Lecithin is the key material attribute in soy bean oil affecting filamentous bioprocesses

Lecithin is the key material attribute in soy bean oil affecting filamentous bioprocesses Complex raw materials are widely used as supplements in biopharmaceutical production processes due to their posi- tive effect on biomass growth and productivity at low cost. However, their use negatively impacts process reproduc- ibility due to high lot-to-lot variability which contradicts current regulatory guidelines. In this study we investigated crude soy bean oil (SBO) which is a common complex raw material for filamentous fungi. We demonstrated that lecithin, which we define as phosphatidylcholines, is in fact the key material attribute in crude SBO positively affect - ing fungal growth and consequently productivity. The methodological toolbox we present here allows the straight- forward isolation of lecithin from crude SBO, its semi-quantification by HPLC and the consequent supplementation thereof in defined amounts. Thus, over-dosage and potential resulting negative impacts on fungal growth and productivity can be omitted. Keywords: Complex raw material, Key material attribute, Soy bean oil, Lecithin, Filamentous fungi An example for complex biological raw materials are Introduction vegetable oils, such as soy bean oil (SBO). It was shown Complex raw materials are commonly used as cheap that SBO leads to an increase in fungal biomass growth media supplements in bacterial, fungal and mammalian and a 50% increase in antibiotics production (Gold- bioprocesses (Gao and Yuan 2011; Millis et al. 1963; Reese schmidt and Koffler 1950; Pan et al. 1959). Although the and Maguire 1969). These raw materials are typically of reasons for this positive effect are still obscure, possible biological origin and thus underlie a high lot-to-lot varia- explanations are: (i) SBO serves as nutritional source, (ii) bility. Furthermore, the specific substance responsible for SBO changes the medium for improved oxygen transfer, the positive impact on the bioprocess is often unknown. and (iii) the antifoam capacity of the oil reduces micro- However, during the International Conference of Har- bial cell damage (Goldschmidt and Koffler 1950; Jones monization, guidelines have been established demanding and Porter 1998). Two critical factors for SBO supple- for the evaluation of critical material attributes (CMAs) mentation are known, namely (i) the time point of addi- and their impact on product quality (ICH 2009). Mate- tion and (ii) the dosage (Anderson et al. 1956; Ohta et al. rial attributes do not only affect product quality but also 1995). It has been shown that low-level supplementation process performance and productivity. We define these of SBO is beneficial, whereas medium to high-level sup - material attributes that have an impact on productivity plementation is disadvantageous. To shed more light on as key material attributes (kMAs). For both, identification that, the effect of crude and refined SBO was tested, how - of CMAs and kMAs, sound science-based knowledge of ever no significant differences in the production of tet - raw material quality has become a necessity. racycline were found (Jones and Porter 1998). Although crude SBO is about 1/5 cheaper than refined oil and con - tains phospholipids, some minerals and tocopherols, it may also contain pesticides, which is why refined SBO is *Correspondence: oliver.spadiut@tuwien.ac.at Research Area Biochemical Engineering, Institute of Chemical, mostly used as medium supplement today. However, sev- Environmental and Bioscience Engineering, Vienna University eral studies have shown the positive impact of phospho- of Technology, Gumpendorferstrasse 1a - 166/4, 1060 Vienna, Austria lipids in crude SBO on fungal growth and productivity Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/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. Hofer et al. AMB Expr (2018) 8:90 Page 2 of 8 (Bateman and Jenkins 1997; Goldschmidt and Koffler assessing correlations between SBO composition and 1950; Tarafdar and Claassen 1988). We hypothesize that byproduct as well as product formation. lecithin, which we define as phosphatidylcholines, is the kMA in SBO causing this positive impact. To investigate Analytical methods this hypothesis, we isolated lecithin from SBO, added HPLC method for crude SBO and lecithin defined amounts thereof to the media for a Penicillium Chromatographic separation was accomplished by adopt- chrysogenum strain and analyzed the effects on fungal ing the method from Jangle et  al. (2013) using a Zorbax growth. For the extraction of lecithin from oily matri- Eclipse Plus C-18 column (3.0 × 150 mm, 3.5 µm; Agilent ces several solid phase extraction (SPE) methods have Technologies) and pre-column (4.6 × 12.5  mm, 5  µm). been reported (Bateman and Jenkins 1997; Bligh and Further details of the chromatographic method are given Dyer 1959; Carelli et  al. 1997; Nash and Frankel 1986; in the Results section. An Ultimate 3000 HPLC system Ruiz-Gutierrez and Perez-Camino 2000). However, SPE (Thermo Fisher Scientific, Waltham, MA, USA) equipped is quite tedious and time consuming and delivers the with a pump (LPG-3400SD), an autosampler (WPS-3000 wanted lecithin in an organic solvent, which is com- SplitLoop), a column oven (Col.Comp. TCC-3000SD), a monly not suitable for media supplementation. Hence, diode array detector (DAD 3000) and a charged aerosol we adopted a simple and cheap extraction procedure, detector (Corona Veo RS) was used. Chromeleon 7.2 was which is currently applied in industrial refinery processes used for control. (Dijkstra and Van Opstal 1989), delivering lecithin in an aqueous background. Additionally, we present a semi- HPLC method for free fatty acids (FFA) quantitative HPLC-CAD method for crude SBO and For sample preparation, FFA were extracted with butyl lecithin analysis, which allows dosage of the extracted acetate at room temperature. The separation of FFA, lecithin without the need of mass spectrometry analysis. namely 18:0, 18:1, 18:2 and 18:3, was achieved by a Zor- The effect of crude SBO and extracted lecithin on fungal bax Eclipse Plus C-18 column (3.0 × 150  mm, 3.5  µm; growth was evaluated in shake flask experiments with P. Agilent Technologies) and pre-column (4.6 × 12.5  mm, chrysogenum. Summarizing, in this study we demonstrate 5 µm). A gradient was implemented over 4 min, starting that lecithin is in fact the kMA in SBO affecting fungal from 90% MQ to 100% acetonitrile, which was held for growth and consequently productivity. another 6  min. FFA were detected by a charged aerosol detector. The method parameters were: 1.0 ml/min flow - Materials and methods rate, 50  °C column oven temperature and 2 µl injection Chemicals volume. All chemicals were of analytical grade and purchased from Carl Roth (Karlsruhe, Germany) or Sigma Aldrich GC method for fatty acid methyl esters (FAME) (St. Louis, MO, USA). Ultra-pure water derived from a An Agilent Gas Chromatography system (7890A, Agi- Milli-Q system from Merck Millipore (Billerica, MA, lent Technologies, Santa Clara, California, USA) with USA) was used for analytical methods. a flame ionisation detector (FID) was used. Chromato - graphic separation was achieved by an HP-88 column SBO samples (60 m × 0.25 mm ID, 0.2 µm; Agilent) and a temperature Unrefined soy bean oil (SBO) originated from two Euro - gradient starting at 40 °C up to 220 °C within 80 min. 1 µl pean manufacturers (called M1 and M2). was injected with an inlet split of 10:1. Prior to injection, the samples were derivatised with trimethylsulfonium SBO lot‑to‑lot variability hydroxide (TMSH) for esterification. Therefore 200  µl For demonstration of SBO lot-to-lot variability, 20 sam- sample, 200  µl internal standard solution and 200  µl ples from manufacturer M1 and M2 were analyzed for TMSH were incubated at 50  °C for 15  