Aerobic and anaerobic ethanol production by Mucor circinelloides during submerged growth

Abstract

The dimorphic organism Mucor circinelloides is currently being investigated as a potential host for heterologous protein production. The production of ethanol on pentose and hexose sugars was studied in submerged batch cultivations to further the general knowledge of Mucor physiology, with a view to the minimisation or elimination of the by-product ethanol for future process design. Large amounts of ethanol were produced during aerobic growth on glucose under non-oxygen limiting conditions, which is indicative of M. circinelloides being a Crabtree-positive organism. Ethanol production on galactose or xylose was less significant. The response of the organism to increased ethanol concentrations, both as the sole carbon source and in the presence of a sugar, was investigated in terms of biomass formation and morphology.

Introduction

Several species of the zygomycete Mucor display dimorphic behaviour (Orlowski 1991), i.e. they are capable of yeast-like or mycelial growth. For most dimorphic Mucor species, the composition of the gas atmosphere constitutes the key factor in determining the morphology the organism will assume (Bartnicki-Garcia and Nickerson 1962a). Filamentous growth prevails under aerobic conditions (Phillips and Borgia 1985), while strict anaerobiosis, often accompanied by an additional need for 30% CO₂, is a requisite for yeast-like growth (Bartnicki-Garcia and Nickerson 1962a). Furthermore, nutritional conditions are important morphological determinants, as yeast growth will only occur on a fermentable hexose (Sypherd et al. 1978). The filamentous form, on the other hand, can utilise a wide range of monosaccharides (including pentoses) and disaccharides as well as complex carbon sources such as cellulose or starch (Bartnicki-Garcia and Nickerson 1962b; McIntyre et al. 2002; Orlowski 1991). The proteolytic and saccharolytic activities of some Mucor species have found industrial applications. One of the best-known examples is the milk-clotting enzyme Mucor rennin, an aspartic proteinase, widely applied in the cheese-making industries (Arima et al. 1967). The use of Mucor circinelloides in the production of γ-linolenic acid, a fatty acid important for human nutrition, is also under investigation (Serrano et al. 2001; Wynn et al. 1999, 2001). More recently, M. circinelloides was suggested as a potential host for heterologous protein production (Wolff and Arnau 2002). This approach aims at exploiting the natural secretion capacity of the filamentous form, taking advantage of the capacity for yeast-like, single cell growth until the desired level of biomass is achieved.

Yeast-like growth of Mucor has been linked with a fermentative metabolism under all known growth conditions. This includes anaerobic cultivations as well as aerobic ones where yeast-like growth has been triggered by the morphogens phenethyl alcohol or dibutyrylic cAMP (reviewed by Orloswki 1991). The mycelial form can grow either oxidatively or fermentatively, depending on the cultivation conditions. Anaerobic growth, also in the mycelial form, has typically been correlated with high levels of ethanol production, while only small amounts of ethanol were produced by aerobic mycelia (Phillips and Borgia 1985). Nonetheless, strains of M. circinelloides also have a high capacity to produce ethanol aerobically, and ethanol yields exceeding those for biomass have been reported (McIntyre et al. 2002).

Therefore, the objective of this study was to investigate the utilisation of different carbon sources or mixtures of sugars in submerged batch cultivations with particular focus on ethanol production. Although a number of investigations have dealt with Mucor growth on various substrates, the performance of the organism in the controlled environment of a fermenter was of interest as most of the previous results have been based on shake flask cultivations. We showed that M. circinelloides was indeed capable of ethanol production from glucose under fully aerobic conditions, a phenomenon often referred to as the "Crabtree effect" (de Deken 1966; van Dijken et al. 1993; van Urk et al. 1990). The attainment of physiological information about Mucor helps in contributing to the general understanding of dimorphic fungi, which are ideal model organisms for studying cellular differentiation processes. More importantly though, it might have implications for improved processing if M. circinelloides is to be used for the production of heterologous proteins. Thus, to meet demands such as low by-product formation and optimal biomass levels, detailed knowledge about ethanol metabolism will be essential. Another aim was to describe ethanol tolerance and utilisation. By applying image analysis techniques initial results from submerged batch fermentations were further substantiated.

Materials and methods

Microorganism

A Mucor circinelloides strain obtained from the American Type Culture Collection (ATCC 1216B) was used throughout this study.

