Growth, productivity, and sugar tolerance of K. marxianus and S. cerevisiae grown on glucose
Sugars can quickly accumulate to very high concentrations during high solids SSF if the fermentative organism is unable to rapidly consume the sugars as they become hydrolyzed by enzymes. High sugar concentrations in the fermentation broth can, in turn, create hyperosmotic stress on the cells.23 Coupling this stress with the need to operate at higher than optimal growth temperatures to foster sufficient enzyme action and ethanol accumulation results in osmotic, temperature, and ethanol stresses.33,34 To understand how these factors impact K. marxianus CBS 6556 and S. cerevisiae D5A, their growth and ethanol production were first evaluated by glucose fermentations when subjected to (i) a higher temperature, (ii) a high osmolarity, and (iii) evaluation of the combined effect of (i) and (ii). First, glucose concentrations of 50 and 150 g/L were fermented by both strains at 37 and 43 °C to determine how temperature and glucose concentration impacted performance. The optical density results in Figure 1 show that at 37 °C, CBS 6556 grew almost twice as fast as D5A for both 50 g/L and 150 g/L glucose concentrations. Thus, although both strains grew on both glucose concentrations, K. marxianus outperformed S. cerevisiae at 37 °C, a temperature typically employed to achieve reasonable enzyme activity in SSF. It is important to note that the growth of both CBS 6556 and D5A were hindered in the presence of high glucose at high temperature. However, the performance of D5A suffered much more under the combined stresses of temperature and higher glucose concentration. This data also reveals that K. marxianus maintained high growth rates at glucose concentrations of 50 g/L and 150 g/L at 43 °C, while S. cerevisiae failed to grow at either concentration at this temperature. These results highlight the unique capabilities of CBS 6556 when compared to D5A and its potential to support higher temperature fermentation where fungal enzyme activity is also higher.
Next, the effect of glucose concentration on ethanol production by each organism was evaluated by fermenting glucose concentrations of 150, 180, and 200 g/L. As shown in Figure 2, D5A and CBS 6556 both performed well for all glucose concentrations at 37 °C. This data also showed that CBS 6556 had a higher initial ethanol productivity at the larger glucose concentration, but performed similarly to D5A at other glucose concentrations. It is interesting to note that despite the slower growth rates for D5A shown in Figure 1, D5A was able to produce ethanol at a similar rate to the faster growing CBS 6556. Furthermore, after 2 days of glucose fermentation by both yeasts, concentrations of ethanol and glucose indicate that CBS 6556 left more glucose in solution than D5A for the two lower starting concentrations of glucose, while residual glucose reached almost 50 g/L for both strains when grown on 200 g/L glucose (Table S1). The ethanol concentration from both yeasts did not increase significantly when the glucose concentration was raised from 180 to 200 g/L, as observed by the significant increase in residual glucose shown in Table S1, suggesting that both yeasts were reaching an ethanol tolerance limit of about 80 g/L.
Ethanol productivity and yields for high solids SSF of CELF pretreated poplar
In light of the glucose fermentation results, CBS 6556 and D5A would be expected to have similar ethanol tolerance and productivity and not be inhibited by the glucose concentrations expected in SSF. However, these results along with the higher growth rate, albeit on glucose, indicated that CBS 6556 should be more suitable than D5A for SSF at higher temperatures. To test whether these attributes would enhance SSF performance, each organism was employed for high solids SSF of CELF pretreated poplar. The CELF pretreated substrate in this study comprised of 88.5 % glucan, 3.0 % xylan, and 2.3% acid-insoluble lignin. SSF experiments were conducted at 13, 17, and 20 wt% insoluble solids corresponded to 11, 15, and 18 wt% glucan-equivalent loadings. Both D5A and CBS 6556 were run at 37 °C, while CBS 6556 was also used in SSF at 43 °C to take advantage of the higher temperature tolerance displayed for glucose fermentations. A Cellic® Ctec 2 enzyme cocktail was employed for each fermentation at a dosage of 15 mg protein per g glucan in raw poplar.