min. Quantifica - free fatty acid (FFA), fatty acid methyl esters (FAME) tion was achieved via external and internal standards, e.g. and phosphorous content as well as for a spectroscopic 19:0. Standards were prepared of 14:0, 16:0, 16:1, 18:0, fingerprint by Fourier transform infrared spectroscopy 18:1, 18:2 and 18:3 fatty acids. (FTIR). Data were analyzed by principal component anal- ysis (PCA). ICP‑OES for phosphor analysis Furthermore, our industrial cooperation partner deliv- The samples were diluted in PremiSolv (SPrep, Ger- ered 29 SBO samples as well as on-line and off-line data many) and indium was added as internal standard. Sam- of an industrial scale fungal production process. In order ples were analyzed on an iCAP 6000 inductively coupled to assess the impact of SBO variability on process per- plasma—optical emission spectrometer (ICP-OES, formance, multivariate data analysis was performed, Thermo Scientific). The instrument was equipped with Hofer et al. AMB Expr (2018) 8:90 Page 3 of 8 a radial optic, an Echelle spectrometer and a CCD-chip the specific composition of SBO as well as the dataset detector. The samples were brought into the plasma using including the byproduct concentration were normalized a Babington-type nebulizer and a peristaltic pump. and mean-centered. From the spectral data, i.e. the fin - gerprints, the second derivative was generated in order to FTIR analysis for SBO fingerprint assure baseline correction and the data were mean-cen- Crude SBO samples were directly analyzed using an tered. For PCA, the final matrix consisted of 12 observa - −1 FTIR instrument equipped with an ATR Platinum, a tions and the FTIR region between 800 and 2000  cm . Tensor 37 photometer and a DTGS detector (Bruker For PLS analysis the x-matrix consisted of 39 observa- optics). Spectra were acquired with a resolution of tions and 12 specific components, namely FFA, FAME −1 4 cm and 32 scans per sample against a background of and phosphor, and the y-vector of 39 observations and air. The acquired wavenumber region ranged from 600 to the maximal byproduct concentration. −1 4000 cm . The design of experiment (DoE) investigating the impact of crude SBO and extracted lecithin as well as Shake flask cultivations interactions on process performance was designed with A P-14 P. chrysogenum candidate strain, descending from MODDE Umetrics (Umeå, Sweden), choosing a full fac- the P-2 P. chrysogenum candidate strain (American Type torial screening with 2 factors, namely crude SBO and Culture Collection with the access number ATCC 48271), extracted lecithin, and 3 levels, namely 0, 4 or 8 g/l, and was used for all experiments. The cultivation medium resulted in 11 experiments including 3 center points. was similar as described previously (Posch et  al. 2013), either with or without supplementation of crude SBO Results or extracted lecithin. In every case glucose was the main SBO as typical complex raw material C-source. For preculture, 500  ml shake flasks were filled Typical characteristics of complex raw materials are their with 30 ml medium and inoculated with 2.8 × 10 spores positive impact on growth and productivity at low cost, of P. chrysogenum. The flasks were incubated at 25 °C and but also their negative impact on process reproducibility 300 rpm for up to 72 h. A flask filled with destilled water due to high lot-to-lot variability. We assessed both char- was added in the incubator in order to increase humidity acteristics for crude SBO in a P. chrysogenum antibiotics and reduce evaporation. The flasks were either harvested production process in shake flask experiments (Fig.  1). In at the end-point only after 72  h or more sample points the preculture, supplementation with up to 4  g/l crude were evaluated in order to get time-resolved data. For the SBO led to an increase in biomass formation. A higher latter, samples were analyzed every 12 h. Due to the low concentration of SBO had a negative effect. This dosage- volume in the flasks, for every time point two separate dependent effect had been reported before, however rea - shake flask were analyzed. sons therefore are still unknown (Jones and Porter 1998). If a main culture for product formation was added after We speculate that a limited oxygen transfer due to the oil the preculture, the preculture was transferred already or too high concentrations of unprofitable ingredients after 55 h. 3 ml of preculture were transferred into 27 ml in the crude oil might be the reason. In the main culture of main culture medium and incubated at 25  °C and phase the supplementation with crude SBO led to an 300 rpm for 120 h. increase in product concentration up to almost 50%. The For determination of cell dry weight (CDW) 5  ml of specific productivity increased by 25% for 1 and 4 g/l and broth were pipetted to preweighed glass tubes, which by 45% for 2  g/l. Again, specific productivity increased were centrifuged at 4800 rpm and 4 °C for 10 min. After the lowest when 6  g/l crude SBO were added, namely decantation, the pellet was washed with 5  ml distilled only by 20% compared to the unsupplemented control. water and centrifuged again. The supernatant was dis - Summarizing, we clearly see a concentration-depend- carded and the pellet was dried at 95  °C for 72  h. The ent positive impact of crude SBO on fungal growth and CDW was then analyzed gravimetrically. CDW analysis productivity. was performed in triplicates. Product titer determination of the main culture supernatant was performed by HPLC SBO lot‑to‑lot variability according to (Posch et al. 2012). The lot-to-lot variability of crude SBO and the asso - ciated negative impact on the process were evalu- Statistical analysis ated with SBO samples which were collected during The lot-to-lot variability of SBO samples was analyzed an industrial fungal production process over one year. by principal component analysis (PCA) and partial least PCA of the FTIR fingerprinting spectra clearly showed squares (PLS) regression using Umetrics SIMCA 4.0 clustering according to the manufacturer along PC2 (Umeå, Sweden). Before analysis, the dataset including Hofer et al. AMB Expr (2018) 8:90 Page 4 of 8 Fig. 1 Two stage shake flask experiments with P. chrysogenum and crude SBO supplemented in the preculture media. At the transfer from preculture to main culture, 3 shake flasks were inoculated from one preculture medium in order to check reproducibility. Plot a shows the CDW after 55 h of preculture at the time point of transfer with respect to different amounts of added SBO and a control run without supplementation (ctrl). Plot b shows the corresponding product concentration at the end of the main culture Fig. 2 PCA of FTIR spectra of crude SBO samples from different manufacturers resulted in 4 PCs (R2 0.999, cum 1.000). The score plot shows clustering according to manufacturer along PC 2 (plot a). A PLS model of the specific composition of SBO and the byproduct resulted in 2 LVs (R2 0.29, 0.49) and showed some correlation between those variables. Hence, lot-to-lot variability of SBO seems to have an impact on process performance attributes (plot b). PCA of FTIR spectra of SBO samples from different manufacturers resulted in 4 PCs (R 0.999, cum 1.000). The score plot shows clustering according to manufacturer along PC 2 (plot a). A PLS model of the specific composition of SBO and the byproduct resulted in 2 LVs (R 0.29, 0.49) and showed some correlation between those variables. Hence, lot-to-lot variability of SBO seems to have an impact on process performance attributes (plot b) Extraction and analysis of lecithin from crude SBO (Fig.  2a). Analysis of the production data and the spe- We hypothesized that lecithin, which we defined as phos - cific SBO composition by PLS showed that there was a phatidylcholines delivering fatty acids, phosphor and certain correlation between the SBO composition and glycerol, might be the kMA in crude SBO triggering the the formation of unwanted byproducts (Fig. 2b) as well positive effect on fungal growth and productivity. To test as maximum product titer (data not shown). Clearly, this hypothesis we developed an extraction procedure as crude SBO is characterized by a high lot-to-lot variabil- well as a quantification method, allowing the supplemen - ity which significantly impacts fungal bioprocesses. tation of defined amounts of lecithin to fungal cultures. Hofer et al. AMB Expr (2018) 8:90 Page 5 of 8 Commonly published methods for lecithin extraction, such as SPE approaches, are quite cumbersome. Hence, we decided to adopt the quite simple method of lecithin degumming, which is a method currently used in indus- try for oil refinery (Dijkstra and Van Opstal 1989; Wie - dermann 1981). 