Mineral Composition of the Sugarcane Juice and Its Influence on the Ethanol Fermentation

Abstract

In the present work, we evaluated the mineral composition of three sugarcane varieties from different areas in northeast Brazil and their influence on the fermentation performance of Saccharomyces cerevisiae. The mineral composition was homogeneous in the different areas investigated. However, large variation coefficients were observed for concentrations of copper, magnesium, zinc and phosphorus. Regarding the fermentation performances, the sugarcane juices with the highest magnesium concentration showed the highest ethanol yield. Synthetic media supplemented with magnesium also showed the highest yield (0.45 g g⁻¹) while the excess of copper led to the lowest yield (0.35 g g⁻¹). According to our results, the magnesium is the principal responsible for the increase on the ethanol yield, and it also seems to be able to disguise the inhibitory effects of the toxic minerals present in the sugarcane juice.

Introduction

The composition of sugarcane juice may vary, affecting several industrial operational parameters, such as the fermentation yield. Several parameters are implicated in determining the final composition of the sugarcane juice, such as the variety of sugarcane, soil type, fertilization strategies, climatic conditions, maturity level of the sugarcane, type of harvest, period of time between burning, cutting and processing, content tips and straw and also by the use or not of vinasse irrigation [1].

Although the ash content in sugarcane juice is somewhat constant, its mineral composition can vary widely, depending on the source of raw material. In order to minimize these differences, many distilleries use the strategy of adding molasses to the broth [1]. This mixture is the most recommended since some stocks have deficient nutritional level and molasses usually contains high concentrations of minerals in its composition [2]. In spite of the importance, the mineral composition in broths of sugarcane and molasses used in distilleries in Brazil are often overlooked by the producers. Several minerals are important in the nutrition of yeast by allowing proper cell metabolism and growth. The major components are ammonium, phosphorus, sulphur, potassium, magnesium, calcium, zinc, manganese, copper and iron. Additionally, in the case of fermentative metabolism, minerals such as potassium, magnesium, calcium, manganese, iron, zinc and copper are somehow essential [3].

Some of these minerals, such as magnesium, manganese, zinc, copper and iron, play important roles as enzyme cofactors and in maintaining cell homeostasis [4]. However, excessive amounts of these components can be toxic and cause damage to the functions they are associated. For example, high levels of potassium and calcium present in sugarcane molasses can cause osmotic stress that impairs yeast fermentation performance [2]. Nitrogen can be assimilated by Saccharomyces cerevisiae as ammonium ion (NH₄ ⁺), amide (urea) or aminic (amino acids) forms, but not in the form of nitrate and little or none in the form of proteins. These components have a direct influence on the budding and specific growth rates and transport of sugars [5].

Potassium (K⁺) is the most abundant monovalent cation in yeast, consisting of 1–2 % of the yeast cell dry weight [3]. This is important in osmoregulation, charge balancing of macromolecules, regulation of phosphate and absorption of divalent cations [6]. It also acts as an activator of a series of reactions in glycolysis and other steps of metabolism, and it is also involved in ionic balance. The monovalent cation phosphorous (P⁺) is essential for the energy metabolism and the synthesis of nucleic acids. This element is considered essential for carbohydrate absorption and transformation of sugar into alcohol and for the production of ATP from the glycolysis and the respiratory chain, and of course for the regulation of enzymes and metabolic mechanisms by kinases [3]. Magnesium (Mg²⁺) is the most abundant intracellular divalent cation and represents around 0.3 % of the yeast cell dry weight. This mineral acts as a cofactor for over 300 enzymes involved in different metabolic reactions, such as DNA and ATP synthesis. Magnesium plays multifaceted roles in the physiology of yeast cells at cytological, biochemical and biophysical levels and is very important in industrial fermentation processes, in which it is necessary for activation of several glycolytic enzymes and in protecting from environmental stresses during fermentation, such as those caused by ethanol, high temperatures or high osmotic pressure [3]. Therefore, its availability in the medium, cellular absorption and subsequent metabolic utilization seems to be a prerequisite for achieving the maximum fermentation activity of the yeast cell [3].

