TY - JOUR
AU - Delvigne, Frank
AB - Abstract The oleaginous yeast Yarrowia lipolytica has the ability to use oils and fats as carbon source, making it a promising cell factory for the design of alternative bioprocesses based on renewable substrates. However, such a multiphasic bioreactor design is rather complex and leads to several constraints when considering emulsification of the oil-in-water mixture, foaming and cell growth/physiology on hydrophobic substrate. This study aims to shed light on the effect of pH changes on the physico-chemical properties of the cultivation medium and on cell physiology. It was indeed observed that at a pH value of 6, cell growth rate and intracellular lipid accumulation were optimized. Additionally, foaming was significantly reduced. In order to avoid over foaming in bioreactor, without impairing cell physiology, the use of alternative processes that can only act on the physical structure of culture medium, seems to be an effective alternative to usual chemical anti-foam agents. bioreactor, population analysis, flow cytometry, foaming, emulsion, anti-foam effect INTRODUCTION The interest of the industry for the yeast Yarrowia lipolytica is justified by its great potential for bioconversion, aroma production, single cell oil production, intermediate-metabolite production, in bioremediation processes and in food technology, based on either recombinant, mutant or wild-type strains (Fickers et al.2005). Additionally, Y. lipolytica has been used as a model organism to study the metabolic pathways involved in hydrophobic substrates (HS) assimilation, given its ability to degrade HS such as alkanes, fats, oils and fatty acids (Fickers et al.2005; Timoumi et al.2018). However, during cultivation of this strictly aerobic yeast (Barth and Gaillardin 1997) under aerobic conditions, intensive aeration, high protein content in the cultivation medium (Delvigne and Lecomte 2010), as well as the metabolites/proteins produced by the microorganism itself (Winterburn and Martin 2012; Kar et al.2012a) can lead to excessive foam formation. Considering its multiple adverse effects, foam formation is a big challenge for process engineer, especially when intensive aeration is required. This phenomenon leads effectively to a decrease of the working volume of the reactor, removal of microbial cells and metabolites, increased risk of contamination, decrease of the oxygen transfer rate after the addition of chemical anti-foam (Kar et al.2012a). Thus, aerobic bioprocesses require finding a good compromise between oxygen transfer rate and foam formation (Kar et al.2012b). Moreover, in order to enhance assimilation of hydrophobic substrate by the yeast, the medium has to be converted into oil-in-water emulsion. Such approach requires careful optimization of the overall mixing applied to the cultures, which also may contribute to occurrence of excessive foam formation. In general, surface-active molecules found in cultivation media are proteins, which are the basis of most industrial complex media, or are released by cell autolysis during culture processes (Schügerl 2000; Winterburn and Martin 2012; Kar et al.2012a). Other compounds with surface tension can also be secreted by microorganisms, for example, Escherichia coli is known to release lipopolysaccharides when it is exposed to starvation stress during culture (Han, Enfors and Häggström 2003), and Bacillus subtilis is able to synthesize several secondary metabolites exhibiting surface active properties (Coutte et al.2017). A wide variety of surface-active compounds excreted during the cultivation of Penicillium herqueii has also been characterized (Delvigne and Lecomte 2010). Proteins are however the most frequently encountered surfactants during fermentation bioprocesses. They are adsorbed at the gas-liquid interfaces and, depending on their surfactant properties, they are capable of promoting the foamation at very low concentrations (even at 1 mg L−1) (Winterburn and Martin 2012). In addition, it has been shown that the sterilization procedure of a protein's solution significantly increases the foaming capacity of the medium (Schügerl 2000). To avoid foam formation in bioreactors, mechanical processes such as foam breakers and ultrasound are often used, as well as chemical methods with the addition of anti-foaming agents (Etoc et al.2006; Delvigne and Lecomte 2010; Routledge 2012). (Vardar-Sukan 1992) and (Junker 2007) considered positive and negative effects of anti-foaming agents described in several investigations. The authors confirmed that anti-foaming agents have potential effects on microbial metabolism, mass transfer, downstream operations and quality of the final product. Moreover, some are known to be toxic to the microorganisms, others can be metabolized by the microorganism and interfere in enzymes’ production bioprocess (Vardar-Sukan 1992; Junker 2007). It has also been reported that anti-foams enhance the bubbles’ coalescence and increase their size, leading to a reduction in the specific surface, hence lower the oxygen transfer coefficient KLa (Routledge 2012). Likewise, anti-foams seem to affect differently the cultures yields, cells growth and the recombinant protein production, depending on the type and the concentration used (Winterburn and Martin 2012). Anti-foam agent concentrations above 1% appear lead to a decrease of recombinant protein production, although some agents at higher concentrations enhance cell growth (Routledge 2012). Overall, anti-foams are often added in bioprocesses without real knowledge on their possible effects. Published studies showed that each anti-foam not only destroys the foam with a specific efficacy range, but can also affect the cells and the produced proteins themselves (Routledge 2012). In this context, the work reported in this paper considers