Analysis of cellodextrin transporters from Neurospora crassa in Saccharomyces cerevisiae for cellobiose fermentation
- First Online:
- Cite this article as:
- Kim, H., Lee, W., Galazka, J.M. et al. Appl Microbiol Biotechnol (2014) 98: 1087. doi:10.1007/s00253-013-5339-2
- 629 Views
Saccharomyces cerevisiae can be engineered to ferment cellodextrins produced by cellulases as a product of cellulose hydrolysis. Direct fermentation of cellodextrins instead of glucose is advantageous because glucose inhibits cellulase activity and represses the fermentation of non-glucose sugars present in cellulosic hydrolyzates. To facilitate cellodextrin utilization by S. cerevisiae, a fungal cellodextrin-utilizing pathway from Neurospora crassa consisting of a cellodextrin transporter and a cellodextrin hydrolase has been introduced into S. cerevisiae. Two cellodextrin transporters (CDT-1 and CDT-2) were previously identified in N. crassa, but their kinetic properties and efficiency for cellobiose fermentation have not been studied in detail. In this study, CDT-1 and CDT-2, which are hypothesized to transport cellodextrin with distinct mechanisms, were introduced into S. cerevisiae along with an intracellular β-glucosidase (GH1-1). Cellobiose transport assays with the resulting strains indicated that CDT-1 is a proton symporter while CDT-2 is a simple facilitator. A strain expressing CDT-1 and GH1-1 (DCDT-1G) showed faster cellobiose fermentation than the strain expressing CDT-2 and GH1-1 (DCDT-2G) under various culture conditions with different medium compositions and aeration levels. While CDT-2 is expected to have energetic benefits, the expression levels and kinetic properties of CDT-1 in S. cerevisiae appears to be optimum for cellobiose fermentation. These results suggest CDT-1 is a more effective cellobiose transporter than CDT-2 for engineering S. cerevisiae to ferment cellobiose.
KeywordsCellulosic ethanolCellodextrin transportersIntracellular β-glucosidaseEngineered S. cerevisiae
Saccharomyces cerevisiae has been widely used to produce ethanol from sugars derived from sugarcane and corn (Shapouri et al. 2006). To avoid further impacting the food and feed markets, producing ethanol from nonedible crops has been proposed (Wyman 2007). However, saccharification of cellulose into glucose requires substantial amounts of cellulolytic enzymes due to the presence of strong β (1 → 4) glycosidic linkages and the crystalline structure of cellulose, which makes cellulose relatively inaccessible to cellulolytic enzymes. Saccharification of cellulose mainly consists of two reactions: (1) hydrolysis of cellulose to cellobiose (and cellodextrins) by endo- and exo-glucanases and (2) saccharification of cellobiose to glucose by β-glucosidase (Kumar et al. 2008). Because the main intermediates of cellulose hydrolysis (cellobiose and cellodextrins) strongly inhibit cellulolytic enzymes, rapid degradation of cellobiose into glucose is essential for efficient hydrolysis of cellulose. However, most cellulase mixtures obtained through cellulolytic fungal cultures exhibit low β-glucosidase activity as compared with their endo and exo-glucanases activity. Thus, supplementation with additional β-glucosidase is often necessary during the saccharification of cellulose into glucose. This may significantly increase the costs of the process, as enzymes constitute a considerable portion (40–50 %) of total costs for ethanol production from cellulose (Haki and Rakshit 2003; Lynd et al. 2002; Shen et al. 2008).
In order to overcome the feedback inhibition by cellobiose on cellulases and relieve high enzyme cost, S. cerevisiae has been engineered to ferment cellobiose directly through the introduction of two heterologous genes that allow transport of cellobiose (and cellodextrin) and intracellular hydrolysis of cellobiose (and cellodextrin) to glucose (Galazka et al. 2010). Direct cellobiose fermentation via intracellular transport and hydrolysis will lead to elimination of cellobiose in the cultures and efficient cellulose degradation without requiring β-glucosidase supplementation. Previously, several genes coding for putative cellobiose transporters including cellodextrin transporters (cdt-1 and cdt-2) from Neurospora crassa (Galazka et al. 2010), lactose transporters (LAC1, LAC2, and LAC3) from Kluyveromyces lactis, and putative hexose transporters (HXT2.1, HXT2.3, HXT2.4, HXT2.5, and HXT2.6) from Pichia stipitis (Ha et al. 2013) were identified. However, the exact mechanisms of cellobiose transport by those transporters have not been fully studied despite these transporters having exhibited diverse patterns of cellobiose utilization when expressed in S. cerevisiae by pairing with intracellular β-glucosidase (GH1-1) from N. crassa. In particular, different patterns of cell growth on cellobiose condition (including cellotriose and cellotetraose conditions) were observed between engineered S. cerevisiae expressing CDT-1 and CDT-2 from N. crassa, although both transporters were reported to have very similar affinity to cellobiose (Galazka et al. 2010). As detailed biochemical characteristics of CDT-1 and CDT-2 have not been reported yet, this study was undertaken to investigate characteristics of engineered S. cerevisiae expressing CDT-1 or CDT-2 with GH1-1 for cellobiose fermentation. The mechanism of cellobiose transport by CDT-1 and CDT-2 was investigated and profiles of sugar utilization by engineered S. cerevisiae during cellobiose fermentation were also investigated.