4  ml crude SBO and 1  ml of distilled water were heated up separately to 80  °C in closed glass eprouvettes for 1  h. Then, the preheated liquids were immediately united and vortexed for 1  min. Afterwards, the eprouvettes were centrifuged at 5000 rpm at 20 °C for 10 min. The lower aqueous phase contained the lecithin. To quantify the extracted lecithin we adopted a RP- HPLC method using a C18 column from literature (Jan- gle et  al. 2013). The elution gradient was optimized, resulting in a total runtime of 20  min (Table  1). The Fig. 3 Chromatogram of crude SBO. The peak area at 13.6 min could flowrate was set to 0.5 ml/min and the column oven was be assigned to lecithin (marked in blue) set to 50 °C. Ten µl of the sample were injected and ana- lyzed at 205 nm as well as with a charged aerosol detector (CAD) using the following settings: an evaporator tem- without SBO or lecithin addition were analyzed over perature of 50 °C, 5 Hz data collection rate, a filter of 3.6 72  h cultivation time. A positive effect of crude SBO and a power function of 1.90. as well as extracted lecithin on fungal cell growth was The method was assessed by injection of crude SBO detected already after 24  h of cultivation. Both sup- and purchased lecithin standards. Lecithin was detected plements increased biomass formation by 12–15% and as various peaks at around 13 min (highlighted in Fig. 3). a higher specific growth rate (µ) in comparison to the The reason for this is that lecithin is a mixture of differ - non-supplemented control was observed (Fig. 4). Nota- ent phospholipids, which show slightly different polar - bly, crude SBO and extracted lecithin showed compa- ity. However, the method allows a relative quantification rable effects demonstrating that lecithin is in fact the of extracted lecithin which was sufficient to test our kMA in crude SBO. hypothesis. In order to get more insights into the beneficial effects In order to evaluate the successful extraction and iden- of crude SBO and extracted lecithin on biomass for- tification of lecithin from crude SBO, the extracted leci - mation a DoE with 3 different concentrations of crude thin was dried and re-dissolved in MeOH for reference SBO and extracted lecithin, respectively, was performed analytics. In short, HPLC chromatograms confirmed (Fig.  5). For data analysis an MLR model was fitted lecithin, which was also identified by ESI–MS (data not including an interaction term and quadratic effects. Reli - shown). Summarizing, simple degumming and HPLC ability of the model was demonstrated by a condition quantification allow media supplementation with defined 2 2 number of 3.082, R of 0.987, Q of 0.920, model validity amounts of extracted lecithin to fungal bioprocesses. of 0.827 and a reproducibility of 0.977. The results of the DoE substantiated former experiments with respect to Eec ff ts of crude SBO versus extracted lecithin the positive effect of both crude SBO and extracted leci - In order to test our hypothesis that lecithin in crude thin on biomass formation and an upper limit of this pos- SBO positively affects fungal bioprocesses, we per - itive effect when too much of the respective supplement formed several supplementation experiments in shake is added to the medium (shown as negative correlation of flasks and analyzed fungal growth. Time-resolved interaction and quadratic terms; Fig. 5). data for both supplements as well as for a control run Table 1 Gradient program for quantification of extracted lecithin t (min) 0 1 2 6 10 10 20 % A 95 95 0 0 0 0 0 % B 5 5 100 100 50 30 30 % C 0 0 0 0 50 70 70 Eluent A is MQ water, eluent B is acetonitrile and eluent C is isopropanol Hofer et al. AMB Expr (2018) 8:90 Page 6 of 8 SBO. The comparability of the effect of crude SBO and extracted lecithin was shown with respect to biomass formation. As growth and productivity are correlated in P. chrysogenum processes, we can assume an equal effect on productivity (Douma et al. 2010). Discussion Complex raw materials, which are multicomponent mixtures of biological origin, are often used as cheap media supplements for different organisms. However, the specific kMA in these mixtures triggering the posi - tive effects on biomass growth and productivity are mostly unknown. Furthermore, these raw materials underlie high lot-to-lot variability (Hofer and Herwig Fig. 4 Shake flask experiments with a P. chrysogenum strain 2017; Hofer et  al. 2018) which naturally leads to vari- and different media supplementation. Media were either not supplemented or supplemented with crude SBO or an equivalent ations in the processes and contradicts QbD guidelines amount of extracted lecithin (Saha and Racine 2010) (ICH 2009). Vegetable oils, such as soy bean oil (SBO), belong to these complex raw materials and are often used as media supplements in Finally, we compared crude SBO and extracted lecithin fungal bioprocesses. However, current QbD guidelines in 6 reproducibility experiments. These results indicated demand for science-based knowledge of raw material less variations in biomass growth when extracted leci- quality and the knowledge of mechanistic correlations thin was used (4% variation for crude SBO and 1% for between raw material attributes and their effects. In extracted lecithin; see standard variations in Fig. 6) argu- this study we hypothesized that lecithin is the kMA in ing for an increased use of this less variable raw material SBO triggering the positive effects on fungal growth compared to crude SBO. and productivity. To test our hypothesis we adapted a In summary, we could identify lecithin as a kMA in rather simple extraction method for lecithin from SBO crude SBO for the evaluated process. A fast and cheap based on lecithin degumming and developed an ana- extraction procedure for lecithin combined with an lytical method allowing the supplementation of defined HPLC method for optimal dosage was developed facili- amounts of either crude SBO or extracted lecithin. In a tating the potential application of a less variable media series of supplementation experiments with a P. chrys- supplement, namely extracted lecithin from crude ogenum strain we could demonstrate that lecithin is in Fig. 5 The data of the 23 DoE were fitted using an MLR model. The coefficient plot (plot a) shows that crude SBO as well as extracted lecithin positively correlate to the CDW. Nevertheless, too high concentrations seem to have a negative impact, which can be seen in the negative interaction and quadratic terms. Plot b represents the observed vs predicted plot of the model, showing a good model fit Hofer et al. AMB Expr (2018) 8:90 Page 7 of 8 Acknowledgements The authors would like to thank Sandoz GmbH for providing the strain and the production process data. Additionally, we want to thank Barbara Mahlberg, Kelly Briffa, Donya Kamravamanesh, Winfried Nischkauer, Sofia Milker and Florian Rudroff for support in the lab. The authors acknowledge the TU Wien University Library for financial support through its Open Access Funding Program. Competing interests The authors