The spore inocula were derived from rice flasks and prepared as follows: a 29 g quantity of rice was autoclaved with 6 ml of Vogel's medium (McIntyre et al. 2002), and 0.5 ml vitamin solution (60 mg ml⁻¹ niacin; 60 mg ml⁻¹ thiamine), 3 ml M9 buffer (60 g l⁻¹ Na₂HPO₄, 30 g l⁻¹ KH₂HPO₄, 5 g l⁻¹ NaCl, 10 g l⁻¹ NH₄Cl) and 7 ml H₂O were added by sterile filtration. Spores were harvested with 1% Tween 20 and stored in aliquots of 1 ml at –80°C in 20% glycerol. The inoculum size was 1×10⁶ spores ml⁻¹.

Cultivation medium

The medium used throughout this study was Vogel's medium (McIntyre et al. 2002), prepared with various carbon sources. Glucose, xylose, galactose and cellulose (Solca-floc) were sterilised in the fermenter vessel or shake flask; the other components were added after sterilisation through a 0.45 µm filter. When provided as a carbon source, ethanol was also added by sterile filtration.

Shake flask cultivations

Cultivations to study ethanol tolerance were performed in 500-ml shake flasks containing 100 ml liquid Vogel's medium at pH 5. The flasks were inoculated with medium containing 5×10⁵ spores ml⁻¹. The cultivation conditions were 25°C and 200 rpm. All experiments were carried out in triplicate.

Fermentation process

The bioreactor used in this study was a Braun BIOSTAT M (Braun, Melsungen, Germany) with a working volume of 1.5 l. The culture pH was maintained at pH 5 by automatic addition of 2 M NaOH. All fermentations were performed at 28°C. Dissolved oxygen in the culture medium was measured using an oxygen probe (Ingold). The agitation was 300 rpm; the stirrer was equipped with three five-blade Rushton turbines. For yeast-like growth, the gas flow was kept constant at 0.3 vvm (volumes of air per volume of fermentation medium per minute) with a gas mixture of 30% CO₂ and 70% N₂. Prior to inoculation, the culture vessels were sparged for 90 min to establish anaerobic conditions. For mycelial growth, the fermenter was sparged with air. The airflow rate was gradually increased up to 1 vvm and the stirrer speed up to 700 rpm.

Analyses: biomass, carbon source and ethanol

Dry weight was estimated by filtering approximately 20 ml of culture broth through preweighed filters (Gelman, 0.45 μm). Filter cakes were washed with 2 vol. 0.9% NaCl, dried at 100°C for 24 h and cooled in a desiccator before weighing.

The concentrations of d-glucose, xylose, galactose and ethanol in the filtrates were analysed by HPLC (Waters) equipped with an Aminex HPX-87H column (Biorad) and UV (Waters 486) and RI detectors (Waters 410).

Flow-through cell experiments

Spores were immobilised in a temperature-controlled flow-through cell mounted on a motorised stage with continuous medium flow to assess fungal growth and branching in a quantitative manner. The construction and set-up of the flow-through cell have been described in detail by Spohr (Spohr et al. 1998). The cell basically consisted of two slides (26 mm × 76 mm) separated by a Parafilm M spacer and clamped into a steel frame equipped with tubes for feed addition and waste withdrawal. To sterilise both the cell and the tubing, 70% ethanol was injected through a filter (0.45 μm); this was followed by addition of distilled water after 20 min to remove the ethanol. A 1 ml volume of 0.1% poly-d-lysine (Sigma) was then filtered into the cell to mediate spore fixation. The cell was inoculated with spores to gain a final number of 50–80 spores. Medium was continuously added and removed from the cell with a flow rate of 3 ml h⁻¹. All experiments were carried out at 28°C. Images of up to 40 hyphal elements were obtained at time intervals of 15 min on a Nikon Optiphot 2 microscope equipped with a CCD camera (Bischke CCD-5230P) connected to the image analysis system (Quantimet 600S, Leica Cambridge Ltd., UK). Automatic image analysis was applied for the detection of hyphal elements and measurements of length or projected area (Spohr et al.1998). Hyphal tips were counted manually. These values were used to calculate the hyphal growth unit length (tips per micrometre). Values for length and area were obtained directly in micrometres or square micrometres, respectively, after calibrating the program for the different microscope objectives using a Nikon objective micrometer.