On the other hand, the divalent cation calcium (Ca²⁺) presents relatively few and more specific biochemical functions, being required in much smaller quantities by the cell when compared with Mg²⁺ [7]. It can induce connections between surface proteins of different cells, causing flocculation that is important in beer fermentation [3]. Although Ca²⁺ is apparently not necessary for the growth of yeast cells [8], it is important in fermentation for their involvement in cellular protection to ethanol stress [9]. It was recently suggested that the assimilation of Ca²⁺ could help in the protection and tolerance to medium acidification [10], since this mineral acts as an intracellular marker of homeostatic regulation [11]. As another relevant divalent cation, zinc (Zn²⁺) is an essential element for normal growth, metabolism and physiology of the yeast cell, besides acting as a cofactor for many proteins [12]. It also plays an important role in the metabolism of the yeast fermentation, because it is essential for the activity of alcohol dehydrogenase [13], also presenting a crucial role in the structure of enzymes and non-catalytic proteins [14]. Zinc in appropriate amounts in the medium promotes the growth of yeast cell and ethanol production [15]. In contrast, deficiency of Zn²⁺ blocks cell growth and fermentative activity, while high concentrations of Zn²⁺ can be toxic to the cell, affecting the membrane permeability to K⁺ causing decrease in cell growth and fermentative activity [4].

Other metals can influence the fermentation, such as iron (Fe³⁺), manganese (Mn²⁺) and copper (Cu²⁺). These are required as cofactors of enzymes (especially Mn⁺²) and yeast respiration as components of redoxisomes (Cu⁺² and Fe⁺³) [16]. Copper acts as cofactor of some enzymes, such as cytochrome C oxidase and Cu/Zn-superoxide dismutase [4]. Manganese is essential for yeast cell as a trace element, playing important roles in the metabolism as part of some enzymes, such as pyruvate carboxylase [4], but also acting as competitor of Mg²⁺ for some enzymes of ATP and DNA synthesis [17]. Moreover, iron is essential for the yeast but may also be toxic, and its utilization by the yeast cells is quite well regulated [18]. S. cerevisiae can easily grow in culture media in which Fe³⁺ is very scarce or very abundant. Yeast cells are also able to grow in Fe³⁺-free medium for several generations, indicating that they express efficient mechanisms for storage and mobilization of iron [19].

At last, it should be pointed out the toxic effects of aluminium (Al³⁺) that inhibits productivity of many crops, and thus, for economic reasons, considerable amount of information is available in the literature about its effects on plants [20]. It was recently identified as stressful element of the yeast in industrial fermentation conditions, causing a simultaneous reduction of the viability and concentration of the yeast trehalose [21].

In the past few years, we have dedicated efforts to study the composition of microbial fermentation processes for ethanol fuel production, especially in the northeast region of Brazil, describing the main species of yeasts and lactic acid bacteria whose presence could risk the efficiency of the industrial process [22–25]. However, in some of these studies, we could observe that the microbiota of these processes was not associated with problems in fermentation. For this reason, we turned our attention to the analysis of the mineral composition of the substrate and also how changes in this composition may influence the fermentative capacity of yeast.

Materials and Methods

Sugarcane Varieties and Juices

Three different varieties of sugarcane that represented the most commonly found in farms providing for the distilleries in the northeast Brazil were analysed. These varieties were bred by the RIDESA consortium (Brazilian Interuniversity Network for the Development of Sugarcane). The sugarcane varieties RB867515, RB92579 and RB863129 had been collected from agricultural fields of three distilleries. The first distillery (site IT) is located in the municipality of Aldeias Altas (04° 37′ 40″ S 43° 28′ 15″ O), state of Maranhão, which presents a yellow latosol-type soil, a typical coastal plain soil with poor nutrient clay material and high cohesion of structural aggregates [26]. The second distillery (site TB) is localized in the municipality of Caaporã (07° 30′ 57″ S 34° 54′ 28″ O), state of Paraiba, and the third distillery (site ST) is in the municipality of Goiana (07° 33′ 39″ S 35° 00′ 10″ O), state of Pernambuco. In both sites TB and ST, the soil is classified as spodosols, a strongly leached ashy grey-type acidic soil with limited suitability for cultivation that normally depends of correct fertilization [26]. Sites TB and ST are close to each other approximately by 17 km, and both are 1,300 km far away from IT site. These sugarcane varieties are commonly cultivated in all the sites, and it has to be stated that those sugarcanes were cultivated under strict controlled conditions of irrigation and fertilization by the distilleries. Samples were collected in the beginning of the harvest season 2011/2012, in august, in a randomized design. It only collected the maturated samples; in other words, the plants had an equally distributed sugar content (BRIX°) in the whole stem. In these analyses, the handheld refractometer was used (IPS 10 T-Impac), and the samples were collected from the apex, middle, and bottom of the stem and then measured the maturation level. Twenty-seven samples were collected for each variety. Then, the broth were extracted from the sugarcane and stored in a freezer (−20) for the subsequent analysis. Part of the sugarcane juice was subjected to digestion with sulphuric acid and hydrogen peroxide to measure the following nutrients [27]: total nitrogen (TN) by Kjeldahl method; total phosphorus by colorimetric method; potassium contents by flame photometer [28]; and calcium, magnesium, iron, zinc, aluminium, copper and manganese by atomic absorption spectrometry.