cultivation of the non-conventional yeast Y. lipolytica on a hydrophobic substrate (oleic acid). Since Y. lipolytica is strictly aerobic and grows efficiently on oleic acid, it is critical to optimize operating conditions in order to keep dissolved oxygen (DO) at a sufficient level and to limit foam overflow. To this end, adjustment of pH of the culture medium was proposed a solution to overcome this problem. Indeed, pH can modify the quality of the emulsion and also impact foam formation (Benelhadj et al.2016). Additionally, modulating the quality of the emulsion can have also some impact at the level of the yeast physiology, i.e. by reducing for example the amount of liposan secreted by the yeast in order to stabilize the emulsion. In order to assess the efficiency of our approach in a comprehensive manner, flow cytometry was used for the simultaneous analysis of yeast cell population and lipid micro-droplets population. By this way, microbial growth and substrate consumption can be efficiently tracked, even in a complex biphasic culture. MATERIAL AND METHODS Strain and culturing media The yeast Y. lipolytica JMY 775 used in this study was obtained by genetic manipulation. It is a lipase overproducing strain (LgX64.81), modified by fusing LIP2 gene with LacZ reporter (Kar et al.2008). The yeast was pre-cultivated in YPD medium for 24 h in 250 mL baffled shaking flasks containing 100 mL of the medium at 30°C, with vertical agitation (185 rpm). The pre-culture medium was composed of 5 g L−1 yeast extract (Organotechnie, La Courneuve, France), 10 g L−1 peptone of casein (Organotechnie, La Courneuve, France) and 20 g L−1 of glucose as carbon source (Cargill, Mechelen, Belgium, dextrose mono hydrate, 98% purity). For the culture, 0.2 × 108 cells were added to 200 mL of a lipid accumulation medium YPDAcO (Bouchedja et al.2017). YPDAcO is a modified version of the YPD medium with the following composition: 5 g L−1 yeast extract, 5 g L−1 casein peptone, 10 g L−1 glucose and 25 g L−1 oleic acid (Sigma-Aldrich, Overijse, Belgium). Emulsion and sonication Effect of pH was tested based on the YPDAcO cultivation medium. For this purpose, medium was prepared by applying two consecutive sonication steps (intensity was set at 85% for 120 s for each cycle). Bioreactor and operating conditions The operating conditions in the reactor were similar to the optimum Y. lipolytica culture conditions described by (Bouchedja et al.2017). The cultures were conducted in stirred mini reactors (Dasgip, 4Unit Technology) of total volume of 260 mL, equipped with digital data control system and sensors for pH, oxygen pressure (PO2), DOand pressure (P). In this work, the investigations were carried out in three successive steps: In the first step we checked whether pH drop has any anti-foam effect. Thus, this step consisted to proceed to a gradual drop in pH every 30 min (pH = 6.8, pH = 6 and pH = 5) in the same reactor (T = 28°C, Stirring/Aeration = 600 rpm/84 mL min−1), which was filled with 150 mL of emulsion. A sample of 5 mL was taken every 15 min and after each pH drop in order to be analyzed with particle sizer and observed under optical microscope. The volume of foam produced in the reactor was also calculated by the formula: \begin{equation*} V\ = \ \pi \times \ \left( {\frac{D}{2}} \right)\ \times H \end{equation*} V: Volume of foam D: Reactor diameter (7 cm) H: Height of the foam In order to complete the observations revealed by the first step concerning the effect of pH value on the foam expansion, in the second step we checked whether stirring time has any impact on the foam neutralization. Experiments in this part were carried out in three mini bioreactors with a working volume of 200 mL (Dasgip 4Unit Technology), filled with 90 mL of emulsion, operating in parallel and stirred mechanically (T = 28°C, Stirring/Aeration = 600 rpm/84 mL min−1). The pH was differently adjusted in each reactor, and was set at respective values of pH = 5, pH = 6 and pH = 6.8. Samples of 5 mL were taken at regular intervals of the experiment from each reactor, namely T1 = 15 min, T2 = 45 min and T3 = 90 min. Samples were analyzed with particle sizer and optical microscope. The volume of produced foam within each reactor was also calculated by the formula described above. In the third step we checked the effect of the studied pH values on the biotic phase. Thus, Y. lipolytica cultures were carried out in the emulsified medium YPDAcO. Cells cultures were carried out for 30 h, in three bioreactors operating in parallel under identical temperature and aeration conditions (T = 28°C, Stirring/Aeration = 600 rpm/84 mL min−1). The pH values were adjusted differently in each bioreactor, namely pH = 5, pH = 6 and pH = 6.8. Sampling was performed at regular time intervals of culture i.e.0, 6, 12 and 30 h. Collected samples were analyzed by flow cytometer. All analyses, measurements and microscopic observations were in triplicates. Granulometry The droplet size distribution of the emulsion was determined with a Malvern Mastersizer 2000 particle size analyzer coupled to a Hydro 2000s device (Malvern Instruments, Malvern, UK). This system allows the characterization of the dispersion by the characteristics diameters d0.1, d0.5 (median particle diameter), d0.9 and d32 (Sauter diameter which corresponds to the diameter of a sphere having the same volume/surface ration as the particle). The d0.1, d0.5 and d0.9 values are size values that correspond to the cumulative distribution at 10%, 50% and 90%, respectively. A sufficient amount of emulsion to reach an obscuration of about 15% was dispersed under moderate stirring in the measuring vessel of the instrument containing distilled water. Measurements were performed at room temperature on the freshly made emulsions. At least three replications were made for each set of