Integrated strain- and process design enable production of 220 g L−1 itaconic acid with Ustilago maydis
Itaconic acid is an unsaturated, dicarboxylic acid which finds a wide range of applications in the polymer industry and as a building block for fuels, solvents and pharmaceuticals. Currently, Aspergillus terreus is used for industrial production, with titers above 100 g L−1 depending on the conditions. Besides A. terreus, Ustilago maydis is also a promising itaconic acid production host due to its yeast-like morphology. Recent strain engineering efforts significantly increased the yield, titer and rate of production.
In this study, itaconate production by U. maydis was further increased by integrated strain- and process engineering. Next-generation itaconate hyper-producing strains were generated using CRISPR/Cas9 and FLP/FRT genome editing tools for gene deletion, promoter replacement, and overexpression of genes. The handling and morphology of this engineered strain were improved by deletion of fuz7, which is part of a regulatory cascade that governs morphology and pathogenicity. These strain modifications enabled the development of an efficient fermentation process with in situ product crystallization with CaCO3. This integrated approach resulted in a maximum itaconate titer of 220 g L−1, with a total acid titer of 248 g L−1, which is a significant improvement compared to best published itaconate titers reached with U. maydis and with A. terreus.
In this study, itaconic acid production could be enhanced significantly by morphological- and metabolic engineering in combination with process development, yielding the highest titer reported with any microorganism.
KeywordsUstilago maydis Itaconic acid Metabolic engineering Morphological engineering Biochemical engineering In situ precipitation
clustered regularly interspaced short palindromic repeats
high performance liquid chromatography
cell dry weight
polymerase chain reaction
- A. terreus
- E. coli
- A. niger
- C. glutamicum
- U. maydis
- U. cynodontis
- U. vetiveriae
regulator of itaconic acid biosynthesis
mitochondrial transporter from Ustilago maydis
mitochondrial transporter from Aspergillus terreus
dual specificity protein kinase
oxidase encoding gene
native promoter of ria1
strong and constitutive promoter
2-(N-morpholino) ethanesulfonic acid
More than 300 potential bio-based building blocks were selected from the U.S. Department of Energy according to criteria such as estimated processing costs, estimated selling price, and the technical complexity, to determine the most important chemicals that can be produced from biomass. In the top selection, nine belong to the group of organic acids , underlining the importance of this class of chemicals. One of these compounds is the unsaturated dicarboxylate itaconic acid. It was first described in 1837  and primary reports about microbial production with Aspergillus itaconicus date back to 1931 . Due to its two functional groups, radical polymerization of the methylene group and/or esterification of the carboxylic acid with different co-monomers is possible [59, 63, 67]. This leads to a wide range of applications in the paper, architectural, pharmaceutical, paint, lacquer, and medical industries [5, 6, 40, 43, 50, 55, 61, 71]. It can also be used as an intermediate for the production of 3-methyltetrahydrofuran, a potential biofuel with advantageous combustion properties . Further, itaconate production by mammalian macrophages is reported, where it plays a key role in the human immune response [11, 57, 69], with possible applications as therapeutic agent for autoimmune diseases .
In spite of this wide variety of potential applications, the market size of itaconic acid in 2011 was relatively small, with 41,400 tons and a market value of $74.5 million . This is caused by the relatively high price of approximately two dollars per kg and the availability of cheaper petro-based alternatives such as acrylic acid. Reduction of this price is, therefore, a major criterion for access to further markets. To be competitive against petro-based products, costs need to reduce to around $0.5 per kg . Assuming that the price would decrease, itaconic acid has the possibility to replace acrylic acid in the production of poly(methyl methacrylate), the production of which is petroleum based with a market worth of $11 billion [39, 43, 59]. Since 1950, Aspergillus terreus is used for the industrial production of itaconate . Charles Pfizer Co. was granted the first patent for the production of itaconate with the filamentous fungus A. terreus by submerged cultivation . During the last decades, the responsible metabolic pathways and regulatory mechanisms of itaconate production in A. terreus were studied in detail . Major advances were achieved through process development. This long history of optimization has enabled titers above 100 g L−1 and yields near the theoretical maximum at low pH, making A. terreus the current best production host for itaconate production [7, 29, 34, 49, 50, 53, 66]. However, despite the long history and experience, itaconate production in A. terreus remains challenging. A specific pellet growth form is required for high productivity [25, 39] and therefore, morphology has to be strictly controlled. A. terreus reacts very sensitively to certain medium impurities, which can induce mycelium formation and stop itaconate production [15, 48, 50]. Thus, medium must be pretreated to remove impurities from production medium, especially when using less pure industrial substrates such as molasses [38, 59]. Consequently, morphological control influences the manufacturing process tremendously, leading to increased operational costs and failed batches.