Samples from 30 independent farms that supplied sugarcane from different varieties to distillery Miriri Agroindustrial, Santa Rita, Paraíba state, were collected in the harvest season 2009/2010 and divided into two groups according to the average of industrial fermentation yield in the moment in which their canes were used by the distilleries: group A corresponded to the group of farms that supplied cane when the industrial yield in two distilleries was coincidently above 85 % and group B corresponded to the farms that supplied sugarcane when the industrial yield was coincidently below 80 %. The blends of group A and group B were used in the fermentation assays.

Yeast Strain and Media of Growth and Fermentation

Industrial strain S. cerevisiae JP1 was used in the present study, since this yeast is the most used in the distilleries of northeastern Brazil [29]. In some experiments, we used the industrial strain S. cerevisiae PE-2 that is commonly used by distilleries in southern Brazil [30]. The cells were maintained and pre-grown in YPD (10 g L⁻¹ yeast extract, 20 g L⁻¹ glucose, 20 g L⁻¹ bacteriological peptone, 20 g L⁻¹ agar) at 32 °C. For the fermentations with sugarcane juice, the broths from the RB varieties were unfrozen and sterilized by autoclavation (121 °C, 15 min). Afterwards, the juices were centrifuged to remove suspended solids, and the supernatant was used for the fermentation. The synthetic media for the fermentations were prepared with YNB (Yeast Nitrogen Base, Difco) without ammonium sulphate and amino acids to final concentration of 1.6 g L⁻¹ and sterilized by filtration (sterile Millipore membrane 0.22 μm). Sucrose was used as carbon source to final concentration of 155 g L⁻¹. Mineral supplementation was elaborated with minerals in the form of salts as NH₄Cl, KH₂PO₄, KCl, CaCl₂·2H₂O, MgSO₄·2H₂O, AlCl₃, CuSO₄·5H₂O, ZnSO₄·7H₂O, MnSO₄·H₂O and FeSO₄·H₂O. The highest concentration of each mineral was based on the average of their concentrations in sugarcane juice in the group A and group B of distilleries described above. The amount of minerals present in the YNB medium formulation was incorporated in this calculation.

Fermentation Assays

The yeast cells were pre-grown in YPD at 32 °C for successive cycles of 24 h in a rotator shaker at 140 rpm. After each cycle, cells were recovered by centrifugation (1,200×g for 5 min at room temperature) and suspended in fresh medium for a new cycle. It was repeated until accumulation of enough cells to perform the fermentation assays. All the fermentations were carried out on a single-batch in a 250-mL flasks containing 200 mL of sugarcane substrate separated in two sets of tubes. In the first set of tubes, a blend of juices from different cane varieties were added cultivated in the same geographical site. In the second set of tubes, a blend of juices from the same cane varieties were added cultivated in different geographical sites. Cells from pre-cultures were suspended to 10⁸ cells mL⁻¹ in each blend and left to ferment for 6 h at 32 °C without agitation. At defined times, samples were withdrawn, centrifuged and the supernatant taken for metabolite analysis. The measurement of carbohydrates, glycerol, acetate and ethanol were carried out by HPLC (Waters Co., USA) using an Aminex HPX-87H column (BioRad, USA) heated at 60 °C. Sulphuric acid at 5 mM was used as mobile phase at a flux of 0.6 L min⁻¹ [31]. The yeast cells were evaluated to viability by direct microscopic count on a Neubauer chamber after dying with methylene blue. The production of CO₂ was measured by the weight loss approach [25] before each sampling, taking into account the loss of volume in the previous sampling.