conditions. Foam scan Foaming properties of the culture medium set at different pH were studied by comparing their foam formation and stability, measured at room temperature, using a Foamscan analyzer (Foamscan IT Concept, Longessaigne, France). In this device, foam formation and foam stability are determined by conductimetric and optical measurements (Fains et al.1997). The foam was generated by blowing air at a flow rate of 100 mL/min through a porous sintered (P3) at the bottom of a glass column (diameter 40 mm); 25 mL of medium was used for each experiment. In all experiments, the foam was allowed to form during 120 min of air bubbling. Afterwards the bubbling was stopped. The evolution of the foam was monitored to assess foam stability; its height was continuously measured by image analysis. The half-life time referring to the time needed to lose half of the volume of the foam was adopted as a criterion to describe the foam stability (Karamoko et al.2013). All measurements were carried out in triplicate. Flow cytometry: simultaneous detection of microbial cells and oil micro-droplets Flow cytometry constitute an interesting alternative strategy to the classical single-cell approach generally relying on the use of biosensors (Lemoine et al.2017). The samples taken from the bioreactors were regularly submitted to cytometric analyzes (Accuri C6 flow cytometer). In order to observe the cell growth dynamics, and lipid assimilation, operating within each reactor. The samples were systematically diluted with distilled water to obtain 40 000 events per sample thus avoid overlap between event and signal distortion induced by high sample concentrations. Overall, flow cytometer measure quantitatively the optical characteristics of cells or particles (also called events) presented in a single file in front of a focused light beam. As particles filter through this latter, three parameters were measured using photomultiplier tubes. They are forward scatter (FSC), side scatter (SSC) and fluorescence (FL). The cell sizes were inversely proportional to the amount of light scattered forward and at right angles. Furthermore, the cell refractibility is related to surface properties and internal structure affecting FSC and SSC. More details describing the on-line flow cytometry process are explained in details in our previous paper (Bouchedja et al.2017). Optical microscopy The microstructure of all the emulsions was observed using microscopy (20× magnification). One droplet of emulsions was placed on a glass slide and covered with a glass cover slip. Samples were imaged at room temperature on an optical microscope (Nikon Eclipse E400 which included a camera type DS-FI2, Kanagawa, Japan). All the images were recorded with the gain switch in the auto position, and digitized using the operating software NIS-D Nikon. Statistical analysis The results of three measurements were expressed as mean ± SD. Differences in foam volumes within bioreactors were tested for significant difference using ANOVA analysis and LSD ad hoc test (P < 0.005). RESULTS Simultaneous monitoring of biotic and abiotic populations based on flow cytometry The purpose of this first section is to show that FC can be effectively used in order to make the distinction between oil micro-droplets and the yeast cells according to an efficient gating procedure. The FC analyses are carried out on the basis of the FSC and SSC without any staining protocol. In order to assess the gating procedure, a series of standard samples have been analyzed (Fig. 1). The first standard is a yeast culture carried out on glucose (YPD medium) without any hydrophobic compounds (Fig. 1A). The FSC/SSC dot plot shows a single population in this case corresponding to yeast cells. The second standard is a yeast culture carried out on YPD medium containing 25mL L−1 of oleate (Fig. 1C). In this case two sub-populations can be observed, the first one corresponding to oil micro-droplets and the second one corresponding to yeast cells. The location of the oil micro-droplets on the FSC/SSC dot plot was further confirmed by analyzing an oleate emulsion stabilized with tween 80 (Fig. 1E). Oil-in-water emulsion shows a very typical, S-shaped, light-scattering diagram, allowing the characterization of the quality of the emulsion, i.e. monodisperse or polydisperse (Fattaccioli et al.2009). It seems thus that on the basis of the standard, an effective gating can be performed without any interference between the biotic and abiotic phases. However, it can be seen that the biotic phase (yeast cells) is subjected to the modification according to the two standard samples analyzed. Indeed, yeast cells cultivated with or without the presence of hydrophobic substrate display different FSC and SSC parameters. This phenomenon can be attributed to the fact that during the consumption of oleate, oil micro-droplets are first adsorbed onto the membrane and then internalized and eventually stored as lipid bodies inside the cells (Fickers et al.2005; Díaz et al.2010). Internalized structures, such as lipid bodies, are known to increase the SSC parameter (Díaz et al.2010; Bouchedja et al.2017). Additionally, yeast cells cultivated on hydrophobic substrate display more complex cell shape (Fig. 1D), contributing to the increase of the SSC parameter as well as the FSC parameter. Yeast growth on hydrophobic substrate is thus a very complex system according to the evolution of the biotic phase. Thus dynamic analytical tools are needed to study this system reliably. The evolution of the cytometric parameters of the biotic phase will then be analyzed more in details in the section 3.3. Figure 1. View largeDownload slide A, FC analysis of Y. lipolityca cultivated in shake flask on YPD medium (B) microscopic observation (40×) of Y. lipolytica cultivated in shake flask on YPD medium