Operation of CBS 6556 at 43 °C initially resulted in higher ethanol productivities (Figure 3 (a)), but 5 day yields for all three experiments were approximately the same (63%) and did not significantly increase at longer times. At a higher initial glucan concentration of 15%, the productivity of CBS 6556 at 43 °C was greater at an even shorter period of time (Figure 3(b)) and when operated at 37 °C, both D5A and CBS 6556 had similar productivities up to day 5, after which D5A increased slightly while CBS 6556 leveled off. However, while the final yields for D5A at both 11 and 15 wt% glucan loadings were about the same, the yields dropped with increased glucan loadings for CBS 6556, particularly for operation at 43 °C. For application of SSF at 18 wt% glucan loadings, D5A demonstrated similar productivities and yields to those for both 11 and 15% glucan. On the other hand, although CBS 6556 operation at 37 °C closely followed the ethanol yields and productivities of D5A for the first 3 days, it virtually stopped ethanol production thereafter. The results show that the yield did not exceed 60% of the theoretical maximum and ethanol production ceased. Thus, these results show that operation of CBS 6556 at 43 °C exhibited the highest initial fermentation rates for 11 and 15 wt% glucan loading, potentially due to higher sugar release by cellulase operated nearer to its optimum temperature. However, CBS 6556 also suffered from a much earlier fermentation arrest, likely due to the combined effects of higher ethanol concentrations and temperature.
The results presented in Figure S1 shed additional light on factors that caused a premature fermentation arrest during high temperature SSF , Figure 3. As shown, D5A completely converted glucose released by the enzymes at 11 and 15 wt% glucan loadings and left only a little glucose in solution at the end of the 18% glucan run. On the other hand, when CBS 6556 was operated at the same temperature as D5A (37 oC), glucose accumulation progressively increased with glucan loading to reach about 30 g/L at the two highest loadings. Furthermore, because ethanol production virtually stopped at the point glucose started building up, the greater amount of ethanol appeared to stop fermentation at these points. However, it is noteworthy that the final ethanol concentration increased with glucan loading, suggesting that faster glucose release from more glucan allowed more ethanol to form before the fermentations stopped. Increasing the temperature to 43 °C resulted in glucose buildup earlier in the fermentation and premature cessation of ethanol production at lower concentrations.
Overall, these results show that operation of CBS 6556 at 43 °C exhibited the highest initial fermentation rates for 11 and 15 wt% glucan, due to faster sugar release by cellulase operated nearer to its optimum temperature. However, CBS 6556 also suffered from a much earlier fermentation arrest due to the combined effects of higher ethanol concentrations and temperature. This outcome is consistent with results with K. marxianus strains capable of fermenting glucose and cane syrup at high temperatures of up to 47 °C that showed that although fermentation was rapid initially, the organism suffered from a rapid rate of cell death at higher temperatures in high gravity fermentations.15 Other studies also observed a high temperature later-stage ethanol fermentation arrest by K. marxianus. 35,36
Impact of glucose, ethanol, and temperature on yeast
Yeasts, in general, are polymorphic organisms and can take many sizes and shapes such as ellipsoidal, spherical, or elongated cylinders, depending on the environment to which they are exposed.37,38 Hyperosmotic stress, due to increased glucose concentration, results in rapid water diffusion from the yeast cells into the surrounding medium, thereby leading to loss of cell wall turgor pressure and cells shrinkage. Higher ethanol concentrations act adversely on the integrity of the cell membrane by increasing membrane fluidity and permeability that result in cellular ion leakage.39 Ethanol also negatively impacts cell metabolism and inhibits cell growth and cell division.40 In response to hyper osmolarity and ethanol shock, the cells can accumulate glycerol or other polyols such as arabitol, mannitol, meso-erythritol, and/or xylitol to alter the equilibrium between the intracellular and extracellular environments and reduce diffusion of intracellular water.23,34,41,42 The result can be an increase in cell volume due to swelling. Heat shock, however, not only increases cell membrane fluidity but also causes protein damage, practically killing the organism unless it possesses heat-shock proteins (HSP), i.e., proteins that enhance thermotolerance of unicellular organisms like yeasts and bacteria. HSPs usually protect thermally damaged proteins from accumulation, unfold aggregated proteins, and refold damaged proteins or efficiently degrade them.33,43
Figure 4 shows that while D5A produced some glycerol initially for SSF at 18% glucan-equivalent solids loadings, glycerol production was relatively unchanged as ethanol production continued at 37 °C. At the same temperature, CBS 6556 coproduced glycerol along with ethanol, and glycerol production plateaued at a 50% higher level than for D5A when ethanol production ceased. Figure 4 also reveals that glycerol production similarly followed ethanol build up for SSF by CBS 6556 at 43 °C and again leveled off when ethanol production stopped. However, the concentrations of ethanol and glycerol stopped building up at somewhat lower concentrations than for operation at 37 °C. Figure S2 reports that for SSF by D5A at 37 °C, glycerol concentrations increased by about 50% when glucan loadings were increased from 11 to 18 wt%. However, Figure S2 also shows that although glycerol levels reached a similar high value for SSF of 11 wt% glucan for CBS 6556 at 37 °C, the amount rose with increasing glucan loadings to reach about 250% of the amount at 18% glucan. Increasing the temperature to 43 °C for SSF by CBS 6556 resulted in a ~50% increase in the maximum glycerol produced with 11 wt% glucan loadings. At higher glucan loadings, glycerol production increased but only modestly. Overall, the higher glycerol concentrations produced by CBS 6556 suggests that it was more stressed by the coupling of ethanol and temperature than D5A.