Statistical Analysis

The mineral composition determination in each distillery was performed in nine samples from each sugarcane variety. Fermentations with sugarcane juice substrates were carried out in duplicate while fermentations with the synthetic media were performed with four biological replicates. The results were subjected to analysis of variance (ANOVA) and were compared by the Tukey test (α = 0.05) using the software ASSISTAT [32].

Results and Discussion

Variation in Mineral Composition of Sugarcane Juice Changes the Yeast Fermentation Performance

During the monitoring of eight distilleries from 2004 to 2012, samples of the feeding substrate (sugarcane juice diluted to around 12 °Brix) were collected throughout the harvesting periods, and fermentation tests were performed in laboratory using fresh cell biomass of the industrial strain S. cerevisiae JP1. This strain was isolated as dominant yeast in industrial processes in northeast of Brazil [29], and it has been used as fermenting strain in several distilleries ever since. Typical results from two distilleries from the state of Paraiba showed the significant variation in the efficiency of fermentation by JP1 strain when using the industrial substrate (Fig. 1). Given that in such experiments, the only yeast was the cells of JP1 strain and that the bacterial population was at low count (<10⁴ cells/mL), thus, the straight explanation for the fall in the ethanol yield was the chemical composition of the industrial substrate.

Fig. 1

Fermentation efficiencies of S. cerevisiae JP1 strain when using sugarcane juice supplied by distilleries Japungu Agroindustrial (a) and Miriri Agroenergy (b) along 21 weeks of harvesting season 2009–2010

Our research group has surveyed for a long time the industrial fermentation processes for fuel ethanol production in the northeast Brazil, as well as in others regions, in the attempt to identify the most important events of microbial contamination that can affect their yields [24, 25, 29, 31]. However, we have observed along these years several episodes of drop in the industrial production of ethanol that could not be simply explained by microbial infection episodes. Despite the sense settled in the literature and among the distilleries worldwide that some contaminating micro-organisms are responsible for these problems, this relationship does not seem to be absolute. For example, declines in industrial yields were identified in distilleries in the northeast of Brazil that are commonly susceptible to episodes of contamination by the yeast Dekkera bruxellensis in periods where the population of this yeast was below the level considered critical or even when the population of S. cerevisiae was dominating the process [33]. From the constant monitoring of several distilleries in the past harvest seasons of sugarcane, it was possible to identify that the quality of the fermentation substrate also has an important role in industrial yield and can interfere with the fermentation capacity of the yeast cells.

From this finding, we conducted a randomized analysis of feeding sugarcane juice belonging to distilleries from the state of Paraiba (recorded as TB samples) and extended the analysis to sugarcane from distilleries in the neighbour states of Pernambuco (ST samples) and Maranhão (IT samples). The analysis of the mineral composition of sugarcane juice in these three sites showed that there is certain homogeneity among substrates. However, large coefficients of variation were observed for concentrations of Cu²⁺, Mg²⁺, Zn²⁺ and P⁺ (Table 1). In a previous work, we showed that sugarcane juice can vary in the content of nitrate assimilated by the cane depending on the physic-chemical composition of the soil, such as total acidity and pH, and such variation could be one of the causes of the settlement of D. bruxellensis contamination in the process [34]. It can be observed the lower concentration of Mg²⁺ and P⁺ in samples from the site IT than in the two other sites (Table 1). It may be a cause of lower production of ethanol observed in fermentation assays of this substrate using industrial strain JP1 (Fig. 2a). It is known that these minerals are essential for the functioning of the enzymes in the central metabolism and fermentation, cell energy metabolism and maintenance of intracellular pH [3]. The higher Cu²⁺ content in the samples of the ST site than in two other sites is also worth noting (Table 1). The optimal concentration of Cu²⁺ in the medium for yeast growth is in the range of 0.09 mg L⁻¹, and higher concentrations of this mineral can lead to metabolic changes with negative effect on yeast growth [16]. The protective effect of Mg²⁺ over toxic effect of Cu²⁺ has already been reported [35], but not using industrial substrates or simulating industrial conditions. Thus, this result corroborates the finding that such inhibitory effect of Cu²⁺ could be suppressed by the presence in high concentrations of some divalent cations in the industrial medium, such as Mg²⁺ (Table 1), leading to increased fermentation yield (Fig. 2a).