showing the prevalence of yeast shaped cells. C, FC analysis of Y. lipolityca cultivated in shake flask on YPD/oleate medium (D) microscopic observation (40×) of Y. lipolytica cultivated in shake flask on YPD/oleate medium. In this case more complex cell shapes due to micro-droplets attached at the cells surface or internalized as lipid bodies can be observed (E) FC analyses of oleate dispersion obtained by sonication in presence of tween 80 used as a stabilizing agent. FC analyses have been carried out on the basis of 40 000 events displayed in dot plot showing the FSC in function of the SSC. F, Microscopic observation of oleate emulsion. The gate P2 corresponds to the yeast cells. Figure 1. View largeDownload slide A, FC analysis of Y. lipolityca cultivated in shake flask on YPD medium (B) microscopic observation (40×) of Y. lipolytica cultivated in shake flask on YPD medium showing the prevalence of yeast shaped cells. C, FC analysis of Y. lipolityca cultivated in shake flask on YPD/oleate medium (D) microscopic observation (40×) of Y. lipolytica cultivated in shake flask on YPD/oleate medium. In this case more complex cell shapes due to micro-droplets attached at the cells surface or internalized as lipid bodies can be observed (E) FC analyses of oleate dispersion obtained by sonication in presence of tween 80 used as a stabilizing agent. FC analyses have been carried out on the basis of 40 000 events displayed in dot plot showing the FSC in function of the SSC. F, Microscopic observation of oleate emulsion. The gate P2 corresponds to the yeast cells. Impact of pH on the oil micro-droplets population and foaming properties In order to determine the origin of the foaming phenomenon (linked to the medium itself or to the cell), a part of experiments were conducted in the absence of the biotic phase. The foam capacity and stability of the sterilized culture media without the biotic phase were first investigated as a function of the adjusted pH. As Y. lipolytica is able to grow both in neutral and acidic conditions, pH adjustment was thought as a solution to avoid foaming. Adjustments of pH have been made in the range 7.0 to 5.0. Results are presented in Table 1. It is clear from those results that the foam production is linked to the culture medium itself and that this phenomenon is highly impacted by the pH. A reduction of pH allowed a reduction both of the foam capacity and stability. Adjustment of pH to 6.0 allowed the avoidance of foam formation. The foaming phenomenon is due to the proteins from the culture medium that act as surfactants owing to their amphiphilic structures. Table 1. Foam capacity and stability of the sterilized culture media analyzed with foam scan in absence of the biotic. Adjusted pH Final foam volume (after 120sec) Time to reach 50mL Time to reach 80mL Half-life time 7.0 190 + $$-$$1 mL 18 + $$-$$0 sec 45.5 + $$-$$0.5 sec >700 sec 6.8 175 + $$-$$2 mL 20 + $$-$$0.5 sec 56 + $$-$$0.5 sec 175 sec 6.5 172 + $$-$$5 mL 21.5 + $$-$$0.5 sec 60 + $$-$$2 sec 160 sec 6.0 5 + $$-$$1 mL – – – 5.0 0 mL – – – Adjusted pH Final foam volume (after 120sec) Time to reach 50mL Time to reach 80mL Half-life time 7.0 190 + $$-$$1 mL 18 + $$-$$0 sec 45.5 + $$-$$0.5 sec >700 sec 6.8 175 + $$-$$2 mL 20 + $$-$$0.5 sec 56 + $$-$$0.5 sec 175 sec 6.5 172 + $$-$$5 mL 21.5 + $$-$$0.5 sec 60 + $$-$$2 sec 160 sec 6.0 5 + $$-$$1 mL – – – 5.0 0 mL – – – View Large Table 1. Foam capacity and stability of the sterilized culture media analyzed with foam scan in absence of the biotic. Adjusted pH Final foam volume (after 120sec) Time to reach 50mL Time to reach 80mL Half-life time 7.0 190 + $$-$$1 mL 18 + $$-$$0 sec 45.5 + $$-$$0.5 sec >700 sec 6.8 175 + $$-$$2 mL 20 + $$-$$0.5 sec 56 + $$-$$0.5 sec 175 sec 6.5 172 + $$-$$5 mL 21.5 + $$-$$0.5 sec 60 + $$-$$2 sec 160 sec 6.0 5 + $$-$$1 mL – – – 5.0 0 mL – – – Adjusted pH Final foam volume (after 120sec) Time to reach 50mL Time to reach 80mL Half-life time 7.0 190 + $$-$$1 mL 18 + $$-$$0 sec 45.5 + $$-$$0.5 sec >700 sec 6.8 175 + $$-$$2 mL 20 + $$-$$0.5 sec 56 + $$-$$0.5 sec 175 sec 6.5 172 + $$-$$5 mL 21.5 + $$-$$0.5 sec 60 + $$-$$2 sec 160 sec 6.0 5 + $$-$$1 mL – – – 5.0 0 mL – – – View Large In order to assess the impact of oleate droplets dispersion on bioreactor performances, a complementary series of tests were carried out in stirred vessel without biotic phase. During these experiments, it was observed that when pH was adjusted to 6.8, emulsion started to foam as soon as stirring was initiated. After 15 min, the foam occupied a volume of 154 mL (±22 mL) and began to overflow. After 3 min of pH reduction to a value of 6, a significant reduction in the foam volume was observed (38% relative to its initial volume, with ΔV (pH6.8-pH6) = 5.8 mL (P < 0.05). Indeed, at pH = 6, the foam volume in reactor, was reduced to a value of 96 mL (± 1.9 mL). At pH = 5, the amount of foam was further significantly reduced, with ΔV (pH6-pH5) = 79 mL, and reached a value of 17 mL (±4 mL). Granulometric analysis showed that the size of the droplets decreased proportionally with pH decreasing, so that at pH = 5, droplet size was completely homogeneous (Fig. 2A and B). The analyses of emulsion droplets dispersions (Fig. 2) showed that, at pH = 6.8, droplets size distribution was highly heterogeneous and was characterized by the presence of small droplets with diameter sizes in the range of [0.5–7.5] μm (median diameter of 2.5 μm), and the presence of large droplets with diameter sizes between [10–100] μm (with a median diameter of about 40 μm), and that represents 50% of the total droplet volume. At pH = 6 (Fig. 2), the same droplet size heterogeneity was observed, however the size of the large droplets was smaller, with a median value of 20 μm. In this case, large droplets occupied only 35% of the total volume, in contrast to the smaller