declare that they have no competing interests. Availability of data and materials The data supporting the conclusions is presented in the main article. Consent for publication Not applicable. Ethics approval and consent to participate Not applicable. Fig. 6 Six experiments were performed either with crude SBO or extracted lecithin as media supplement. The plot shows the variance Funding between those experiments concerning biomass formation This study was funded by the Austrian research funding association (FFG) under the scope of the COMET program within the research project “Industrial Methods for Process Analytical Chemistry - From Measurement Technologies to Information Systems (imPACts)” (grant number # 843546). fact the kMA in crude SBO causing beneficial effects on fungal growth. We assume an equal positive effect Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- on productivity due to the correlation between growth lished maps and institutional affiliations. an productivity in this process (Douma et  al. 2010). For both crude SBO an extracted lecithin, these effects Received: 14 May 2018 Accepted: 28 May 2018 were concentration-dependent. In reproducibility stud- ies we found that extracted lecithin caused less varia- tion in biomass growth than crude SBO, which would References be highly favorable in production processes. As the Anderson RF, Tornqvist EGM, Peterson WH (1956) Eec ff t of oil in pilot plant method of lecithin degumming is already used in large fermentations. J Agr Food Chem 4(6):556–559. https ://doi.org/10.1021/ scale refinery processes, we believe that it can also be jf600 64a01 0 Bateman HG, Jenkins TC (1997) Method for extraction and separation by solid introduced as raw material preparation step for bio- phase extraction of neutral lipid, free fatty acids, and polar lipid from processes. Alternatively, we propose to quantify the mixed microbial cultures. J Agr Food Chem 45(1):132–134. https ://doi. amount of lecithin in crude SBO to have a means of org/10.1021/jf960 3339 Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purifica- evaluation for this complex raw material and to be able tion. Can J Biochem Phys 37(8):911–917 to avoid over-dosage of the supplement and thus nega- Carelli AA, Brevedan MIV, Crapiste GH (1997) Quantitative determination of tive effects. phospholipids in sunflower oil. J Am Oil Chem Soc 74(5):511–514. https :// doi.org/10.1007/s1174 6-997-0173-2 Dijkstra AJ, Van Opstal M (1989) The total degumming process. J Am Oil Chem Soc 66(7):1002–1009. https ://doi.org/10.1007/bf026 82627 Abbreviations Douma RD, Verheijen PJ, de Laat WT, Heijnen JJ, van Gulik WM (2010) Dynamic CAD: charged aerosol detection; CMA: critical material attribute; DoE: design gene expression regulation model for growth and penicillin production of experiment; ESI–MS: electrospray ionization–mass spectrometry; FFA: free in Penicillium chrysogenum. Biotechnol Bioeng 106(4):608–618. https :// fatty acids; FAME: fatty acid methyl esters; FTIR: fourier transform infrared doi.org/10.1002/bit.22689 spectroscopy; HPLC: high performance liquid chromatography; ICP-OES: Gao Y, Yuan YJ (2011) Comprehensive quality evaluation of corn steep liquor in inductively coupled plasma-optical emission spectrometer; kMA: key material 2-Keto-L-gulonic acid fermentation. J Agr Food Chem 59(18):9845–9853. attribute; µ: growth rate; PCA: principal component analysis; PLS: partial least https ://doi.org/10.1021/jf201 792u squares; QbD: quality by design; SBO: soy bean oil. Goldschmidt MC, Koffler H (1950) Eec ff t of surface-active agents on penicillin yields. Ind Eng Chem 42(9):1819–1823. https ://doi.org/10.1021/ie504 Authors’ contributions 89a03 5 AHO and OSP planned the experiments and wrote the manuscript. AHO Hofer A, Herwig C (2017) Quantitative determination of nine water-soluble performed the experiments and analysed the data. CHE was responsible for vitamins in the complex matrix of corn steep liquor for rawmaterial initiation and supervision of the study. All authors read and approved the final quality assessment. J Chem Technol Biot 92(8):2106–2113. https ://doi. manuscript. org/10.1002/jctb.5211 Hofer A, Kamravamanesh D, Bona-Lovasz J, Limbeck A, Lendl B, Herwig C, Author details Fricke J (2018) Prediction of filamentous process performance attrib - Research Area Biochemical Engineering, Institute of Chemical, Environmental utes by CSL quality assessment using mid-infrared spectroscopy and and Bioscience Engineering, Vienna University of Technology, Gumpendor- chemometrics. J Biotechnol 265:93–100. https ://doi.org/10.1016/j.jbiot ferstrasse 1a - 166/4, 1060 Vienna, Austria. CD Laboratory on Mechanistic ec.2017.11.010 and Physiological Methods for Improved Bioprocesses, Vienna University of Technology, Gumpendorferstrasse 1a - 166/4, 1060 Vienna, Austria. Hofer et al. AMB Expr (2018) 8:90 Page 8 of 8 ICH (2009) Harmonised Tripartite Guideline: Pharmaceutical Development Q8 Posch AE, Koch C, Helmel M, Marchetti-Deschmann M, Macfelda K, Lendl (R2) B, Allmaier G, Herwig C (2013) Combining light microscopy, dielectric Jangle RD, Galge RV, Patil VV, Thorat BN (2013) Selective HPLC method devel- spectroscopy, MALDI intact cell mass spectrometry, FTIR spectromicros- opment for soy phosphatidylcholine fatty acids and its mass spectrom- copy and multivariate data mining for morphological and physiological etry. Indian J Pharm Sci 75(3):339–345. https ://doi.org/10.4103/0250- bioprocess characterization of filamentous organisms. Fungal Genet Biol 474x.11743 5 51:1–11. https ://doi.org/10.1016/j.fgb.2012.11.008 Jones AM, Porter MA (1998) Vegetable oils in fermentation: beneficial effects Reese ET, Maguire A (1969) Surfactants as stimulantes of enzyme production of low-level supplementation. J Ind Microbiol Biot 21(4–5):203–207. https by microorganisms. Appl Microbiol 17(2):242–245 ://doi.org/10.1038/sj.jim.29005 71 Ruiz-Gutierrez V, Perez-Camino MC (2000) Update on solid-phase extraction Millis NF, Palmer BM, Trumpy BH (1963) Eec ff t of lipids on citric acid production for the analysis of lipid classes and related compounds. J Chromatogr A by an Aspergillus Niger mutant. J Gen Microbiol 30(3):365–379 885(1–2):321–341. https ://doi.org/10.1016/S0021 -9673(00)00181 -3 Nash AM, Frankel EN (1986) Limited extraction of soybeans with hexane. J Am Saha BC, Racine FM (2010) Eec ff ts of pH and corn steep liquor variability on Oil Chem Soc 63(2):244–246. https ://doi.org/10.1007/Bf025 46147 mannitol production by Lactobacillus intermedius NRRL B-3693. Appl Ohta N, Park YS, Yahiro K, Okabe M (1995) Comparison of neomycin produc- Microbiol Biot 87(2):553–560. https ://doi.org/10.1007/s0025 3-010-2552-0 tion from Streptomyces-Fradiae cultivation using soybean oil as the sole Tarafdar JC, Claassen N (1988) Organic phosphorus compounds as a phospho- carbon source in an airlift bioreactor and a stirred-tank reactor. J Ferment rus source for higher plants through the activity of phosphatases pro- Bioeng 79(5):443–448. https ://doi.org/10.1016/0922-338x(95)91259 -8 duced by plant roots and microorganisms. Biol Fertility Soils 5(4):308–312. Pan SC, Bonanno S, Wagman GH (1959) Efficient utilization of fatty oils as https ://doi.org/10.1007/bf002 62137 energy source in penicillin fermentation. Appl Microbiol 7(3):176–180 Wiedermann LH (1981) Degumming, refining and bleaching soybean oil. J Am Posch AE, Spadiut O, Herwig C (2012) Switching industrial production pro- Oil Chem Soc 58(3):159–166. https ://doi.org/10.1007/bf025 82328 cesses from complex to defined media: method development and case study using the example of Penicillium chrysogenum. Microb Cell Fact. https ://doi.org/10.1186/1475-2859-11-88 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png AMB Express Springer Journals

Lecithin is the key material attribute in soy bean oil affecting filamentous bioprocesses

Free
8 pages
Loading next page...