Table 1 Mineral composition of sugarcane juices in three different locations in the northeast of Brazil and its influence in the fermentation efficiency of S. cerevisiae JP1 industrial strain

Fig. 2

Fermentation efficiencies of the S. cerevisiae industrial strain JP1 using sugarcane juice from different origins. a Juice from three different sites of study in the states of Maranhão (IT), Paraiba (TB) and Pernambuco (ST) were composed by the pool of samples from three different sugarcane varieties. b Juice from the three different sugarcane varieties RB867515 (RB15), RB863129 (RB29) and RB92579 (RB79) were composed by the pool of samples from three different sites of study

Additionally, samples from these three sites were also grouped according to the three cane variety commonly cultivated: RB92579, RB863129 and RB867515. The blend of the sugarcane juices for variety RB867515 from three different sites showed higher content of Cu²⁺ (4.9 mg L⁻¹) than the varieties RB92579 (2.23 mg L⁻¹) and RB863129 (1.43 mg L⁻¹). This could be the cause of the lower fermentation efficiency of JP1 strain when the juice of RB867515 was used as substrate (Fig. 2b). Moreover, the juice from RB863129 variety was converted to ethanol at lower efficiency than the one from RB92579 (Fig. 2b), probably as the result of its lower content of Mg²⁺ (528 and 957 mg L⁻¹, respectively). Thus, it seems clear that there is a balance between the concentrations of these two minerals in sugarcane juice that affects fermentative ability of yeast cells. Once again, the results indicate that Mg²⁺ supplementation of the industrial wort that presents Cu²⁺ higher than 3 mg L⁻¹ would help to mitigate the inhibitory effect of the metal.

We did not detect any differences in the production of glycerol or acetate, two metabolites related to redox balance of the yeast cells, when the juices of different sites or different varieties of sugarcane were tested (data not shown). No statistical differences to the results for all these fermentation parameters showed in Fig. 2 were observed for the industrial strain PE-2 (data not shown), widely used in industrial processes in southern Brazil [30].

Excess of Minerals in the Sugarcane Negatively Affects Yeast Fermentation

After analysing samples of sugarcane juice with no apparent history of negative influence on industrial fermentation (Table 1), we evaluated samples of sugarcane juice that are constantly related to fermentation problems without any clear relationship with the settlement of contamination episodes either by bacteria or yeast (see Material and Methods). The blends of juices from groups A and B of farms (see Material and Methods for definitions) were analysed for mineral composition. A large variation among the mineral concentrations was observed when comparing samples from groups A and B (Fig. 3). In this case, the concentration of Cu²⁺ did not differ between these two groups, and so, this mineral could not be responsible alone for the fall in the industrial yield used as the initial parameter for definition of sugarcane group. However, it can be verified that in the group B, there was an excess of nitrogen (total and ammonium), P⁺, Ca²⁺, Mn²⁺ and Fe³⁺ (Fig. 3). Total nitrogen content of the juice may consider the presence of ammonium, free amino acids, nitrate and other forms of nitrogenous compounds in minor concentrations. Ammonium and nitrate can be found up to 250 mg L⁻¹ each in sugarcane juices, depending on the type of soil in which sugarcane is cultivated [34]. However, S. cerevisiae is incapable to assimilate nitrate and has very little capacity to assimilate traces of proteins, peptides and peptide-like compound that can be present in the medium [36]. Therefore, only ammonium may account as nitrogen source for the observed effect on fermentation.

Fig. 3

Mineral composition of the pool of juices from two groups of sugarcane supplier farms divided according to their influence in the industrial fermentation yield. Group A corresponded to the group of farms supplying cane when the industrial yield was above 85 % (light grey columns) and group B corresponded to the farms supplying sugarcane when the industrial output was below 80 % (dark grey columns)

Afterwards, two synthetic fermentation media that were formulated by mimicking the mineral composition of juices from groups A and B were used for fermentation assays. Their fermentation efficiency yields were compared to those using complete synthetic laboratory medium (YNB medium). The results showed fermentation efficiencies around 90 % when the cells of JP1 and PE-2 industrial strains fermented group A-like media, similar to that observed for complete synthetic medium (Fig. 4). On the other hand, the combination of various minerals in excess recorded in the pool of juices from group B promoted a reduction of the yeast fermentative capacity for the two strains for less than 50 % of the theoretical maximum (Fig. 4). The average of cell viability at the end of the fermentation assays was 85 % for the complete synthetic and group A-like media, while it was around 65 % in the group B-like media. This last group also showed higher buffering effect that avoided the drop of external pH below 4.0, which may be also detrimental for fermentation. Residual sucrose of 2.3 to 2.5 % was detected at the end of fermentation, which is more than ten times higher than the acceptable level of residual assimilable sugar in the distilleries. Even though the Mg²⁺ concentration was higher in the group B-like than that in group A-like media (Fig. 4), this was probably not enough to suppress the inhibitory effects of other minerals in excess. It is noteworthy that the Mg²⁺ concentration in group B-like medium was lower than observed in juice from ST or TB sites (Table 1).