ones, which were occupying a volume of 50%. This clearly explains lesser foam expansion at this value of pH (pH = 6). At pH = 5, droplets were homogeneous and the smallest, with an average size of 2.5 μm, occupying 80% of total volume (Fig. 2). Figure 2. View largeDownload slide A, Granulometry of the studied emulsion under regulated temperature and aeration conditions. (T = 28°C, Stirring/Aeration = 600 rpm/84 mL min−1) and different pH conditions (pH = 6.8, pH = 6 and pH = 5). The pH was dropped within the same reactor, in the absence of the biotic phase (B) optical microscopy (G20×) corresponding to the samples displayed in the panel A. Figure 2. View largeDownload slide A, Granulometry of the studied emulsion under regulated temperature and aeration conditions. (T = 28°C, Stirring/Aeration = 600 rpm/84 mL min−1) and different pH conditions (pH = 6.8, pH = 6 and pH = 5). The pH was dropped within the same reactor, in the absence of the biotic phase (B) optical microscopy (G20×) corresponding to the samples displayed in the panel A. The same results were obtained by comparing the expansion of the emulsion, tested in three reactors operating in parallel under different pH conditions, i.e. with pH values of 5, 6 and 6.8 respectively (Figs. 3 and 4). Therefore, results depicted in Fig. 4 confirmed that decreasing pH leads to a significant foam reduction effect, which is totally independent of the stirring intensity. Indeed, Fig. 4 showed that, at identical pH levels, droplets size curves followed the same trends in function of time, regardless of the duration of agitation. Figure 3. View largeDownload slide Granulometry of the emulsion sampled from three identical reactors operating in parallel under regulated conditions of temperature (T = 28°C) and Stirring/Aeration = 600rpm/84mL.min−1. The pH was differently adjusted in each reactor (pH = 6.8, pH = 6 and pH = 5). Figure 3. View largeDownload slide Granulometry of the emulsion sampled from three identical reactors operating in parallel under regulated conditions of temperature (T = 28°C) and Stirring/Aeration = 600rpm/84mL.min−1. The pH was differently adjusted in each reactor (pH = 6.8, pH = 6 and pH = 5). Figure 4. View largeDownload slide Optical microscopy (G20×) corresponding to the samples described in the Fig. 2, showing the bubbles’ sizes of emulsion sampled from each reactor operating in parallel under regulated conditions of temperature (T = 28°C) and Stirring/Aeration = 600 rpm/84 mL min−1. The pH was differently adjusted in each reactor (pH = 6.8, pH = 6 and pH = 5). Figure 4. View largeDownload slide Optical microscopy (G20×) corresponding to the samples described in the Fig. 2, showing the bubbles’ sizes of emulsion sampled from each reactor operating in parallel under regulated conditions of temperature (T = 28°C) and Stirring/Aeration = 600 rpm/84 mL min−1. The pH was differently adjusted in each reactor (pH = 6.8, pH = 6 and pH = 5). Impact of pH on Y. lipolytica population The pH value in the culture medium is an important factor controlling several cellular functions and it is thus important to study the effect of this parameter not only on the micro-droplets population, but also on the microbial cell population. At pH = 6.8, excessive foaming within the bioreactor caused overflow of cultivation medium. Thus, at this pH value it was impossible to carry out any investigations on microbial population. However, investigations at pH = 5 and pH = 6 were possible. The study was carried out in two bioreactors operating in parallel under the same conditions of temperature (T = 30°C), stirring speed (600 min−1) and aeration (84 mL min−1). The objective was to study the effect of the tested pH on microbial growth and intracellular lipid accumulation. Indeed, as it was stated before (Fig. 1), flow cytometry enabled to follow the relative proportion of oil micro-droplets and cell population in bioreactor. Then, following the evolution of the respective sub-populations from time to time allows determining the assimilation of the hydrophobic substrate and microbial growth. The cytometric results showed that at pH = 6, cell growth (i.e. appearance of events in P2 gate) and lipid assimilation (i.e. disappearance of events in gate P1) was faster by comparison with cultivation operated at pH = 5 (Fig. 5). Figure 5. View largeDownload slide Evolution of the oil micro-droplets (gate P1) and yeast cell population (gates P2) during a culture of Y. lipolytica JMY 775 carried out in two stirred mini-bioreactors under controlled temperature (28°C) and stirring/aeration conditions = 600 rpm/84 mL min−1. The pH was adjusted at different level in each mini bioreactor (pH = 6 and pH = 5). Analyses were carried out with online flow cytometry and results displayed on FSC/SSCdot plot on the basis of 40 000 events. The gate P1 corresponds to the oil droplets and the gate P2 corresponds to the yeast cells. Figure 5. View largeDownload slide Evolution of the oil micro-droplets (gate P1) and yeast cell population (gates P2) during a culture of Y. lipolytica JMY 775 carried out in two stirred mini-bioreactors under controlled temperature (28°C) and stirring/aeration conditions = 600 rpm/84 mL min−1. The pH was adjusted at different level in each mini bioreactor (pH = 6 and pH = 5). Analyses were carried out with online flow cytometry and results displayed on FSC/SSCdot plot on the basis of 40 000 events. The gate P1 corresponds to the oil droplets and the gate P2 corresponds to the yeast cells. DISCUSSION An emulsion is a colloidal dispersion of liquid droplets in a liquid continuous phase in which they are immiscible. From a quantitative point of view, it is characterized by the volume fraction of the dispersed phase (Φ) and the size distribution of the droplets. Emulsion thus consists of two separate phases in a pseudo-homogeneous