Besides A. terreus, numerous itaconate producers have been engineered in recent years, such as E. coli , A. niger , and C. glutamicum . Besides these heterologous hosts, Ustilaginaceae like the pH-tolerant Ustilago cynodontis or the yeast-like Ustilago maydis are natural itaconate producers which have recently been engineered to higher efficiency [21, 24, 33, 75]. Among the Ustilaginaceae, U. maydis is the most studied species in the fields of plant pathogenicity, cell biology, and biotechnology [17, 18, 52, 68, 70]. The Ustilaginaceae produce a broad spectrum of interesting products such as organic acids [21, 24, 76], glycolipids [14, 58], polyols [21, 35], and enzymes . This, along with their yeast-like growth, makes them attractive for biotechnological applications .
That said, certain stresses can induce filamentous growth in U. maydis [45, 54] but efficient itaconate production with this species is, at least at small scale, not coupled to a specific morphology. In wild-type U. maydis, itaconate production is induced by nitrogen limitation  and requires pH values above five . Like in A. terreus, the genes encoding the itaconate production pathway in U. maydis are clustered and co-regulated [19, 53]. Considerable progress has been made in increasing the yield, titer, and rate of itaconate production in U. maydis and related species by metabolic engineering and process development. Geiser et al.  characterized the itaconate production pathway and identified an itaconate oxidase Cyp3, which produces the downstream product (S)-2-hydroxyparaconate. The disruption of this oxidase, and overexpression of the cluster-associated regulator Ria1, led to 4.5-fold increase in ITA production in U. maydis . In U. vetiveriae, itaconate production from glycerol could be increased 2.5-fold by overexpression of ria1 or 1.5-fold by overexpression of the mitochondrial transporter mtt1 .
In another study, we could show that heterologous expression of the mitochondrial transporter MttA from A. terreus in U. maydis enables more efficient itaconate production than the native mitochondrial transporter . Further, by deletion of fuz7 in U. cynodontis, a stable yeast-like growth could be established for several relevant itaconic acid production conditions . This is especially favorable for large-scale fermentation . Furthermore, with optimization of growth media and the fermentation process, such as pulsed fed-batch strategies, product titers can be significantly increased [20, 21]. This is especially effective when combined with in situ product removal approaches such as reactive extraction or calcium precipitation [23, 43, 46, 75].
These optimizations have individually made a significant impact on the efficiency of itaconate production in Ustilago. In this study, we consolidate several of these metabolic and bioprocess engineering strategies to achieve itaconate titers that surpass those currently achieved by any other host.
Results and discussion
Engineering of a marker-free U. maydis MB215 for enhanced itaconate production
In the cultures of these overproducing strains, we also observed a degree of filamentous growth. Although this is by far not as prominent as described for U. cynodontis , elongated cells and filaments were formed in all tested U. maydis strains for all conditions shown in Fig. 1, especially upon addition of CaCO3.