 
/lp/springer_journal/lecithin-is-the-key-material-attribute-in-soy-bean-oil-affecting-zGdTKrrJog
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2018 by The Author(s)
Subject
Life Sciences; Microbiology; Microbial Genetics and Genomics; Biotechnology
eISSN
2191-0855
D.O.I.
10.1186/s13568-018-0625-0
Publisher site
See Article on Publisher Site

Abstract

Complex raw materials are widely used as supplements in biopharmaceutical production processes due to their posi- tive effect on biomass growth and productivity at low cost. However, their use negatively impacts process reproduc- ibility due to high lot-to-lot variability which contradicts current regulatory guidelines. In this study we investigated crude soy bean oil (SBO) which is a common complex raw material for filamentous fungi. We demonstrated that lecithin, which we define as phosphatidylcholines, is in fact the key material attribute in crude SBO positively affect - ing fungal growth and consequently productivity. The methodological toolbox we present here allows the straight- forward isolation of lecithin from crude SBO, its semi-quantification by HPLC and the consequent supplementation thereof in defined amounts. Thus, over-dosage and potential resulting negative impacts on fungal growth and productivity can be omitted. Keywords: Complex raw material, Key material attribute, Soy bean oil, Lecithin, Filamentous fungi An example for complex biological raw materials are Introduction vegetable oils, such as soy bean oil (SBO). It was shown Complex raw materials are commonly used as cheap that SBO leads to an increase in fungal biomass growth media supplements in bacterial, fungal and mammalian and a 50% increase in antibiotics production (Gold- bioprocesses (Gao and Yuan 2011; Millis et al. 1963; Reese schmidt and Koffler 1950; Pan et al. 1959). Although the and Maguire 1969). These raw materials are typically of reasons for this positive effect are still obscure, possible biological origin and thus underlie a high lot-to-lot varia- explanations are: (i) SBO serves as nutritional source, (ii) bility. Furthermore, the specific substance responsible for SBO changes the medium for improved oxygen transfer, the positive impact on the bioprocess is often unknown. and (iii) the antifoam capacity of the oil reduces micro- However, during the International Conference of Har- bial cell damage (Goldschmidt and Koffler 1950; Jones monization, guidelines have been established demanding and Porter 1998). Two critical factors for SBO supple- for the evaluation of critical material attributes (CMAs) mentation are known, namely (i) the time point of addi- and their impact on product quality (ICH 2009). Mate- tion and (ii) the dosage (Anderson et al. 1956; Ohta et al. rial attributes do not only affect product quality but also 1995). It has been shown that low-level supplementation process performance and productivity. We define these of SBO is beneficial, whereas medium to high-level sup - material attributes that have an impact on productivity plementation is disadvantageous. To shed more light on as key material attributes (kMAs). For both, identification that, the effect of crude and refined SBO was tested, how - of CMAs and kMAs, sound science-based knowledge of ever no significant differences in the production of tet - raw material quality has become a necessity. racycline were found (Jones and Porter 1998). Although crude SBO is about 1/5 cheaper than refined oil and con - tains phospholipids, some minerals and tocopherols, it may also contain pesticides, which is why refined SBO is *Correspondence: oliver.spadiut@tuwien.ac.at Research Area Biochemical Engineering, Institute of Chemical, mostly used as medium supplement today. However, sev- Environmental and Bioscience Engineering, Vienna University eral studies have shown the positive impact of phospho- of Technology, Gumpendorferstrasse 1a - 166/4, 1060 Vienna, Austria lipids in crude SBO on fungal growth and productivity Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/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. Hofer et al. AMB Expr (2018) 8:90 Page 2 of 8 (Bateman and Jenkins 1997; Goldschmidt and Koffler assessing correlations between SBO composition and 1950; Tarafdar and Claassen 1988). We hypothesize that byproduct as well as product formation. lecithin, which we define as phosphatidylcholines, is the kMA in SBO causing this positive impact. To investigate Analytical methods this hypothesis, we isolated lecithin from SBO, added HPLC method for crude SBO and lecithin defined amounts thereof to the media for a Penicillium Chromatographic separation was accomplished by adopt- chrysogenum strain and analyzed the effects on fungal ing the method from Jangle et  al. (2013) using a Zorbax growth. For the extraction of lecithin from oily matri- Eclipse Plus C-18 column (3.0 × 150 mm, 3.5 µm; Agilent ces several solid phase extraction (SPE) methods have Technologies) and pre-column (4.6 × 12.5  mm, 5  µm). been reported (Bateman and Jenkins 1997; Bligh and Further details of the chromatographic method are given Dyer 1959; Carelli et  al. 1997; Nash and Frankel 1986; in the Results section. An Ultimate 3000 HPLC system Ruiz-Gutierrez and Perez-Camino 2000). However, SPE (Thermo Fisher Scientific, Waltham, MA, USA) equipped is quite tedious and time consuming and delivers the with a pump (LPG-3400SD), an autosampler (WPS-3000 wanted lecithin in an organic solvent, which is com- SplitLoop), a column oven (Col.Comp. TCC-3000SD), a monly not suitable for media supplementation. Hence, diode array detector (DAD 3000) and a charged aerosol we adopted a simple and cheap extraction procedure, detector (Corona Veo RS) was used. Chromeleon 7.2 was which is currently applied in industrial refinery processes used for control. (Dijkstra and Van Opstal 1989), delivering lecithin in an aqueous background. Additionally, we present a semi- HPLC method for free fatty acids (FFA) quantitative HPLC-CAD method for crude SBO and For sample preparation, FFA were extracted with butyl lecithin analysis, which allows dosage of the extracted acetate at room temperature. The separation of FFA, lecithin without the need of mass spectrometry analysis. namely 18:0, 18:1, 18:2 and 18:3, was achieved by a Zor- The effect of crude SBO and extracted lecithin on fungal bax Eclipse Plus C-18 column (3.0 × 150  mm, 3.5  µm; growth was evaluated in shake flask experiments with P. Agilent Technologies) and pre-column (4.6 × 12.5  mm, chrysogenum. Summarizing, in this study we demonstrate 5 µm). A gradient was implemented over 4 min, starting that lecithin is in fact the kMA in SBO affecting fungal from 90% MQ to 100% acetonitrile, which was held for growth and consequently productivity. another 6  min. FFA were detected by a charged aerosol detector. The method parameters were: 1.0 ml/min flow - Materials and methods rate, 50  °C column oven temperature and 2 µl injection Chemicals volume. All chemicals were of analytical grade and purchased from Carl Roth (Karlsruhe, Germany) or Sigma Aldrich GC method for fatty acid methyl esters (FAME) (St. Louis, MO, USA). Ultra-pure water derived from a An Agilent Gas Chromatography system (7890A, Agi- Milli-Q system from Merck Millipore (Billerica, MA, lent Technologies, Santa Clara, California, USA) with USA) was used for analytical methods. a flame ionisation detector (FID) was used. Chromato - graphic separation was achieved by an HP-88 column SBO samples (60 m × 0.25 mm ID, 0.2 µm; Agilent) and a temperature Unrefined soy bean oil (SBO) originated from two Euro - gradient starting at 40 °C up to 220 °C within 80 min. 1 µl pean manufacturers (called M1 and M2). was injected with an inlet split of 10:1. Prior to injection, the samples were derivatised with trimethylsulfonium SBO lot‑to‑lot variability hydroxide (TMSH) for esterification. Therefore 200  µl For demonstration of SBO lot-to-lot