system characterized by a large interfacial area (Mcclements 2007). It is almost never a spontaneous process, and requires mechanical action. Surface tension σ is the work required to increase the area of an interface by one unit. From the thermodynamic point of view, (σ) corresponds to the free energy necessary to carry out this operation. As for the expansion, it is the dispersion of a gas phase that is formed in situ in the form of bubbles in a liquid phase or in pasty phase. An expanded product can therefore be defined, such as a dispersion of a gas in a continuous medium. When it comes to an emulsion, it is about an expanded emulsion. This study deals with an expanded emulsion that from a physico-chemical point of view is a multi-phasic medium with three phases, two of which are dispersed in a continuous medium. The first phase is oil dispersed in water-based culture medium in a form of the emulsion (oil in water: O/W) comprising lipid droplets with a maximum size in the order of 1 μm; the second is the gas phase (air) dispersed as millimetre-sized bubbles. So, expanded emulsions are very heterogeneous structures at the microscopic scale that belong to the colloid family, since the emulsion is of a colloidal nature. The dispersed gas phase is generally assigned to a colloidal system, although the average size of the bubbles (10 to 100 μm) exceeds the size limit that is attributed to these structures (approximately 1 μm) and that this qualifier must first be reserved for dry foams (Mcclements 2007). Indeed, the expansion of emulsions forms essentially wet foams, in which bubbles are spherical and the interfaces gas-liquid are non-contiguous. However, the importance of the adsorbed layers (proteins, surfactants, etc.) in the foamed emulsions and their interactions with the continuous phase results in a physical behaviour dominated by surface effects and colloidal interactions (Mcclements 2007). Therefore, expanded emulsions are multiple colloids, and their stability increases when the bubbles size decreases, when protein concentration increases and when the expansion rate decreases. It should be known that the stability of the gas bubbles in an expanded emulsion is much lower than that of the dispersed fatty phase for the following reasons: The difference in densities of continuous and dispersed phases is higher; The viscosity ratio between the continuous phase and the dispersed phase is greater; The surface tension at the water/air (W/A) interfaces is much higher than the interfacial tension of an oil/air (O/A) interface. The bubble sizes are therefore much larger than those of the fat globules of the emulsion; The solubility of the gas in the water is higher. Proteins containing in the culture medium have two essential functional properties in the production of emulsions and foams (foaming emulsions), which are their emulsifying and foaming power resulting from their ability to be adsorbed in the water in oil interfaces (W/O), and their ability to stabilize these interfaces. That's why they are used for their surfactant properties (Lazidis et al.2015). The proteins are amphiphilic polymers which, in addition to being formed of hydrophilic or hydrophobic amino acids, are capable to form between the ammonium and carboxylate groups, inter and intra-molecular bonds, as ionic bonds (Martin et al.2002). So, proteins are polyelectrolytes, and their charges depend strongly on their degree of protonation, thus they depend on the pH of the medium. In addition, caseins (our culture medium contain 5g L−1 casein peptones) are quickly adsorbed proteins (Martin et al.2002)), so to explain the mechanisms of emulsions stabilization by proteins, Dickinson (2001) arguments that at a pH far from their isoelectric pH (pI), proteins are electrically charged, once adsorbed on the surface of the fat globules, they cause electrostatic stabilization. As well, by their size, the proteins hinder the approximation of the fat globules and contribute to their stabilization by steric hindrance. And finally, the adsorption of proteins increases the density of fat globules that reduces the difference in density between the fat and the aqueous phase. During expansion, it has been shown that the adsorption of fat globules at the interfaces takes place only in a second stage and is of non-competitive type. The air bubbles are first stabilized by the adsorption of the proteins at the water/air interface, the adsorption of the fat globules becomes significant only later. This difference in the kinetics of adsorption of proteins and fat globules comes essentially from their difference in size (1–10 μm for fat globules and 50–300 nm for casein micelles). Thus, proteins have an essential role in stabilizing the dispersed phase. In fact, during the production of an expanded emulsion, they form a coherent adsorbed film on the bubble surface. pH and ionic strength are other parameters that govern the stability of emulsions, in addition to protein concentration, (Dickinson 2001; Zhang, Dalgleish and Goff 2004). Although it has been confirmed that increases in ionic strength and protein content have generally a positive effect (Kinsella 1984; Luck, Bray and Foegeding 2002), pH effect appears to be more complex. There are two antagonistic effects that may explain the role of pH on protein foams. At the isoelectric point, the protein-protein affinity is maximal because the electrostatic repulsion is minimal, as well as the compactness, the viscosity and the stability of the adsorbed layers, which should favour the expansion (Carp et al.2001; Davis, Foegeding and Hansen 2004; Zhang, Dalgleish and Goff 2004). However, this effect is counterbalanced by the fact that at the isoelectric point the proteins solubility is minimal, which reduces the amount of protein stabilizing bubbles (Patino, Delgado