Morphological engineering in U. maydis ∆cyp3 ∆Pria1::Petef
Usually, filamentous growth in U. maydis is investigated in terms of pathogenicity. In its natural habitat, filamentous growth is indispensable to U. maydis for infection of Zea mays. This is strongly coupled with sexual development including a complex regulatory system [36, 51, 52]. Filamentous growth can also occur in haploid cells when they encounter stresses such as low pH, nitrogen limitation, or the presence of sunflower oil [45, 56]. This ability to grow filamentously is an obstacle in a biotechnological context, as it strongly influences bioprocess parameters such as oxygen transfer, viscosity, and clogging, and it increases the sensitivity to hydro-mechanical stress . To solve this problem and to restore robust yeast-like growth, the fuz7 gene was deleted in the marker-free ∆cyp3 ∆Pria1::Petef #2 strain by replacement with a hygromycin marker through homologous recombination, followed by FLP/FRT-mediated marker excision . Fuz7 is part of the Ras/mitogen-activated protein kinase (MAPK) pathway, which plays an important role in conjugation tube formation and filamentous growth . By deletion of fuz7 in the strongly filamentous U. cynodontis, filamentous growth was repressed without influencing itaconate production and cell fitness under biotechnologically relevant conditions . Deletion of fuz7 in U. maydis is known to abolish filamentous growth, and it also renders the strain completely apathogenic [3, 45]. This inability to colonize the maize plant is an additional advantage in a biotechnological context, as it may alleviate possible regulatory hurdles for industrial application.
Mitochondrial transporter engineering in U. maydis ∆cyp3 ∆Pria1::Petef ∆fuz7
An even more pronounced effect was observed with similar cultures using glycerol as C-source (Additional file 1: Fig. S1). Glycerol is a very poor substrate for wild-type U. maydis MB215 , and it invokes a high degree of filamentation and pigmentation in U. cynodontis . The fuz7 deletion had a very positive effect on the glycerol uptake rate and itaconate production, with the ∆cyp3 ∆Pria1::Petef ∆fuz7 strain producing 13.1 ± 0.04 g L−1, compared to 4.3 ± 0.4 g L−1 produced by the ∆cyp3 ∆Pria1::Petef control strain. Titers could be further increased with U. maydis ∆cyp3 ∆Pria1::Petef ∆fuz7 PetefmttA to 16.1 ± 0.4 g L−1 itaconate.
Optimized itaconate production in a stirred bioreactor
In a similar approach where pH was controlled by titration with NaOH, a much lower level of itaconate production was observed (Additional file 1: Fig. S2), reaching a maximum titer of only 35.9 ± 1.5 g L−1 with a yield of 0.2 ± 0.01 gITA g GLC −1 and an overall productivity of 0.12 ± 0.004 g L−1 h−1. In this titrated fermenter less than 1 g L−1 malate was produced, supporting the hypothesis that the additional CO2 from CaCO3 increases malate production. The overall decrease of productivity in the titrated culture is likely caused by the overexpression of mttA, which significantly stresses the cells leading to reduced growth and productivity as described previously . The application of in situ itaconate crystallization with CaCO3, greatly reduced product inhibition, which is especially relevant with this deeply engineered strain, leading to almost threefold higher production rates.
In this study, the combination of metabolic and morphological engineering together with in situ crystallization of itaconate yielded a titer of 220 g L−1 itaconate, which corresponds to 284 g L−1 calcium itaconate. This titer exceeds the 160 g L−1 achieved with A. terreus , although the yield and production rate achieved with A. terreus are still higher . Especially, the yield achieved with U. maydis could be further increased by the reduction of byproduct formation, as illustrated by the relatively high levels of malate production under these conditions. The strategy of in situ crystallization has not been reported in a biotechnological context with A. terreus, likely because the used pH values and the presence of solids strongly affect its morphology . The use of in situ crystallization greatly enhanced itaconate production, but it will also pose new bioprocessing challenges such as solid/solid separation of biomass, CaCO3 and Ca-itaconate, or pH shifts for resolubilization of itaconate prior to purification [47, 62]. In all, this study demonstrates the power of an integrated approach of strain and process engineering by greatly enhancing Ustilago-based itaconate production.