In order to further study the impact of temperature and ethanol concentration on CBS 6556 and D5A performance, electron micrographs were taken of both strains following fermentation of pure glucose and SSF of CELF pretreated poplar. As shown in Figure 5(a-h), both D5A and CBS 6556 cells maintained ellipsoidal or yeast-like morphologies when grown in an anaerobic environment. Therefore, we assumed the cells to be prolate ellipsoids and estimated their total surface areas and volumes based on their vertical and horizontal dimensions.12 Although it was difficult to precisely image fibrous biomass in the SSF broth, it appeared that the oval structures highlighted in the yellow boxes (Figures 5 c, g, and h) are similar in shape to the native ellipsoidal yeast. The cell volume estimations are calculated based on an elliptical geometry (Figure 6). Figure 5-f also revealed that CBS 6556 cells suffered substantial surface damage including shrinking and wrinkling, likely due to greater shock at 43 °C compared to the behavior of this yeast (Figure 5d) and D5A (Figure 5b) under similar stresses at 37 °C. Figure 6 further indicates that when subjected to a 150 g/L glucose concentration at 37 °C, the cell volumes of D5A and CBS 6556 increased by 66.0% and 46.64%, respectively, as compared to their sizes at seed culture conditions. However, when subjected to higher ethanol concentrations at 43 °C, the average volume of CBS 6556 cells dramatically shrunk by almost 64%. These observations further indicate that CBS 6556 was more stressed by high concentrations of ethanol than D5A and the adverse impact was more pronounced at a higher temperature resulting in shrinking of CBS 6556 cells to an abnormally small size with quite noticeable surface damage.
Figure 6 shows similar observations from SSF of 18 wt% glucan at the end of 5-day glucose fermentations in that D5A and CBS 6556 volumes expanded by 16.8 % and 6.97 %, respectively, at 37 °C, while CBS 6556 contracted by 43.66% at 43 °C. However, as shown in Figure 4 and Table S2, the glucose concentration remaining at the end of 5 days of SSF at 43 °C was less than 50 g/L, a value within the tolerance limit of CBS 6556. This outcome indicated that the lower ethanol productivity could be attributed to reduced ethanol tolerance of CBS 6556 cells at higher temperatures.44
Overall, these results suggest that CBS 6556 cells suffered major cell damage due to the combined effects of ethanol and heat shock. Because the cells were unable to make sufficient glycerol and/or maintain the turgor pressure of the cell wall, they shrunk to an abnormally small size. In addition, yeast cells need a critical size that is characteristic for the growth medium to initiate budding, and extremely small cells are incapable of budding, thereby arresting the cell cycle.23 The atypically small cell size at high temperature and higher ethanol concentrations appeared to limit growth and metabolism of CBS 6556, thereby causing premature cessation of sugar uptake and fermentation at elevated temperature.
These observations are consistent with an analysis by Li et al.36 of protein samples collected during K. marxianus fermentations at 45 °C that revealed some biochemical and enzymatic modifications triggered by stress conditions. They observed that some of the proteins related to gene transcription and translation, along with some of the proteins involved in oxidative phosphorylation, were down-regulated in K. marxianus after fermentation arrest. The repression of transcription and translation can be attributed to a self-defense mechanism to cope with stress condition during the late fermentation. Potentially, up-regulation of some molecular chaperones and proteasome proteins involved in the protein quality control (PQC) system after fermentation arrest could also be a limiting factor. The interactions of the proteins in the PQC system are responsible for the folding of proteins, refolding of misfolded proteins, and degradation of misfolded and damaged proteins. These observations provide some explanation for the observed fermentation halt and offer possible opportunities for metabolic engineering towards improvement of the stress tolerance in K. marxianus.