started as displayed in Fig. 2b. Beuker et al. (2016). Analytical methods Culture conditions Sampling and processing Media Taken bioreactor samples were mixed with equal vol- Tetracycline was added to all media to an end concentra- umes of hexane and centrifuged (4700 rpm, 15 min, 4 °C) tion of 20  mg/L. For the first culture LB medium (5  g/L to remove antifoaming agent. The hexane phase was dis - yeast extract (BD, Heidelberg, Germany), 10  g/L tryp- carded and the cell free aqueous supernatant was used tone (BD), 5  g/L NaCl; pH 7.0) was utilized. For seed for rhamnolipid, glucose and ammonium detection. The culture SupM medium (4.4 g/L Na HPO ∙2 H O, 1.5 g/L cell pellet was washed with 1:1 (v/v) 9 g/L NaCl solution, 2 4 2 KH PO , 1  g/L NH Cl, 0.2  g/L MgSO ∙7 H O, 0.02  g/L centrifuged (4700 rpm, 15 min, 4 °C), dried at 100 °C and 2 4 4 4 2 CaCl ∙2 H O, 0.006 g/L FeCl , 30 g/L glucose, 10 g/L yeast used for gravimetrical cell dry weight determination. 2 2 3 extract, 1  mL/L trace element solution 2, pH 6.8; trace Rhamnolipid detection was performed as described element solution 2: 0.3 g/L H BO , 0.2 g/L CoCl ∙6 H O, by Schenk et  al. (1995) with minor adjustments. Part 3 3 2 2 0.1  g/L ZnSO ∙7 H O, 0.03  g/L MnCl ∙4 H O, 0.01  g/L of the liquid phase was acidified with 1:100 (v/v) phos - 4 2 2 2 CuCl ∙2 H O, 0.03 g/L Na MoO ∙2 H O, 0.02 g/L NiCl ∙6 phoric acid and rhamnolipids were extracted twice with 2 2 2 4 2 2 H O) was applied. In the bioreactor cultivation ModR 1.25:1 (v/v) ethyl acetate. For rhamnolipid measure- medium (22  g/L KH PO , 2.6  g/L (NH ) HPO , 1.4  g/L ment appropriate amount of this ethyl acetate extract 2 4 4 2 4 MgSO ∙7 H O, 0.87 g/L citric acid, 0.01 g/L FeSO ∙7 H O, was evaporated. Rhamnolipids were resolved in ace- 4 2 4 2 5 g/L glucose, 10 mL/L trace element solution 2, pH 6.8) tonitrile and derivatized for 90  min at 1400  rpm and was used. 15  g/L antifoaming agent Tego KS 53 (Evonik 60  °C using a 1:1 mixture of 40  mM bromphenacylbro- Industries, Essen, Germany) was added to the bioreac- mid and 20  mM tri-ethyl-ammonium/-amin. Detection tor cultivation. Feed medium contained 400 g/L glucose, of rhamnolipids was performed using a HPLC device 22 g/L KH PO , 3.43 g/L KH PO , 0.01 g/L FeSO ∙7 H O (Agilent 1100 Series, Agilent, Waldbronn, Germany) 2 4 2 4 4 2 −3 and 0.1∙10  g/L CuCl ∙2 H O. equipped with a 15  cm reversed phase column (Supel- 2 2 cosil LC-18, Supelco, Deisenhofen, Germany) at 30 °C. Preparation of seed culture The mobile phase was composed of 100% methanol and All shake flasks were inoculated in a shake incuba - ultrapure water. For rhamnolipid detection a gradient tor chamber (Multitron II, HT Infors, Bottmingen, was applied. During the first 17  min methanol concen - Switzerland) at 30  °C and 120  rpm. First 25  mL LB in a tration was increased starting at 80 to 100%. This meth - 100  mL baffled shake flask were inoculated with 50  µL anol concentration was held for 8 min and decreased to from a glycerol stock solution of P. putida KT2440 80% during the next 5 min. Rhamnolipids were detected pSynPro8oT_rhlAB and incubated for 24 h. Seed cultures at a wave length of 254  nm at 30  °C. For calibration Beuker et al. AMB Expr (2016) 6:124 Page 3 of 7 standard solutions of rhamnolipid in acetonitrile were Biomass used. Biomass mass growth was described via an autocatalytic The concentration of glucose and ammonium were process and a jump function to model lag times (Eq.  1). detected from the aqueous phase of samples using glu- The jump function depended on the lag time (τ) and the cose (R-Biopharm AG, Darmstadt, Germany) and autocatalytic process on existing biomass (X) and specific ammonium (Merck KGaA, Darmstadt, Germany) growth rate (µ). assay kits, respectively, according to the manufacturers’ dX 1 instructions. = ∗ μ ∗ X (1) 100∗(−t+τ ) dt 1 + e Software for modeling and parameter optimization The specific growth rate was represented by a combi - Modeling was performed in a mathematical/numeri- nation of Monod and Luong kinetic (Eq.  2). Unhindered cal computing environment (Matlab, The MathWorks, growth was described by Monod kinetic depending on Natick, MA, USA). To solve the ordinary differential the concentration of the growth limiting substrate (c ), Glu equations the numeric solver “ode45” embedded in Mat- a Monod constant (K ) and a maximal growth rate (µ ) S max lab environment was used based on an explicit Runge– (Monod 1949). Inhibiting effects of high glucose concen - Kutta formula. Parameter optimization was performed tration were integrated using Luong kinetic including the using the Matlab embedded function “lsqnonlin” apply- concentration of the growth limiting substrate, an inhibi- ing a trust-region-reflective algorithm, which is based on tory constant (K ) and a shape factor (n) (Luong 1987). the interior reflective Newton method. The parameter c c Glu Glu optimization was executed by minimizing the sum of μ = μ ∗ ∗ 1 − max (2) c + K K S i Glu squares of the errors between simulation data and meas- ured data considering biomass, rhamnolipid and glucose. Therefore, biomass growth was modeled using Eq. 3. Equations for modeling dX 1 Glu/V BR The general model setup is depicted in Fig.  1 outlining = ∗ μ ∗ max 100∗(−t+τ ) dt 1 + e Glu/V + K BR s the interactions between glucose consumption, biomass Glu/V BR (3) growth, rhamnolipid production, feed and volume of the ∗ 1 − ∗ X cultivation broth. All parameters used for modeling are K summarized in Table 1. Fig. 1 Dependencies between volume, feed, glucose, biomass and rhamnolipid; dependent parameter are depicted in framed boxes, dashed boxes depict type of dependencies, arrows show the source(s) and sink of dependencies Beuker et al. AMB Expr (2016) 6:124 Page 4 of 7 Table 1 Parameters for modeling Bioreactor volume For changes in the overall bioreactor volume just changes Parameter Meaning Unit due to the feed medium were considered (Eq.  6). Vol- α* Growth dependent production term – ume addition by acid, base and antifoaming agent −1 β* Growth independent production term h were assumed to level out by volume reduction due to −1 µ Growth rate h sampling. −1 µ * Maximal growth rate h max dV m ˙ −1 BR Feed ρ Density g L (6) dt ρ τ* Lag time h BR Bioreactor – To describe the time course of the bioreactor cultiva- −1 c Concentration of index g L tion the 10 parameter shown in Table  1 were optimized Index Glu Glucose g using the developed software. Optimization was imple- −1 K * Inhibiting concentration of glucose g L mented using measuring data of the double determina- −1 K * Monod constant g L tion cultivation applying a dual phase feeding profile and −1 m ˙ Mass flux g h the set of ODEs (Eq. 3, 4, 5, 6). −1 −1 m * Maintenance coefficient g g L S/X n* Shape factor for Luong kinetic – Results RL Rhamnolipid g Time courses of overall biomass, rhamnolipid and glucose t Time h during bioreactor cultivation V Volume of index L Time courses of generated biomass, rhamnolipid and Index X Biomass g consumed glucose are depicted in Fig.  2. In bioreactor −1 Y * Rhamnolipid yield coefficient g g cultivations applying the dual phase feeding profile glu - P/S −1 Y * Biomass yield coefficient g g cose amount decreased in the beginning, stayed almost X/S constant at 0.2 g from 12 h till 28 h and decreased after- * Parameter determined via optimization wards to limiting levels. Biomass grew constantly with a leveling trend in the end reaching 35.3 ± 3.2 g in the end of cultivation. Rhamnolipid generation started after 24 h Glucose and increased steadily to a maximal value of 22.2 ± 4.3 g The time course of glucose was modeled taking various after 72 h, which equals a concentration of 