Materials and methods
Media and culture conditions
U. maydis MB215 strains used in this study
Ustilago maydis MB215
Ustilago maydis ∆cyp3 Petefria1
Ustilago maydis ∆cyp3
Ustilago maydis ∆cyp3 ∆Pria1::Petef #1
Ustilago maydis ∆cyp3 ∆Pria1::Petef #2
Ustilago maydis ∆cyp3 ∆fuz7 ∆Pria1::Petef
Ustilago maydis ∆cyp3 ∆Pria1::Petef ∆fuz7 PetefmttA
Controlled batch cultivations were performed in a New Brunswick BioFlo® 115 bioreactor (Eppendorf, Germany) with a total volume of 1.3 L and a working volume of 0.5 L or a total volume of 2.0 L and a starting volume of 1.0 L if CaCO3 was used. All cultivations were performed in batch medium containing 0.2 g L−1 MgSO4·7H2O, 0.01 g L−1 FeSO4·7H2O, 0.5 g L−1 KH2PO4, 1 g L−1 yeast extract (Merck Millipore, Germany) 1 mL L−1 vitamin solution, and 1 mL L−1 trace element solution and varying concentrations of glucose and NH4Cl as indicated. During cultivation, pH 6.0 was maintained by automatic addition of 10 M NaOH or pH was kept above 6.2 by manual addition of CaCO3. The stirring rate was kept constant at 1000 rpm with 2 Rushton impeller. The bioreactor was aerated with an aeration rate of 1 L min−1 (2 vvm) for working volume of 0.5 L or 2 L min−1 (1 vvm) for total volume of 2 L, while evaporation was limited by sparging the air through a water bottle. The temperature was set at 30 °C. The bioreactor was inoculated to a final OD600 of 0.75 with cells from an overnight culture in 50 mL screening medium containing 50 g L−1 glucose and 100 mM MES buffer.
When using CaCO3 as buffer, 1 mL of culture broth was taken for OD600 determination and HPLC analysis. The CaCO3 was dissolved with HCL prior to further measurements, basically as described by Zambanini et al. .
Cell densities were measured by determining the absorption at 600 nm with an Ultrospec 10 Cell Density Meter (Amersham Biosciences, Chalfont St Giles, UK).
For CDW determination of controlled high-density pulsed fed-batch fermentation of U. maydis MB215 ∆cyp3 ∆Pria1::Petef ∆fuz7 PetefmttA with NaOH titration 1 mL culture broth was centrifuged at maximum speed (Heraeus Megafuge 16R, TX-400 rotor, Thermo Scientific) and the pellet was dried (Scan Speed 40 lyophilizer, Labogene ApS) for 24 h at 38 °C and weighed afterwards.
Off-gas analysis for online monitoring of CO2 content were performed with BCpreFerm sensors (BlueSens gas sensor GmbH).
Differential interference contrast (DIC) microscopy was performed with a Leica DM500 light microscope (Leica Microsystems). Images were recorded with a Leica ICC50 digital microscope camera (Leica Microsystems). Images were taken at 630-fold magnification. The cell morphology was analyzed by microscopy at different time points in all cultivations.
The ammonium concentration in the culture supernatant was measured by a colorimetric method according to Willis et al.  using salicylate and nitroprusside.
Products in the supernatants were analyzed in a DIONEX UltiMate 3000 High-Performance Liquid Chromatography System (Thermo Scientific, Germany) with an ISERA Metab AAC column 300 × 7.8 mm column (ISERA, Germany). As solvent, 5 mM H2SO4 with a flow rate of 0.6 mL min−1 and a temperature of 40 °C was used. Samples were filtered with Rotilabo® (CA, 0.20 µm, Ø 15 mm) or Acrodisc® (GHP 0.20 µm, Ø 13 mm) syringe filters and afterwards diluted up to 1:30 with 5 mM H2SO4. Itaconate and malate were determined with a DIONEX UltiMate 3000 Variable Wavelength Detector set to 210 nm, and glucose with a refractive index detector SHODEX RI-101 (Showa Denko Europe GmbH, Germany). Analytes were identified via retention time and UV/RI quotient compared to corresponding standards. All values are the arithmetic mean of at least three biological replicates instead of CaCO3 fermentations (n = 1). Error bars indicate the deviation from the mean for n = 2, if n > 2 error bars indicate the standard error of the mean. Statistical significance was assessed by t-test (two-tailed distribution, heteroscedastic, p ≤ 0.05).