and Fernández 1995). Changes in the conformation of globular proteins as function of pH are also possible, which may positively or negatively impact their foaming power (Zhang, Dalgleish and Goff 2004). It has been confirmed that foaming expansion of a protein solution is maximum near the isoelectric point for a serum protein isolate (pI ≈ 5), but it is minimal at pI for skim milk powder (pI ≈ 4,6) because of the low solubility of caseins at this pH. For serum protein isolate, the maximum is shifted to 4.5 by a conformational change of α-lactoglobulin (pI = 4.1–4.8) that increases its foaming power, when the pH is lower at 5 (Zhang, Dalgleish and Goff 2004). The work of (Allen, Dickinson and Murray 2006) on emulsions containing sodium caseinate, showed better stability of the gas phase, resulting in less expansion, when the pH is close to the pH of the protein. Namely that, casein pI is close to 4.6. Therefore the change in pH affects strongly the microstructure of the emulsions, so their abundant properties (Thakur, Vial and Djelveh 2006). Concerning the pH effect on cells growth and lipid cells assimilation in the biotic phase, information available in the literature indicates that a low pH variation in the culture medium influences the lipid composition rather than the accumulated quantity (Barth and Gaillardin 1997), and that the response to pH variations depends of each yeast species. However, Y. lipolytica is able to grow at low pH, around 4, its growth is slowed down to pH 7 and inhibited at pH above 8 (Barth and Gaillardin 1997; Moeller et al.2007). Another explanation about the impact of pH on foaming properties of the broth is linked with the physiology of Y. lipolytica itself. Indeed, the total proteome of Y. lipolytica exhibits an average isoelectric point of 6.34 (analysis performed based on the Proteome-pI) (Kozlowski 2016), potentially explaining the reduction in foaming intensity observed when pH was decreased to a value of 6. Online flow cytometry process seems to be an efficient tool for monitoring a bioreactor culture of Y. lipolytica. So, this emerging and versatile technique, not only for the study of the involved parameters cell culture evolution, but also into study of the various phenomena induced by modifications in each culture condition i.e cells growth, cells lipid assimilation and cells dimorphism. Thus, it offer a considerable time saving for obtaining and processing results in comparison with the conventional techniques (Bouchedja et al.2017). CONCLUSION The experiments conducted in this study have highlighted the complex interplay between biotic and abiotic factors in complex, multiphasic bioreactor. More precisely, we have shown that pH had a strong impact on the quality of dispersion of oil micro-droplets population, which in turn affects its assimilation by the yeast population. From an applied perspective, this study has shown that the use of alternative methods based on the modification of the physical structure of the oil micro-droplet population by modulating the pH of culture medium, without modifying its chemical composition and its performances, significantly limits the undesirable phenomenon of foaming within the bioreactor, without defecting the quality of the culture. This alternative appears to be an effective strategy to common chemical agents such as emulsifiers and anti-foams. Acknowledgements Authors deeply thank Pr. Nouredine KACEM-CHAOUCHE, Dr. Anis CHIKHOUNE, Dr. Mounia YOUCEF-ALI and Mr Samuel TELEK for their valuable help and assistance. FUNDING This work was financially supported by University of Liège/Gembloux Agro-Bio Tech, and also by the support given to DNB from the Algerian Ministry of high education and research and University Frères Mentouri-Constantine1. Conflict of interest. None declared. REFERENCES Allen KE , Dickinson E , Murray B . Acidified sodium caseinate emulsion foams containing liquid fat: a comparison with whipped cream . LWT - Food Sci Technol 2006 ; 39 : 225 – 34 . Google Scholar CrossRef Search ADS Barth G , Gaillardin C . Physiology and genetics of the dimorphic fungus Yarrowia lipolytica . FEMS Microbiol Rev 1997 ; 19 : 219 – 37 . Google Scholar CrossRef Search ADS PubMed Benelhadj S , Gharsallaoui A , Degraeve P et al. Effect of pH on the functional properties of Arthrospira (Spirulina) platensis protein isolate . Food Chem 2016 ; 194 : 1056 – 63 . Google Scholar CrossRef Search ADS PubMed Bouchedja DN , Danthine S , Kar T et al. Online flow cytometry, an interesting investigation process for monitoring lipid accumulation, dimorphism, and cells’ growth in the oleaginous yeast Yarrowia lipolytica JMY 775 . Bioresour Bioprocess 2017 ; 4 : 3 . Google Scholar CrossRef Search ADS PubMed Carp DJ , Bartholomai GB , Relkin P et al. Effects of denaturation on soy protein–xanthan interactions: comparison of a whipping–rheological and a bubbling method . Colloids Surf B 2001 ; 21 : 163 – 71 . Google Scholar CrossRef Search ADS Coutte F , Lecouturier D , Dimitrov K et al. Microbial lipopeptide production and purification bioprocesses, current progress and future challenges . Biotechnol J 2017 ; 12 . Davis JP , Foegeding EA , Hansen FK . Electrostatic effects on the yield stress of whey protein isolate foams . Colloids Surf B 2004 ; 34 : 13 – 23 . Google Scholar CrossRef Search ADS Delvigne F , Lecomte JP . Encyclopedia on Industrial Biotechnology: Bioprocess, Bioseparation and Cell Technologies. Foam Formation and Control in Bioreactors . New York, NY : Wiley , 2010 . Díaz M , Herrero M , García LA et al. Application of flow cytometry to industrial microbial bioprocesses . Biochem Eng J 2010 ; 48 : 385 – 407 . Google Scholar CrossRef Search ADS Dickinson E . Milk protein interfacial layers and the relationship to emulsion stability and rheology . Colloids Surf B 2001 ; 20 : 197 – 210 . Google