14.7 ± 2.6 g/L. glucose sinks as well as sources into account (Eq. 4). Glu- Until the end of cultivation 253.0  ±  0.1  g glucose was cose was assumed to be utilized by biomass and rham- added to the bioreactor and in total 252.0 ± 0.6 g glucose nolipid generation as well as by maintenance of the was metabolized. biomass. Conversion efficiencies of glucose to biomass or rhamnolipid were determined by biomass and rham- Parameter optimization nolipid yield coefficients (Y and Y , respectively). X/S P/S To describe the time course of bioreactor cultivations Glucose amount used for maintenance was described by a parameter optimization was conducted using the set a maintenance coefficient (m ). Applied feed (m ˙ ) Feed S/X of ODEs (Eq.  3, 4, 5, 6) and the experimental data. The with its glucose feed concentration (c ) was deter- Glu in Feed results are depicted in Fig. 2 and Table 2. mined as glucose input. Biomass and glucose time course were satisfactorily dGlu dX 1 dRL described using the model with the optimized parameter =˙ m ∗ c − ∗ + Feed Glu in Feed with end values of 33.5 and 0.0 g, respectively. However, dt dt Y dt X/S rhamnolipid production was just described properly in (4) ∗ + m ∗ X S/X the end of cultivation with a final rhamnolipid mass of P/S 23.2  g which equals a concentration of 14.9  ±  0.5  g/L. The overall rhamnolipid production rate reached 0.3 g/h Rhamnolipid and the final production rate 0.4  g/h. The volume of the Rhamnolipid kinetic was described by a growth depend- bioreactor was underestimated after 14 h, because mod- ent (α) and growth independent (β) part of the specific eled volume just considered changes caused by feeding biomass productivity (Eq. 5). medium. dRL In Fig. 3 the model-derived time courses of the growth = (α ∗ μ + β) ∗ X (5) dt rate (µ), the specific rhamnolipid production rate (q ) RL Beuker et al. AMB Expr (2016) 6:124 Page 5 of 7 Fig. 2 Time courses of P. putida KT2440 pSynPro8oT_rhlAB bioreactor cultivation measurement data compared to modeled data. a experimentally determined (black circles) and modeled total biomass time courses (black line), b experimentally determined (grey triangles), modeled glucose time courses (black line) and total fed glucose amount (grey dashed line), c measured rhamnolipid (empty squares) and modeled rhamnolipid time course (black line), d calculated bioreactor volume at each sampling point (black circles) and modeled bioreactor volume time course (black line) applying parameters depicted in Table 2 Table 2 Optimized model parameters 1/h were computed. Because rhamnolipid amounts in the fermentation broth accumulate only in the fed-batch Parameter Optimized value Unit phase after 14  h the values of the specific rhamnolipid α 0.0477 – production rate and the yield coefficient are only mean - −1 β 0.0125 h ingful after that time. The highest value for the specific −1 µ 0.4386 h max productivity (18  mg/(g  h)) was registered immediately τ 0.4686 h after switching to the feeding of glucose. Thereafter it −1 K 107.3604 g L i steadily reduced its value until reaching 10  mg/(g  h) at −1 K 0.2034 g L S the end of the process. The product yield increased from −1 −1 m 0.0599 g g L S/X 0.075 to 0.104  g/g during the fed-batch phase. This is n 0.6167 – an expected finding as the growth rates in the fed-batch −1 Y 0.2637 g g phase decreased steadily from 0.052 to 0.002 1/h. It may P/S −1 Y 0.3253 g g be assumed that a reduced growth rate leads to a reduced X/S glucose demand derived for cell growth. This may led to a better yield of product formation. and the yield coefficient (Y ) are shown. The model- P/S Discussion derived values for the growth rate show three distinct In this study a bioreactor fed-batch cultivation for het- phases, reflecting the batch phase before 14  h and the erologous rhamnolipid production in a defined