Fig. 4

Fermentation efficiencies of the S. cerevisiae industrial strains JP1 (light grey columns) and PE-2 (dark grey columns) when using synthetic media supplemented with minerals according to the composition of sugarcane juice from group A (normal mineral) or group B (high mineral), as defined in the legend of Fig. 4. Synthetic medium without mineral supplementation was used as reference

Effects of Individual Minerals in the Fermentation

With the aim to study the effects of each mineral on the fermentation capacity of the yeast cells, we formulated ten synthetic fermentation media containing only one element in high concentration and one synthetic medium containing very low concentration of all the minerals (Table 2).

Table 2 Formulation of synthetic fermentation media containing the excess of one of the minerals measured in sugarcane juice samples provided by different farms in the city of Santa Rita, Paraiba state, Brazil

From this point, we defined high concentration of a mineral based on its average concentration in juices from over 30 sugarcane farms studied (Fig. 3) and its effect on the fermentation yield (Fig. 4). In the period of 6 h of fermentation, the yeast biomass did not vary significantly among the media tested (data not shown). The media could be sorted in four groups regarding the ethanol yields (Table 3). The highest ethanol yield as well as ethanol-specific productivity was observed in high-Mg²⁺ and in high-Mn²⁺medium, respectively (Table 3). The kinetics of fermentation in this medium showed no statistical difference in the consumption of sucrose (Fig. 5b) and the specific consumption rates of glucose (98 mg h⁻¹ g DW⁻¹) and fructose (115 mg h⁻¹ g DW⁻¹) released by the extracellular invertase activity compared to the fermentation kinetics in the reference medium (Fig. 5a) with specific consumption rates of glucose (105 mg h⁻¹ g DW⁻¹) and fructose (120 mg h⁻¹ g DW⁻¹). Magnesium absorption and subsequent metabolic utilization seem to be a prerequisite for achieving the maximum fermentation activity of the yeast cell [37]. Three key enzymes that subsequently work for ethanol biosynthesis are dependent on the Mg²⁺: enolase that converts 2-phosphoglycerate into 2-phosphoenolpyruvate [38], pyruvate kinase that converts 2-phophoenolpyruvate into pyruvate [39] and pyruvate decarboxylase that further converts pyruvate into acetaldehyde [40]. Therefore, we concluded that the high concentration of Mg²⁺ in the fermentation medium did not increase the transport of sugars, but rather incremented the metabolic conversion of sugar into ethanol. This might explain the highest ethanol yield achieved when fermenting sugarcane juices with higher contents of Mg²⁺, as we observed for the juice from RB92579 sugarcane variety (Table 1).

Table 3 Physiological parameters of batch fermentation of S. cerevisiae JP1 strain in synthetic medium with excess of one compound

Fig. 5

Fermentation kinetics of synthetic reference medium (a) and supplemented media with magnesium (b) and copper (c). All these fermentations were carried out with the industrial yeast JP1 (S. cerevisiae). Ethanol (white square) and glycerol produced (white diamond), as well as consumption of sucrose (black diamond), glucose (black square) and fructose (black triangle). Results represent the average value of four biological replicates