variability, 20 sam- sample, 200  µl internal standard solution and 200  µl ples from manufacturer M1 and M2 were analyzed for TMSH were incubated at 50  °C for 15  min. Quantifica - free fatty acid (FFA), fatty acid methyl esters (FAME) tion was achieved via external and internal standards, e.g. and phosphorous content as well as for a spectroscopic 19:0. Standards were prepared of 14:0, 16:0, 16:1, 18:0, fingerprint by Fourier transform infrared spectroscopy 18:1, 18:2 and 18:3 fatty acids. (FTIR). Data were analyzed by principal component anal- ysis (PCA). ICP‑OES for phosphor analysis Furthermore, our industrial cooperation partner deliv- The samples were diluted in PremiSolv (SPrep, Ger- ered 29 SBO samples as well as on-line and off-line data many) and indium was added as internal standard. Sam- of an industrial scale fungal production process. In order ples were analyzed on an iCAP 6000 inductively coupled to assess the impact of SBO variability on process per- plasma—optical emission spectrometer (ICP-OES, formance, multivariate data analysis was performed, Thermo Scientific). The instrument was equipped with Hofer et al. AMB Expr (2018) 8:90 Page 3 of 8 a radial optic, an Echelle spectrometer and a CCD-chip the specific composition of SBO as well as the dataset detector. The samples were brought into the plasma using including the byproduct concentration were normalized a Babington-type nebulizer and a peristaltic pump. and mean-centered. From the spectral data, i.e. the fin - gerprints, the second derivative was generated in order to FTIR analysis for SBO fingerprint assure baseline correction and the data were mean-cen- Crude SBO samples were directly analyzed using an tered. For PCA, the final matrix consisted of 12 observa - −1 FTIR instrument equipped with an ATR Platinum, a tions and the FTIR region between 800 and 2000  cm . Tensor 37 photometer and a DTGS detector (Bruker For PLS analysis the x-matrix consisted of 39 observa- optics). Spectra were acquired with a resolution of tions and 12 specific components, namely FFA, FAME −1 4 cm and 32 scans per sample against a background of and phosphor, and the y-vector of 39 observations and air. The acquired wavenumber region ranged from 600 to the maximal byproduct concentration. −1 4000 cm . The design of experiment (DoE) investigating the impact of crude SBO and extracted lecithin as well as Shake flask cultivations interactions on process performance was designed with A P-14 P. chrysogenum candidate strain, descending from MODDE Umetrics (Umeå, Sweden), choosing a full fac- the P-2 P. chrysogenum candidate strain (American Type torial screening with 2 factors, namely crude SBO and Culture Collection with the access number ATCC 48271), extracted lecithin, and 3 levels, namely 0, 4 or 8 g/l, and was used for all experiments. The cultivation medium resulted in 11 experiments including 3 center points. was similar as described previously (Posch et  al. 2013), either with or without supplementation of crude SBO Results or extracted lecithin. In every case glucose was the main SBO as typical complex raw material C-source. For preculture, 500  ml shake flasks were filled Typical characteristics of complex raw materials are their with 30 ml medium and inoculated with 2.8 × 10 spores positive impact on growth and productivity at low cost, of P. chrysogenum. The flasks were incubated at 25 °C and but also their negative impact on process reproducibility 300 rpm for up to 72 h. A flask filled with destilled water due to high lot-to-lot variability. We assessed both char- was added in the incubator in order to increase humidity acteristics for crude SBO in a P. chrysogenum antibiotics and reduce evaporation. The flasks were either harvested production process in shake flask experiments (Fig.  1). In at the end-point only after 72  h or more sample points the preculture, supplementation with up to 4  g/l crude were evaluated in order to get time-resolved data. For the SBO led to an increase in biomass formation. A higher latter, samples were analyzed every 12 h. Due to the low concentration of SBO had a negative effect. This dosage- volume in the flasks, for every time point two separate dependent effect had been reported before, however rea - shake flask were analyzed. sons therefore are still unknown (Jones and Porter 1998). If a main culture for product formation was added after We speculate that a limited oxygen transfer due to the oil the preculture, the preculture was transferred already or too high concentrations of unprofitable ingredients after 55 h. 3 ml of preculture were transferred into 27 ml in the crude oil might be the reason. In the main culture of main culture medium and incubated at 25  °C and phase the supplementation with crude SBO led to an 300 rpm for 120 h. increase in product concentration up to almost 50%. The For determination of cell dry weight (CDW) 5  ml of specific productivity increased by 25% for 1 and 4 g/l and broth were pipetted to preweighed glass tubes, which by 45% for 2  g/l. Again, specific productivity increased were centrifuged at 4800 rpm and 4 °C for 10 min. After the lowest when 6  g/l crude SBO were added, namely decantation, the pellet was washed with 5  ml distilled only by 20% compared to the unsupplemented control. water and centrifuged again. The supernatant was dis - Summarizing, we clearly see a concentration-depend- carded and the pellet was dried at 95  °C for 72  h. The ent positive impact of crude SBO on fungal growth and CDW was then analyzed gravimetrically. CDW analysis productivity. was performed in triplicates. Product titer determination of the main culture supernatant was performed by HPLC SBO lot‑to‑lot variability according to (Posch et al. 2012). The lot-to-lot variability of crude SBO and the asso - ciated negative impact on the process were evalu- Statistical analysis ated with SBO samples which were collected during The lot-to-lot variability of SBO samples was analyzed an industrial fungal production process over one year. by principal component analysis (PCA) and partial least PCA of the FTIR fingerprinting spectra clearly showed squares (PLS) regression using Umetrics SIMCA 4.0 clustering according to the manufacturer along PC2 (Umeå, Sweden). Before analysis, the dataset including Hofer et al. AMB Expr (2018) 8:90 Page 4 of 8 Fig. 1 Two stage shake flask experiments with P. chrysogenum and crude SBO supplemented in the preculture media. At the transfer from preculture to main culture, 3 shake flasks were inoculated from one preculture medium in order to check reproducibility. Plot a shows the CDW after 55 h of preculture at the time point of transfer with respect to different amounts of added SBO and a control run without supplementation (ctrl). Plot b shows the corresponding product concentration at the end of the main culture Fig. 2 PCA of FTIR spectra of crude SBO samples from different manufacturers resulted in 4 PCs (R2 0.999, cum 1.000). The score plot shows clustering according to manufacturer along PC 2 (plot a). A PLS model of the specific composition of SBO and the byproduct resulted in 2 LVs (R2 0.29, 0.49) and showed some correlation between those variables. Hence, lot-to-lot variability of SBO seems to have an impact on process performance attributes (plot b). PCA of FTIR spectra of SBO samples from different manufacturers resulted in 4 PCs (R 0.999, cum 1.000). The score plot shows clustering according to manufacturer along PC 2 (plot a). A PLS model of the specific composition of SBO and the byproduct resulted in 2 LVs (R 0.29, 0.49) and showed some correlation between those variables. Hence, lot-to-lot variability of SBO seems to have an impact on process performance attributes (plot b) Extraction and analysis of lecithin from crude SBO (Fig.  2a). Analysis of the production data and the spe- We hypothesized that lecithin, which we defined as phos - cific SBO composition by PLS showed that there was a phatidylcholines delivering fatty acids, phosphor and certain correlation between the SBO composition and glycerol, might be the kMA in crude SBO triggering the the formation of unwanted byproducts (Fig. 2b) as well positive effect on fungal growth and productivity. To test as maximum product titer (data not shown). Clearly, this hypothesis we developed an extraction procedure as crude SBO is characterized by a high lot-to-lot variabil- well as a quantification method, allowing the supplemen - ity which significantly impacts fungal bioprocesses. tation of defined amounts of lecithin to fungal cultures. Hofer et al. AMB Expr (2018) 8:90 Page 5 of 8 Commonly published methods for lecithin extraction, such as SPE approaches, are quite cumbersome. Hence, we decided to adopt the quite simple method of lecithin degumming, which is a method currently used in indus- try for oil refinery (Dijkstra and Van Opstal 1989; Wie - dermann 1981). 