Results

Growth and ethanol production on different carbon sources

Aerobic and anaerobic growth of M. circinelloides on Vogel's medium with 20 g l⁻¹ glucose in submerged batch cultivations are shown in Fig. 1. The culture vessels were sparged with air or a gas mixture of 30% CO₂ and 70% N₂ to gain mycelial or yeast-like growth, respectively. Ethanol was produced under both gas atmospheres, with the yield coefficient being higher aerobically than anaerobically (Table 1). The maximum ethanol production was approximately 6 g l⁻¹ during aerobic filamentous growth. The exact opposite was observed for growth on galactose where ethanol yields were higher anaerobically. The lowest amounts of ethanol were produced on xylose. For preliminary fermentation studies performed on cellulose under aerobic conditions (results not shown), ethanol accumulation was not detected. Low levels of glucose were present in the medium initially (<0.1 g l⁻¹); however during the course of the cultivation glucose could not be detected by HPLC analysis.

Fig. 1.

Cultivation of Mucor. circinelloides ATCC 1216B in Vogel's medium sparged with air (filamentous growth) or 30% CO₂/70% N₂ (yeast-like growth) respectively. Values for biomass (grams of dry weight per kilogram), ethanol (grams per litre) and residual glucose (grams per litre) over the time course of the fermentation are shown. y Yeast growth, f filamentous growth, bm biomass, etoh ethanol, glc glucose

Table 1. Specific growth rate (μ), yield of biomass (Y X/S) and ethanol (Y EtOH/S) on substrate for cultivations of Mucor circinelloides ATCC 1216B in Vogel's medium. n.d. Not determined

Yield coefficients for ethanol decreased and those for biomass concomitantly increased when less sugar was supplied during aerobic cultivations (10 instead of 20 g l⁻¹).

None of the aerobic batch cultivations performed became oxygen-limited. For each of the aerobic cultivations, the ethanol produced was utilised after depletion of the sugars; this was in contrast to the yeast cells, which did not have this ability due to their strict hexose requirements.

Ethanol tolerance and utilisation

Following the finding that ethanol produced aerobically served as a carbon source after sugar depletion, the utilisation of ethanol was screened in shake flasks (results not shown). On Vogel's medium containing up to 10 g l⁻¹ ethanol as the sole carbon source, final biomass levels of 4.8 g l⁻¹ were obtained. Increasing the concentration above this threshold proved to be inhibitory and led to decreased biomass yields. Moreover, samples from these cultivations showed aberrant morphologies, with hyphae being uncharacteristically thin and having few branches. At a concentration of 30 g l⁻¹ ethanol, final biomass levels were below 0.25 g l⁻¹ and at higher concentrations growth was totally suppressed.

In order to evaluate ethanol tolerance, different amounts of ethanol were then added to Vogel's medium containing 10 g l⁻¹ glucose (Table 2). Under these conditions 2.52 g l⁻¹ biomass was obtained at 30 g l⁻¹ ethanol; at higher concentrations germination failed and no growth occurred during the cultivation period of 72 h. With increasing ethanol concentration the cultures turned from a purely filamentous to a mixed morphology, consisting of both mycelial and yeast forms.

Table 2. Biomass level, biomass yield and morphological observation in shake flask cultivations of M. circinelloides in Vogel's medium containing 10 g l⁻¹ glucose supplemented with a range of ethanol concentrations. Shake flasks were harvested 72 h after inoculation

A series of batch cultivations in 1.5-l fermenters was conducted subsequently to investigate growth of M. circinelloides on ethanol in greater detail (Table 3). Cultivations with ethanol as the sole carbon source (7 g l⁻¹ and 12 g l⁻¹) displayed increased lag-phases (23 h and 35 h, respectively) before exponential growth began. However, supplementation with 1 g l⁻¹ glucose markedly reduced this time interval, and the duration of the lag phase approached that obtained for cultivations where glucose was the sole source of carbon. Supplementation with xylose caused a similar, though less pronounced effect. Higher concentrations of xylose were necessary to gain a significant decrease in the lag phase. For each of the cultivations supplied with a sugar and ethanol, consumption of ethanol was observed after sugar depletion.