The second group corresponded to synthetic media containing high concentrations of Zn²⁺, Mn²⁺, Ca²⁺ and Al³⁺ that showed ethanol yield in the range of 0.40 to 0.42 g g⁻¹ (Table 3). This value corresponds to the average range of industrial fermentation processes amongst the production plants. Zinc is an important cofactor for the activity of alcohol dehydrogenase [13], which converts acetaldehyde into ethanol, and thus acts synergistically with Mg²⁺. Manganese is a potential competitor of Mg²⁺ in binding to enzymes, and its excess in the medium can lead to reduced ethanol production [3]. We observed that the highest Mn²⁺/Mg²⁺ ratio in the juice, the lower the ethanol yield, as in the case of the juice of IT site (Table 1; Fig. 2a). Calcium is present at higher concentrations in sugarcane molasses than in sugarcane juice [2]. It counteracts with Mg²⁺ by competitively inhibiting the activity of some key enzymes, and its toxic effect is suppressed by increasing the availability of Mg²⁺ in the fermentation medium [37]. In the present work, Ca²⁺ concentration is low and adequate for the maintenance of cellular homeostasis [11] and might not affect yeast fermentation performance. Therefore, this cation may be only relevant in molasses-fermenting process. Al³⁺ was reported as only mildly toxic for yeast cells at 21.5 mg L⁻¹ [41] and inhibited yeast by 72 % at concentration of 54 mg L⁻¹ [21], which are four to 12 times higher than the highest concentration we found in sugarcane juices (Fig. 3). Thus, like Ca²⁺, this cation seems not to be a concern for sugarcane juice fermentation.

The third group was composed by media containing high concentrations of P⁺, Fe³⁺, K⁺ and NH4⁺, showing lower ethanol yields in the range of 0.37 to 0.39 g g⁻¹ (Table 3). However, the consumption of sugars in these media was slightly higher (88.5 % ± 2.5) than that in Mg²⁺-enriched medium (83 %). It seems that sugar was diverted to a metabolite other than glycerol (Table 3). Moreover, acetate and lactate were not produced, and yeast biomass remained unaltered in these media (data not shown). Thus, CO₂ production may be increased in this medium. In the case of Fe³⁺, the reduction in ethanol yield could be explained by the induction of oxidative stress in cell yeast [16]. The ideal concentration of K⁺ for yeast growth is around 10 g L⁻¹, while no toxic effect was observed even at 100 g L⁻¹ in the medium due to the strict control of the membrane permeability to this ion by the action of Na⁺/K⁺ pump [42]. These concentrations are far higher than we observed in sugarcane juices, and so, its effect on the reduction of ethanol yield could not be determined. Tthe negative effect of phosphorus in the fermentation yield remains unclear, since no toxic effect was reported for the concentration range used in the present work. Instead, taking the results from P⁺ concentrations in juices from different locations (Table 1) and the yields of the correspondent juices (Fig. 2a), we can assume that such concentrations are still below the optimal for ethanol production, especially when in combination with low concentration of Mg²⁺.

At last, the fourth group refers to fermentation media containing high concentrations of Cu²⁺ with ethanol yield as low as 0.35 g g⁻¹ (Table 3). In an industrial point of view, the results represent an industrial efficiency of only 68 % of sugar conversion, imposing severe economic loss to the industry. Therefore, the use of sugarcane varieties that accumulate high Cu⁺² content in their juices independently of the site of cultivation, such as the RB867515 variety (Fig. 2b), can be problematic for the industrial process. The kinetic analysis of the fermentation showed that sucrose consumption was similar to that of Mg²⁺-enriched medium (Fig. 5c). However, a low specific consumption rate of glucose was calculated (89 mg h⁻¹ g DW⁻¹) and even lower specific consumption rate of fructose (75 mg h⁻¹ g DW⁻¹), when compared to the reference and Mg²⁺-enriched media. Thus, it seems that Cu²⁺ significantly reduced fructose uptake by the yeast cells, among other metabolic effects. Early report showed that supplementation of the medium with Cu²⁺ to 43 mg L⁻¹ stimulated ethanol production in YNB medium with glucose, although it extended the time of the experiments to days of fermentations [43]. This result is incompatible with the industrial approach of 8 to 12 h of fermentation. Moreover, the production of glycerol was lower than that in other media (Table 3), which indicates that Cu⁺² affected the redox metabolism. It is known that Cu²⁺ acts protecting the yeast cells against oxidative damages by the activation of Cu-Zn superoxide dismutase. Thus, the elevated external Cu²⁺ concentration can deregulate the delicate balance that separates its metabolic beneficial from deleterious effects for the yeast cells [44]. As stated above, such toxic effect could be suppressed by the adequate supplementation of Mg²⁺ in the fermentation medium.

Conclusion

In view of the results, we propose that the monitoring of Mg²⁺ concentration in the wort is the most important issue for industrial control of fuel ethanol fermentation process when using sugarcane juice. Keeping its concentration in the level above 600 mg L⁻¹ may prevent falls in ethanol production by the presence of toxic cations in the fermentation substrate.