4  ml crude SBO and 1  ml of distilled water were heated up separately to 80  °C in closed glass eprouvettes for 1  h. Then, the preheated liquids were immediately united and vortexed for 1  min. Afterwards, the eprouvettes were centrifuged at 5000 rpm at 20 °C for 10 min. The lower aqueous phase contained the lecithin. To quantify the extracted lecithin we adopted a RP- HPLC method using a C18 column from literature (Jan- gle et  al. 2013). The elution gradient was optimized, resulting in a total runtime of 20  min (Table  1). The Fig. 3 Chromatogram of crude SBO. The peak area at 13.6 min could flowrate was set to 0.5 ml/min and the column oven was be assigned to lecithin (marked in blue) set to 50 °C. Ten µl of the sample were injected and ana- lyzed at 205 nm as well as with a charged aerosol detector (CAD) using the following settings: an evaporator tem- without SBO or lecithin addition were analyzed over perature of 50 °C, 5 Hz data collection rate, a filter of 3.6 72  h cultivation time. A positive effect of crude SBO and a power function of 1.90. as well as extracted lecithin on fungal cell growth was The method was assessed by injection of crude SBO detected already after 24  h of cultivation. Both sup- and purchased lecithin standards. Lecithin was detected plements increased biomass formation by 12–15% and as various peaks at around 13 min (highlighted in Fig. 3). a higher specific growth rate (µ) in comparison to the The reason for this is that lecithin is a mixture of differ - non-supplemented control was observed (Fig. 4). Nota- ent phospholipids, which show slightly different polar - bly, crude SBO and extracted lecithin showed compa- ity. However, the method allows a relative quantification rable effects demonstrating that lecithin is in fact the of extracted lecithin which was sufficient to test our kMA in crude SBO. hypothesis. In order to get more insights into the beneficial effects In order to evaluate the successful extraction and iden- of crude SBO and extracted lecithin on biomass for- tification of lecithin from crude SBO, the extracted leci - mation a DoE with 3 different concentrations of crude thin was dried and re-dissolved in MeOH for reference SBO and extracted lecithin, respectively, was performed analytics. In short, HPLC chromatograms confirmed (Fig.  5). For data analysis an MLR model was fitted lecithin, which was also identified by ESI–MS (data not including an interaction term and quadratic effects. Reli - shown). Summarizing, simple degumming and HPLC ability of the model was demonstrated by a condition quantification allow media supplementation with defined 2 2 number of 3.082, R of 0.987, Q of 0.920, model validity amounts of extracted lecithin to fungal bioprocesses. of 0.827 and a reproducibility of 0.977. The results of the DoE substantiated former experiments with respect to Eec ff ts of crude SBO versus extracted lecithin the positive effect of both crude SBO and extracted leci - In order to test our hypothesis that lecithin in crude thin on biomass formation and an upper limit of this pos- SBO positively affects fungal bioprocesses, we per - itive effect when too much of the respective supplement formed several supplementation experiments in shake is added to the medium (shown as negative correlation of flasks and analyzed fungal growth. Time-resolved interaction and quadratic terms; Fig. 5). data for both supplements as well as for a control run Table 1 Gradient program for quantification of extracted lecithin t (min) 0 1 2 6 10 10 20 % A 95 95 0 0 0 0 0 % B 5 5 100 100 50 30 30 % C 0 0 0 0 50 70 70 Eluent A is MQ water, eluent B is acetonitrile and eluent C is isopropanol Hofer et al. AMB Expr (2018) 8:90 Page 6 of 8 SBO. The comparability of the effect of crude SBO and extracted lecithin was shown with respect to biomass formation. As growth and productivity are correlated in P. chrysogenum processes, we can assume an equal effect on productivity (Douma et al. 2010). Discussion Complex raw materials, which are multicomponent mixtures of biological origin, are often used as cheap media supplements for different organisms. However, the specific kMA in these mixtures triggering the posi - tive effects on biomass growth and productivity are mostly unknown. Furthermore, these raw materials underlie high lot-to-lot variability (Hofer and Herwig Fig. 4 Shake flask experiments with a P. chrysogenum strain 2017; Hofer et  al. 2018) which naturally leads to vari- and different media supplementation. Media were either not supplemented or supplemented with crude SBO or an equivalent ations in the processes and contradicts QbD guidelines amount of extracted lecithin (Saha and Racine 2010) (ICH 2009). Vegetable oils, such as soy bean oil (SBO), belong to these complex raw materials and are often used as media supplements in Finally, we compared crude SBO and extracted lecithin fungal bioprocesses. However, current QbD guidelines in 6 reproducibility experiments. These results indicated demand for science-based knowledge of raw material less variations in biomass growth when extracted leci- quality and the knowledge of mechanistic correlations thin was used (4% variation for crude SBO and 1% for between raw material attributes and their effects. In extracted lecithin; see standard variations in Fig. 6) argu- this study we hypothesized that lecithin is the kMA in ing for an increased use of this less variable raw material SBO triggering the positive effects on fungal growth compared to crude SBO. and productivity. To test our hypothesis we adapted a In summary, we could identify lecithin as a kMA in rather simple extraction method for lecithin from SBO crude SBO for the evaluated process. A fast and cheap based on lecithin degumming and developed an ana- extraction procedure for lecithin combined with an lytical method allowing the supplementation of defined HPLC method for optimal dosage was developed facili- amounts of either crude SBO or extracted lecithin. In a tating the potential application of a less variable media series of supplementation experiments with a P. chrys- supplement, namely extracted lecithin from crude ogenum strain we could demonstrate that lecithin is in Fig. 5 The data of the 23 DoE were fitted using an MLR model. The coefficient plot (plot a) shows that crude SBO as well as extracted lecithin positively correlate to the CDW. Nevertheless, too high concentrations seem to have a negative impact, which can be seen in the negative interaction and quadratic terms. Plot b represents the observed vs predicted plot of the model, showing a good model fit Hofer et al. AMB Expr (2018) 8:90 Page 7 of 8 Acknowledgements The authors would like to thank Sandoz GmbH for providing the strain and the production process data. Additionally, we want to thank Barbara Mahlberg, Kelly Briffa, Donya Kamravamanesh, Winfried Nischkauer, Sofia Milker and Florian Rudroff for support in the lab. The authors acknowledge the TU