Table 3. Start of exponential growth after inoculation for cultivations of M. circinelloides ATCC 1216B with a range of carbon sources in Vogel's medium. E Ethanol, G glucose, X xylose

Flow-through cell experiments were performed with Vogel's medium containing 10 g l⁻¹ ethanol (results not shown) to follow the development of individual spores. A specific growth rate of 0.5±0.05 h⁻¹ (based on the increase in hyphal length) was obtained for the 10 g l⁻¹ ethanol cultivation, but spore swelling only occurred 30 h after inoculation into the flow-through cell. The exponential growth phase was reached after 35–38 h, as was also observed during submerged cultivations (Table 3).

Growth on mixed sugars

In order to study the performance of M. circinelloides on a mixture of sugars, submerged batch cultivations were performed with Vogel's medium containing 10 g l⁻¹ glucose and 10 g l⁻¹ xylose (Fig. 2). Consumption of xylose did not commence before glucose was depleted entirely; the organism was able to switch rapidly from one sugar to another as illustrated by the biomass growth curve that continued without visible disturbance. The ethanol produced during growth on glucose was consumed after xylose depletion. Flow-through cell experiments where the medium feed was changed from 10 g l⁻¹ glucose to 10 g l⁻¹ xylose or vice versa were then carried out to assess any changes in hyphal growth during this transition phase. The specific growth rate (μ) was more than halved during the first 60 min after the shift from glucose to xylose. Following this adaptation to the new carbon source, μ rose to almost its original value (Fig. 3). For the hyphal element presented in Fig. 3 this meant a temporary decrease from μ=0.62 h⁻¹ (growth on glucose) to μ=0.25 h⁻¹ (adaptation to xylose) followed by an increase to μ=0.61 h⁻¹ (growth on xylose).

Fig. 2.

Cultivation of M. circinelloides ATCC1216B in Vogel's medium containing a mixture 10 g l⁻¹ glucose and 10 g l⁻¹ xylose as carbon source sparged with air (filamentous growth). Values for biomass (BM), ethanol and residual substrates over the time course of the process are shown

Fig. 3.

Flow-through cell cultivation of M. circinelloides ATCC 1216B where the medium was shifted from containing 10 g l⁻¹ glucose to 10 g l⁻¹ xylose (xyl). Increase of hyphal length (micrometres) before and after the shift are shown. HGU L Hyphal growth unit length

Discussion

Ethanol production/Crabtree effect

In contrast to previous reports (Philipps and Borgia 1985), the strain of Mucor employed in this study is capable of aerobic ethanol production. Yield coefficients for aerobic growth on glucose even exceeded those obtained under anaerobiosis (Fig. 1). However, anaerobic growth on glucose, whether in the mycelial or yeast-like form, has been reported to be accompanied by ethanol accumulation (Phillips and Borgia 1985), while little or no ethanol production has been linked with aerobic growth of the fungus. This discrepancy could either be due to a true difference in the metabolism of the strains or more likely due to the fact that most studies on ethanol production have been performed in shake flasks that were sampled at a stage after consumption of both the added carbon source and the metabolic product ethanol.

An important outcome of these experiments was the fact that an abundance of sugar in the cultivation medium, in particular glucose, resulted in increased ethanol yields at the expense of biomass development (Table 1). Bearing in mind an industrial application for the production of heterologous proteins, these are important findings. A high rate of conversion of substrate into biomass, and subsequently the product of interest, is not only desirable to make a process economically feasible, but, furthermore, minimisation of by-product formation would avert detrimental effects on growth and productivity, which could be caused by high levels of accumulated ethanol. The organism was still capable of growth in the presence of up to 30 g l⁻¹ ethanol, but this level restricted growth severely (Table 2).

In addition, these results further substantiate the suggestion that M. circinelloides exhibits the so-called "Crabtree effect", i.e. ethanol production occurring under full aerobiosis when sugar is present in excess. This phenomenon has been studied most thoroughly for Saccharomyces cerevisiae (de Deken 1966; van Dijken et al. 1993; van Urk et al. 1990). Nevertheless a certain controversy surrounds the term, as the underlying regulatory mechanisms are perceived differently (Alexander and Jeffries 1990), and long- and short-term Crabtree effects have been defined (Petrik et al. 1983). The data presented here are in line with those found for S. cerevisiae, where glucose repression is known to play a role in the Crabtree effect, particularly the findings that the ethanol yield increases with increasing glucose concentration and is lower during growth on glucose.