Plasmid cloning and strain engineering
Plasmids were assembled by Gibson assembly  using the NEBbuilder HiFi DNA Assembly kit (New England Biolabs, Ipswich, MA, USA). Primers were ordered as unmodified DNA oligonucleotides from Eurofins Genomics (Ebersberg, Germany). As polymerase, Q5 High-Fidelity Polymerase was used. Detailed information about utilized primers and plasmid are listed in Additional file 1: Table S3 and S4. All assembled plasmids were subcloned into E. coli 10β from New England Biolab and confirmed by PCR, restriction or sequencing. Standard cloning techniques for E. coli were performed according Sambrook et al. . For transformation, preparation of protoplasts and isolation of genomic DNA of U. maydis protocols according to Brachmann et al.  were used. For deletion of fuz7 in U. maydis, homologous recombination with 1000 bp flanking regions (F1, F2) including FRT sites and a hygromycin resistance cassette were used. For integration of pETEF_CbxR_At_mttA , the plasmid was linearized with SspI and integrated into the genome. For exchange of the promoter of ria1, CRISPR/Cas9 system was used according to Schuster et al.  and sgRNA has been selected online with http://www.e-crisp.org/E-CRISP/ . A donor template was used to exchange the native promoter with the strong and constitutive Petef. Successful integration and deletion were verified by PCR and sequencing.
We thank Dr. Mariana Schuster and Prof. Dr. Regine Kahmann (Max Planck Institute for Terrestrial Microbiology, Department of Organismic Interactions, Marburg) for providing the plasmid pCas9_sgRNA_0 and Dr. Kerstin Schipper and Prof. Dr. Michael Feldbrügge (Institute for Microbiology, Heinrich Heine University Düsseldorf) for pstorI_1rh_WT (pUMa1522). We thank Prof. Dr. Jochen Büchs (Aachener Verfahrenstechnik, RWTH-Aachen) and Dr. Lars Regestein (Biotechnikum, Hans-Knöll-Institut Jena) for advice on Bioprocess development.
All authors contributed significantly to the work. NW conceived and supervised the study. HHT designed and performed experiments and analyzed results with the help of NW and LMB. HHT wrote the manuscript with help of NW and LMB. IB and AL engineered the strains and SM and KAS performed fermentation experiments. All authors read and approved the final manuscript.
This work was funded by the German Federal Ministry of Food and Agriculture (BMEL), through the Specialist agency renewable raw materials e. V. (FNR) as part of the ERA-IB project “TTRAFFIC”. (FKZ 22030515). The laboratory of Lars M. Blank was partially funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany´s Excellence Strategy within the Cluster of Excellence 236 “TMFB” and Exzellenzcluster 2186 “The Fuel Science Center”.
Ethics approval and consent to participate
Consent for publication
All authors have seen and approved the manuscript. All authors have contributed significantly to the work. The manuscript has not been published and is not being considered for publication elsewhere. The authors declare that they have no competing interests.