Wien University Library for financial support through its Open Access Funding Program. Competing interests The authors declare that they have no competing interests. Availability of data and materials The data supporting the conclusions is presented in the main article. Consent for publication Not applicable. Ethics approval and consent to participate Not applicable. Fig. 6 Six experiments were performed either with crude SBO or extracted lecithin as media supplement. The plot shows the variance Funding between those experiments concerning biomass formation This study was funded by the Austrian research funding association (FFG) under the scope of the COMET program within the research project “Industrial Methods for Process Analytical Chemistry - From Measurement Technologies to Information Systems (imPACts)” (grant number # 843546). fact the kMA in crude SBO causing beneficial effects on fungal growth. We assume an equal positive effect Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- on productivity due to the correlation between growth lished maps and institutional affiliations. an productivity in this process (Douma et  al. 2010). For both crude SBO an extracted lecithin, these effects Received: 14 May 2018 Accepted: 28 May 2018 were concentration-dependent. In reproducibility stud- ies we found that extracted lecithin caused less varia- tion in biomass growth than crude SBO, which would References be highly favorable in production processes. As the Anderson RF, Tornqvist EGM, Peterson WH (1956) Eec ff t of oil in pilot plant method of lecithin degumming is already used in large fermentations. J Agr Food Chem 4(6):556–559. https ://doi.org/10.1021/ scale refinery processes, we believe that it can also be jf600 64a01 0 Bateman HG, Jenkins TC (1997) Method for extraction and separation by solid introduced as raw material preparation step for bio- phase extraction of neutral lipid, free fatty acids, and polar lipid from processes. Alternatively, we propose to quantify the mixed microbial cultures. J Agr Food Chem 45(1):132–134. https ://doi. amount of lecithin in crude SBO to have a means of org/10.1021/jf960 3339 Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purifica- evaluation for this complex raw material and to be able tion. Can J Biochem Phys 37(8):911–917 to avoid over-dosage of the supplement and thus nega- Carelli AA, Brevedan MIV, Crapiste GH (1997) Quantitative determination of tive effects. phospholipids in sunflower oil. J Am Oil Chem Soc 74(5):511–514. https :// doi.org/10.1007/s1174 6-997-0173-2 Dijkstra AJ, Van Opstal M (1989) The total degumming process. J Am Oil Chem Soc 66(7):1002–1009. https ://doi.org/10.1007/bf026 82627 Abbreviations Douma RD, Verheijen PJ, de Laat WT, Heijnen JJ, van Gulik WM (2010) Dynamic CAD: charged aerosol detection; CMA: critical material attribute; DoE: design gene expression regulation model for growth and penicillin production of experiment; ESI–MS: electrospray ionization–mass spectrometry; FFA: free in Penicillium chrysogenum. Biotechnol Bioeng 106(4):608–618. https :// fatty acids; FAME: fatty acid methyl esters; FTIR: fourier transform infrared doi.org/10.1002/bit.22689 spectroscopy; HPLC: high performance liquid chromatography; ICP-OES: Gao Y, Yuan YJ (2011) Comprehensive quality evaluation of corn steep liquor in inductively coupled plasma-optical emission spectrometer; kMA: key material 2-Keto-L-gulonic acid fermentation. J Agr Food Chem 59(18):9845–9853. attribute; µ: growth rate; PCA: principal component analysis; PLS: partial least https ://doi.org/10.1021/jf201 792u squares; QbD: quality by design; SBO: soy bean oil. Goldschmidt MC, Koffler H (1950) Eec ff t of surface-active agents on penicillin yields. Ind Eng Chem 42(9):1819–1823. https ://doi.org/10.1021/ie504 Authors’ contributions 89a03 5 AHO and OSP planned the experiments and wrote the manuscript. AHO Hofer A, Herwig C (2017) Quantitative determination of nine water-soluble performed the experiments and analysed the data. CHE was responsible for vitamins in the complex matrix of corn steep liquor for rawmaterial initiation and supervision of the study. All authors read and approved the final quality assessment. J Chem Technol Biot 92(8):2106–2113. https ://doi. manuscript. org/10.1002/jctb.5211 Hofer A, Kamravamanesh D, Bona-Lovasz J, Limbeck A, Lendl B, Herwig C, Author details Fricke J (2018) Prediction of filamentous process performance attrib - Research Area Biochemical Engineering, Institute of Chemical, Environmental utes by CSL quality assessment using mid-infrared spectroscopy and and Bioscience Engineering, Vienna University of Technology, Gumpendor- chemometrics. J Biotechnol 265:93–100. https ://doi.org/10.1016/j.jbiot ferstrasse 1a - 166/4, 1060 Vienna, Austria. CD Laboratory on Mechanistic ec.2017.11.010 and Physiological Methods for Improved Bioprocesses, Vienna University of Technology, Gumpendorferstrasse 1a - 166/4, 1060 Vienna, Austria. Hofer et al. AMB Expr (2018) 8:90 Page 8 of 8 ICH (2009) Harmonised Tripartite Guideline: Pharmaceutical Development Q8 Posch AE, Koch C, Helmel M, Marchetti-Deschmann M, Macfelda K, Lendl (R2) B, Allmaier G, Herwig C (2013) Combining light microscopy, dielectric Jangle RD, Galge RV, Patil VV, Thorat BN (2013) Selective HPLC method devel- spectroscopy, MALDI intact cell mass spectrometry, FTIR spectromicros- opment for soy phosphatidylcholine fatty acids and its mass spectrom- copy and multivariate data mining for morphological and physiological etry. Indian J Pharm Sci 75(3):339–345. https ://doi.org/10.4103/0250- bioprocess characterization of filamentous organisms. Fungal Genet Biol 474x.11743 5 51:1–11. https ://doi.org/10.1016/j.fgb.2012.11.008 Jones AM, Porter MA (1998) Vegetable oils in fermentation: beneficial effects Reese ET, Maguire A (1969) Surfactants as stimulantes of enzyme production of low-level supplementation. J Ind Microbiol Biot 21(4–5):203–207. https by microorganisms. Appl Microbiol 17(2):242–245 ://doi.org/10.1038/sj.jim.29005 71 Ruiz-Gutierrez V, Perez-Camino MC (2000) Update on solid-phase extraction Millis NF, Palmer BM, Trumpy BH (1963) Eec ff t of lipids on citric acid production for the analysis of lipid classes and related compounds. J Chromatogr A by an Aspergillus Niger mutant. J Gen Microbiol 30(3):365–379 885(1–2):321–341. https ://doi.org/10.1016/S0021 -9673(00)00181 -3 Nash AM, Frankel EN (1986) Limited extraction of soybeans with hexane. J Am Saha BC, Racine FM (2010) Eec ff ts of pH and corn steep liquor variability on Oil Chem Soc 63(2):244–246. https ://doi.org/10.1007/Bf025 46147 mannitol production by Lactobacillus intermedius NRRL B-3693. Appl Ohta N, Park YS, Yahiro K, Okabe M (1995) Comparison of neomycin produc- Microbiol Biot 87(2):553–560. https ://doi.org/10.1007/s0025 3-010-2552-0 tion from Streptomyces-Fradiae cultivation using soybean oil as the sole Tarafdar JC, Claassen N (1988) Organic phosphorus compounds as a phospho- carbon source in an airlift bioreactor and a stirred-tank reactor. J Ferment rus source for higher plants through the activity of phosphatases pro- Bioeng 79(5):443–448. https ://doi.org/10.1016/0922-338x(95)91259 -8 duced by plant roots and microorganisms. Biol Fertility Soils 5(4):308–312. Pan SC, Bonanno S, Wagman GH (1959) Efficient utilization of fatty oils as https ://doi.org/10.1007/bf002 62137 energy source in penicillin fermentation. Appl Microbiol 7(3):176–180 Wiedermann LH (1981) Degumming, refining and bleaching soybean oil. J Am Posch AE, Spadiut O, Herwig C (2012) Switching industrial production pro- Oil Chem Soc 58(3):159–166. https ://doi.org/10.1007/bf025 82328 cesses from complex to defined media: method development and case study using the example of Penicillium chrysogenum. Microb Cell Fact. https ://doi.org/10.1186/1475-2859-11-88

Journal

AMB ExpressSpringer Journals

Published: Jun 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off