medium fed-batch phase thereafter. The highest value for the and a process model was established. As a result up to growth rates in the batch phase were between 0.38 and 14.9  g/L rhamnolipid were obtained. Utilizing a dual 0.19 1/h while in the fed-batch phase values below 0.052 Beuker et al. AMB Expr (2016) 6:124 Page 6 of 7 metabolic capacity of the construct P. putida KT2440 pSynPro8oT_rhlAB. Differences in rhamnolipid amounts in experimental and modeled data were mainly caused by a lag time in rhamnolipid production in the beginning of cultivation. This lag time did not occur in the model. This lag time in rhamnolipid production could be caused by several rea- sons. One explanation could be the availability of one or both rhamnolipid precursors, which may be impaired or shifted at high growth rates thus resulting in a reduced specific production rate. It is noticeable that rhamnolipid production started when glucose started to be limit- ing corresponding to the onset of the glucose feeding. The recorded value of 18  mg/(g  h) is comparable to the maximum value for the P. putida KT42C1 pVLT31_rhlAB reported by Wittgens et al. (2011) and about 38% for the very similar construct P. putida KT2440 pSynPro8_rhlAB grown on a complex rich medium containing additional glucose (Tiso et  al. 2016). However, the comparison to experimental data reported on complex rich medium must be considered with caution. The established model could be used to describe time courses of biomass and glucose during the overall bio- reactor cultivation and of rhamnolipid in the end of bio- reactor cultivation. Additionally the time courses of the growth rate, the specific productivity and the yield coeffi - cient could be derived. To determine which factors influ - ences biomass as well as rhamnolipid production further experiments should be conducted concerning medium components, e.g. the influence of iron and copper. Addi - tionally, further investigations should be made to deter- mine if other essential components should be added to the cultivation due to depletion by high biomass growth (e.g. sulphur, magnesium, calcium, sodium or trace elements). For future work, strain engineering should focus on enhancing the effective yield that reached a value of 10% in relation to the theoretical maximum value of 49% (Henkel et al. 2012). Therefore, the putative genera - tion of polysaccharides as byproducts should be mainly Fig. 3 Time courses of modeled-derived process parameters for P. addressed. Another approach will be the combination putida KT2440 pSynPro8oT_rhlAB bioreactor cultivation a time course of the strategy presented here using fed-batch process- of growth rate, b time course of specific rhamnolipid production rate, ing, with in  situ product removal by foam fractionation c time course of product yield (Beuker et  al. 2016) and thus avoiding the necessity of chemical antifoaming agents. This could lead to a further enhanced semi-continuous process. phase feeding profile in bioreactor cultivations resulted Authors’ contributions in an overall rhamnolipid productivity of 0.3  g/h and JB planned and executed the experiments, collected and calculated all rhamnolipid titer of 14.9  g/L. So far, in literature the data, created the graphs and figures and drafted this manuscript. AW and FR generated the plasmid pSynPro8_rhlAB. AS generated the plasmid highest rhamnolipid titer reported using heterologeous pSynPro8oT_rhlAB. TB conducted her Bachelor thesis under the supervision production hosts was 7.3  g/L (Cha et  al. 2008). There - of JB and conducted experiments applying the two staged feeding profile. fore, the effectiveness of this bioreactor cultivation apply - MH significantly contributed to data evaluation and interpretation of the experiment as well as graphs and figures. RH substantially contributed to ing the