- 2.Bambouskova M, Gorvel L, Lampropoulou V, Sergushichev A, Loginicheva E, Johnson K, Korenfeld D, Mathyer ME, Kim H, Huang LH, Duncan D, Bregman H, Keskin A, Santeford A, Apte RS, Sehgal R, Johnson B, Amarasinghe GK, Soares MP, Satoh T, Akira S, Hai T, de Guzman Strong C, Auclair K, Roddy TP, Biller SA, Jovanovic M, Klechevsky E, Stewart KM, Randolph GJ, Artyomov MN. Electrophilic properties of itaconate and derivatives regulate the IκBζ-ATF3 inflammatory axis. Nature. 2018;556:501–4.PubMedPubMedCentralCrossRefGoogle Scholar
- 4.Baup S. Über eine neue Pyrogen-Citronensäure, und über Benennung der Pyrogen Säure überhaupt. Annales de Chimie et de Physique. 1837;19:29–38.Google Scholar
- 6.Betancourt T, Pardo J, Soo K, Peppas NA. Characterization of pH-responsive hydrogels of poly(itaconic acid-g-ethylene glycol) prepared by UV-initiated free radical polymerization as biomaterials for oral delivery of bioactive agents. J Biomed Mater Res. 2010;93:175–88.Google Scholar
- 18.Geiser E, Przybilla SK, Engel M, Kleineberg W, Buttner L, Sarikaya E, Den Hartog T, Klankermayer J, Leitner W, Bölker M, Blank LM, Wierckx N. Genetic and biochemical insights into the itaconate pathway of Ustilago maydis enable enhanced production. Metab Eng. 2016;38:427–35.PubMedCrossRefPubMedCentralGoogle Scholar
- 24.Guevarra ED, Tabuchi T. Accumulation of itaconic, 2-hydroxyparaconic, itatartaric, and malic-acids by strains of the genus Ustilago. Agric Biol Chem. 1990;54:2353–8.Google Scholar
- 37.Kane JH, Finlay AC, Amann PF. Production of itaconic acid. Assignee: Pfizer Inc. Patent number: US2385283A, 1945.Google Scholar
- 42.Kinoshita K. Über die Produktion von Itaconsäure und Mannit durch einen neuen Schimmelpilz. Acta Phytochim. 1932;5:271–87.Google Scholar
- 47.Kristiansen B, Mattey M, Linden J, editors. Citric acid biotechnology. London: Taylor & Francis Ltd; 1999. ISBN 0-203-48339-1.Google Scholar
- 57.Michelucci A, Cordes T, Ghelfi J, Pailot A, Reiling N, Goldmann O, Binz T, Wegner A, Tallam A, Rausell A, Buttini M, Linster CL, Medina E, Balling R, Hiller K. Immune-responsive gene 1 protein links metabolism to immunity by catalyzing itaconic acid production. Proc Natl Acad Sci. 2013;110:7820–5.PubMedCrossRefPubMedCentralGoogle Scholar
- 62.Regestein L, Klement T, Grande P, Kreyenschulte D, Heyman B, Maßmann T, Eggert A, Sengpiel R, Wang Y, Wierckx N, Blank LM, Spiess A, Leitner W, Bolm C, Wessling M, Jupke A, Rosenbaum M, Büchs J. From beech wood to itaconic acid: case study on biorefinery process integration. Biotechnol Biofuels. 2018;11:279.PubMedPubMedCentralCrossRefGoogle Scholar
- 64.Sambrook J, Russell DW. Molecular cloning a laboratory manual. New York: Cold Spring Harbor Laboratory Press; 2001. ISBN 978-1-93-611342-2.Google Scholar
- 71.Weastra. Wp 8.1, determination of market potential for selected platform chemicals: itaconic acid, succinic acid, 2,5-furandicarboxylic acid. Technical report, Bioconsept. 2013.Google Scholar
- 72.Werpy T, Peterson G. Top value added chemicals from biomass. Volume 1—results of screening for potential candidates from sugars and synthesis gas. Washington: US-DoE; 2004.Google Scholar
- 74.Zambanini T, Hartmann SK, Schmitz LM, Büttner L, Hosseinpour Tehrani H, Geiser E, Beudels M, Venc D, Wandrey G, Büchs J, Schwarzländer M, Blank LM, Wierckx N. Promoters from the itaconate cluster of Ustilago maydis are induced by nitrogen depletion. Fungal Biol Biotechnol. 2017;4:11.PubMedPubMedCentralCrossRefGoogle Scholar
Open AccessThis 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. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.