Scholar CrossRef Search ADS Etoc A , Delvigne F , Lecomte J-P et al. Foam control in fermentation bioprocess . Twenty-Seventh Symposium on Biotechnology for Fuels and Chemicals . Berlin, Heidelberg : Springer , 2006 , 392 – 404 . Fains A , Bertrand D , Baniel A et al. Stability and texture of protein foams: a study by video image analysis . Food Hydrocolloid 1997 ; 11 : 63 – 9 . Google Scholar CrossRef Search ADS Fattaccioli J , Baudry J , Émerard J-D et al. Size and fluorescence measurements of individual droplets by flow cytometry . Soft Matter 2009 ; 5 : 2232 – 8 . Google Scholar CrossRef Search ADS Fickers P , Benetti P-H , Wache Y et al. Hydrophobic substrate utilisation by the yeast, and its potential applications . FEMS Yeast Res 2005 ; 5 : 527 – 43 . Google Scholar CrossRef Search ADS PubMed Han L , Enfors S-O , Häggström L . Escherichia coli high-cell-density culture: carbon mass balances and release of outer membrane components . Bioprocess Biosyst Eng 2003 ; 25 : 205 – 12 . Google Scholar CrossRef Search ADS PubMed Junker B . Foam and its mitigation in fermentation systems . Biotechnol Progress 2007 ; 23 : 767 – 84 . Google Scholar CrossRef Search ADS Kar T , Delvigne F , Masson M et al. Investigation of the effect of different extracellular factors on the lipase production by Yarrowia lipolityca on the basis of a scale-down approach . J Ind Microbiol Biotechnol 2008 ; 35 : 1053 – 9 . Google Scholar CrossRef Search ADS PubMed Kar T , Destain J , Thonart P et al. Physical and physiological impacts of different foam control strategies during a process involving hydrophobic substrate for the lipase production by Yarrowia lipolytica . Bioprocess Biosyst Eng 2012 ; 35 : 483 – 92 . Google Scholar CrossRef Search ADS PubMed Kar T , Destain J , Thonart P et al. Scale-down assessment of the sensitivity of Yarrowia lipolytica to oxygen transfer and foam management in bioreactors: investigation of the underlying physiological mechanisms . J Ind Microbiol Biotechnol 2012 ; 39 : 337 – 46 . Google Scholar CrossRef Search ADS PubMed Karamoko G , Danthine S , Olive G et al. Interfacial and foaming properties of two types of total proteose-peptone fractions . Food Bioprocess Technol 2013 ; 6 : 1944 – 52 . Google Scholar CrossRef Search ADS Kinsella JE . Milk proteins: physicochemical and functional properties . Crit Rev Food SciNutr 1984 ; 21 : 197 – 262 . Google Scholar CrossRef Search ADS Kozlowski LP . Proteome-pI: proteome isoelectric point database . Nucleic Acids Res 2017 ; 45 : D1112 – 6 . Google Scholar CrossRef Search ADS PubMed Lazidis A , Hancocks RD , Spyropoulos F et al. Whey protein fluid gels for the stabilisation of foams . Food Hydrocolloid 2015 ; 30 : 1e9 . Lemoine A , Delvigne F , Bockisch A et al. Tools for the determination of population heterogeneity caused by inhomogeneous cultivation conditions . J Biotechnol 2017 ; 251 : 84 – 93 . Google Scholar CrossRef Search ADS PubMed Luck PJ , Bray N , Foegeding EA . Factors determining yield stress and overrun of whey protein foams . J Food Sci 2002 ; 67 : 1677 – 81 . Google Scholar CrossRef Search ADS Martin AH , Grolle K , Bos MA et al. Network forming properties of various proteins adsorbed at the air/water interface in relation to foam stability . J Colloid Interface Sci 2002 ; 254 : 175 – 83 . Google Scholar CrossRef Search ADS PubMed Mcclements DJ . Critical review of techniques and methodologies for characterization of emulsion stability . Crit Rev Food Sci Nutr 2007 ; 47 : 611 – 49 . Google Scholar CrossRef Search ADS PubMed Moeller L , Strehlitz B , Aurich A et al. Optimization of citric acid production from glucose byYarrowia lipolytica . Eng Life Sci 2007 ; 7 : 504 – 11 . Google Scholar CrossRef Search ADS Patino JMR , Delgado MDN , Fernández JL . Stability and mechanical strength of aqueous foams containing food proteins . Colloids Surf A 1995 ; 99 : 65 – 78 . Google Scholar CrossRef Search ADS Routledge SJ . Beyond de-foaming: the effects of antifoams on bioprocess productivity . Comput Struct Biotechnol J 2012 ; 3 : e201210001 . Google Scholar CrossRef Search ADS PubMed Schügerl K . Recovery of proteins and microorganisms from cultivation media by foam flotation . New Products and New Areas of Bioprocess Engineering . Berlin, Heidelberg : Springer , 2000 ; 191 – 233 . Google Scholar CrossRef Search ADS Thakur RK , Vial C , Djelveh G . Effect of pH of food emulsions on their continuous foaming using a mechanically agitated column . Innov Food Sci Emerg 2006 ; 7 : 203 – 10 . Google Scholar CrossRef Search ADS Timoumi A , Guillouet SE , Molina-Jouve C et al. Impacts of environmental conditions on product formation and morphology of Yarrowia lipolytica . Appl Microbiol Biotechnol 2018 ; 102 : 3831 – 48 . Google Scholar CrossRef Search ADS PubMed Vardar-Sukan F . Foaming and its control in bioprocesses . Recent Advances in Biotechnology . Berlin, Heidelberg : Springer , 1992 , 113 – 46 . Google Scholar CrossRef Search ADS Winterburn JB , Martin PJ . Foam mitigation and exploitation in biosurfactant production . Biotechnol Lett 2012 ; 34 : 187 – 95 . Google Scholar CrossRef Search ADS PubMed Zhang Z , Dalgleish DG , Goff HD . Effect of pH and ionic strength on competitive protein adsorption to air/water interfaces in aqueous foams made with mixed milk proteins . Colloids Surf B 2004 ; 34 : 113 – 21 . Google Scholar CrossRef Search ADS © FEMS 2018. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
TI - pH level has a strong impact on population dynamics of the yeast Yarrowia lipolytica and oil micro-droplets in multiphasic bioreactor
JF - FEMS Microbiology Letters
DO - 10.1093/femsle/fny173
DA - 2018-07-04
UR - https://www.deepdyve.com/lp/oxford-university-press/ph-level-has-a-strong-impact-on-population-dynamics-of-the-yeast-zAbIydaHsl
SP - 1
VL - Advance Article
IS - 16
DP - DeepDyve
ER -