feeding profile was demonstrated as well as the Beuker et al. AMB Expr (2016) 6:124 Page 7 of 7 conception and design of the conducted experiments and contributed in Cha M, Lee N, Kim MM, Kim MM, Lee S (2008) Heterologous production of manuscript writing. All authors read and approved the final manuscript. Pseudomonas aeruginosa EMS1 biosurfactant in Pseudomonas putida. Bioresour Technol 99:2192–2199. doi:10.1016/j.biortech.2007.05.035 Author details Hausmann R, Syldatk C (2014) Types and classification of microbial surfactants. Department of Bioprocess Engineering (150k), Institute of Food Science In: Biosurfactants. CRC Press, Boca Raton, pp 3–18 and Biotechnology, University of Hohenheim, Fruwirthstr. 12, 70599 Stuttgart, Henkel M, Müller MM, Kügler JH, Lovaglio RB, Contiero J, Syldatk C, Hausmann Germany. Ulm Center for Peptide Pharmaceuticals (U-PEP), Ulm University, R (2012) Rhamnolipids as biosurfactants from renewable resources: Albert-Einstein-Allee 11, 89081 Ulm, Germany. Present Address: Evonik Indus- concepts for next-generation rhamnolipid production. Process Biochem tries, Evonik Technology and Infrastructure GmbH, Rodenbacher Chaussee 4, 47:1207–1219. doi:10.1016/j.procbio.2012.04.018 63457 Hanau-Wolfgang, Germany. Kovárová-Kovar K, Gehlen S, Kunze A, Keller T, von Däniken R, Kolb M, van Loon APGM (2000) Application of model-predictive control based on artificial Competing interests neural networks to optimize the fed-batch process for riboflavin produc- The authors declare that they have no competing interests. This article does tion. J Biotechnol 79:39–52. doi:10.1016/S0168-1656(00)00211-X not contain any studies with human participants or animals performed by any Luong JH (1987) Generalization of monod kinetics for analysis of growth data of the authors. with substrate inhibition. Biotechnol Bioeng 29:242–248. doi:10.1002/ bit.260290215 Received: 8 November 2016 Accepted: 1 December 2016 Monod J (1949) The growth of bacterial cultures. Annu Rev Microbiol 3:371–394. doi:10.1146/annurev.mi.03.100149.002103 Müller MM, Hausmann R (2011) Regulatory and metabolic network of rhamnolipid biosynthesis: traditional and advanced engineering towards biotechnological production. Appl Microbiol Biotechnol 91(2):251–264 Schenk T, Schuphan I, Schmidt B (1995) High-performance liquid chromato- References graphic determination of the rhamnolipids produced by Pseudomonas Abdel-Mawgoud AM, Hausmann R, Lépine F, Müller MM, Dèziel E (2011) aeruginosa. J Chromatogr A 693:7–13. doi:10.1016/0021-9673(94)01127-Z Rhamnolipids: detection, analysis, biosynthesis, genetic regulation, and Tiso T, Sabelhaus P, Behrens B, Wittgens A, Rosenau F, Hayen H, Blank LM (2016) bioengineering of production. In: Soberón-Chavez G (ed) Biosurfactants: Creating metabolic demand as an engineering strategy in Pseudomonas from genes to applications. Springer, Berlin putida—Rhamnolipid synthesis as an example. Metab Eng Commun Beuker J, Syldatk C, Hausmann R (2014) Bioreactors for the Production of 3:234–244. doi:10.1016/j.meteno.2016.08.002 biosurfactants. In: Biosurfactants: production and utilization—processes, Wittgens A, Tiso T, Arndt TT, Wenk P, Hemmerich J, Müller C, Wichmann R, Küp- technologies, and economics. CRC Press, Boca Raton, p 117 per B, Zwick M, Wilhelm S, Hausmann R, Syldatk C, Rosenau F, Blank LM Beuker J, Steier A, Wittgens A, Rosenau F, Henkel M, Hausmann R (2016) (2011) Growth independent rhamnolipid production from glucose using Integrated foam fractionation for heterologous rhamnolipid production the non-pathogenic Pseudomonas putida KT2440. Microb Cell Fact 10:80. with recombinant Pseudomonas putida in a bioreactor. AMB Express 6:11. doi:10.1186/1475-2859-10-80 doi:10.1186/s13568-016-0183-2

Journal

AMB Express – Springer Journals

Published: Dec 12, 2016

There are no references for this article.