Whilst non-Saccharomyces yeasts cease to grow in the complete absence of oxygen, though still require oxygen-limited conditions to trigger alcoholic fermentation (van Dijken et al. 1993), M. circinelloides can grow under complete anaerobiosis and also produce ethanol when oxygen is not limited, as was seen for the experiments reported in this study. More evidence for the role of abundant sugar in actuating aerobic fermentation in M. circinelloides was provided by the fact that no ethanol was produced during growth on cellulose. This would indicate that alcoholic fermentation was prevented because consumption of glucose proceeded at almost the same rate as it was liberated; thus an excess of sugar would not be available at any point during the cultivation.

Xylose consumption/glucose repression

Generally, yields of ethanol on the hexoses tested were much higher than those obtained on xylose (Table 1). This carbon source was not readily fermentable under aerobic conditions and biomass yields were high. M. circinelloides is not only able to grow on a wide range of carbon sources aerobically (Bartnicki-Garcia and Nickerson 1962b; McIntyre et al. 2002; Orlowski 1991), it also has the capacity to readily adapt to a new carbon source. Experiments in the flow-through cell showed that the organism requires no more than 75–90 min to resume growth at almost its previous μ after a shift from glucose to xylose (Fig. 3). However, growth on xylose was clearly subject to glucose repression as glucose was consumed first when both sugars were present in the growth medium. This repression is mediated by the product of the creA gene in Aspergilli (Ruijter and Visser 1997). Similarly, as for other filamentous fungi (Felenbok et al. 2001), ethanol consumption by Mucor was subject to repression in the presence of sugars and it was only metabolised once other carbon sources were exhausted.

Ethanol toxicity and effects on morphology

Ethanol is one of the metabolic products of M. circinelloides and could be utilised as the sole source of carbon by the mycelial form, but the organism was still relatively intolerant of ethanol. With increasing concentrations of ethanol, a decrease in biomass levels and a delay in germination occurred (Tables 2, 3), as previously reported for Mucor fragilis (Serrano et al. 2001). Additionally, elevated ethanol levels caused both strains to turn from a strictly filamentous to a mixed morphology.

Above a critical concentration M. fragilis has been shown to revert exclusively to a yeast-like mode (Serrano et al. 2001), whereas M. circinelloides cultures consisted of both mycelial and yeast cells until spore swelling finally ceased entirely at a concentration of 50 g l⁻¹ ethanol or above (this study). Despite the fact that anaerobic conditions are typically required to facilitate yeast-growth, a number of morphogens are known to mediate aerobic yeast development, including cAMP and phenethyl alcohol (Orlowski 1991). So far, a common link in the way these compounds alter cell morphology has not been established. Increasing lipid unsaturation was demonstrated for M. fragilis with increased ethanol addition, a mechanism also displayed by bacteria and yeasts to maintain an effective plasma membrane (Serrano et al. 2001). M. circinelloides might respond similarly in an adverse environment. For large-scale processes, such a morphological conversion in an attempt to overcome alcohol toxicity would constitute a major obstacle if the secretory capacity of the filamentous form is to be exploited.

Conclusion

The results presented confirm that M. circinelloides is a Crabtree-positive organism, i.e. it is capable of fully aerobic ethanol production on glucose. It would seem advantageous for future process design to further investigate growth and physiology of M. circinelloides on xylose, as high biomass yields with virtually no ethanol production were obtained in submerged batch cultivations. Rather than deleting genes involved in alcohol metabolism, simply using a carbon source such as xylose that does not stimulate ethanol production while giving satisfactory yields would be a much easier and faster approach. Although of quite a different nature, the findings might also be regarded from a viewpoint where ethanol production is desirable. With an increased demand for cheap sustainable fuels, Mucor might also be considered a potential organism to be engineered into a biocatalyst for ethanol production. Currently lacking the ability to ferment xylose, which might be introduced by genetic modification, the organism possesses many important traits required, e.g. anaerobic growth and ethanol production and a broad substrate utilisation range including complex carbon sources, which could be exploited provided the strain is successfully adapted to tolerate higher levels of ethanol.