Materials and methods
Strains, plasmids, and culture conditions
The list of plasmids and S. cerevisiae strains used in this study
Plasmids and strains
Sources and references
LEU2, PPGK -MCS- TCYC, 2 μ origin, Ampr
Galazka et al. (2010)
URA3, PPGK -MCS- TCYC, 2 μ origin, Ampr
Galazka et al. (2010)
LEU2, PPGK-gh1-1-TCYC, 2 μ origin, Ampr
Galazka et al. (2010)
URA3, PPGK-cdt-2-TCYC, 2 μ origin, Ampr
Galazka et al. (2010)
URA3, PPGK-cdt-1-TCYC, 2 μ origin, Ampr
Galazka et al. (2010)
MATa ura3-52 lys2-801_amber ade2-101_ochre trp1-Δhis3-Δ200 leu2-Δ1
Sikorski and Hieter (1989)
MATα leu2 his3 ura3 can1
Hosaka et al. (1992)
Cellobiose transport assay
To determine the sensitivity of cellobiose transport by CDT-1 or CDT-2 to the strength of the plasma membrane proton gradient, strains expressing cdt-1 or cdt-2 were grown to OD600 of 1.5 to 3.0 in selective media, washed three times with assay buffer (30 mM MES (2-(N-morpholino) ethanesulfonic acid)-NaOH [pH 5.6] and 50 mM ethanol), and resuspended to an OD600 of 20 (Arendt et al. 2007). Then, 350 μL aliquots of cells were held on ice. To assay the transport rate at a given concentration of carbonyl cyanide m-chlorophenyl hydrazine (CCCP), 0.35 μL of a 1,000× CCCP stock made in 100 % ethanol was added to a single 350 μL aliquot of cells. This aliquot of cells was then incubated for 10 min in a room temperature water bath. After incubation, a transport assay was performed by adding 50 μL of cells to 50 μL of buffered [3H]-cellobiose with an appropriate concentration of CCCP that was layered above 100 μL of silicone oil. The same process was repeated sequentially for each tested concentration of CCCP. Controls with no CCCP were incubated for 10 min with 0.35 μL of 100 % ethanol.
A single colony after transformation was picked and grown in YP medium (10 g/L yeast extract and 20 g/L peptone) with 20 g/L of cellobiose at 200 rpm and 30 °C to prepare inocula (L) for cellobiose fermentation experiments. Yeast cells at their mid-exponential phase were harvested by centrifugation and washed with sterilized water before inoculations. Fermentation experiments were performed in YP (10 g/L yeast extract and 20 g/L peptone) or SC (6.7 g/L of yeast nitrogen base, 0.42 g/L of Tween 80, and 0.01 g/L of ergosterol with CSM or without CSM, pH 6.0) media in an oxygen-limited condition or anoxic condition. Appropriate concentrations of cellobiose (20, 40, or 80 g/L) were used. Oxygen-limited fermentation experiments were performed in 25 mL culture using a 125 mL Erlenmeyer flask tightly covered by aluminum foil at 30 °C and an agitation speed of 100 rpm. Anaerobic fermentation was done by 20 mL culture using a 100-mL serum vial sealed with a rubber cap and nitrogen purged for 5 min with gentle swirling. Vials were incubated in 30 °C at 100 rpm as well. Initial cell densities of all fermentation experiments were adjusted to OD600 of 1. All fermentation experiments were performed in duplicates.
Cell growth was monitored by measuring optical density at 600 nm using a UV-visible spectrophotometer (Biomate5, Thermo, Rochester, NY). Cellobiose, glycerol, acetate, ethanol, and cellodextrin concentrations were determined by high performance liquid chromatography (Agilent Technologies 1200 Series, Mississauga, CA) equipped with a refractive index detector using a Rezex ROA-Organic Acid H+ (8 %) column (Phenomenex Inc., Torrance, CA). The column was eluted with 0.005 N of H2SO4 at a flow rate of 0.6 ml/min at 50 °C.
Determination of a mode of transportation through cellobiose transport assay
As observed in previous studies (Galazka et al. 2010), cellodextrin transporters in N. crassa (CDT-1 and CDT-2) with high affinity to cellobiose were selected for further characterization in S. cerevisiae. When the cellodextrin transporters were expressed with GH1-1 in S. cerevisiae strains, the resulting engineered strains were able to utilize and grow with cellodextrin as the sole carbon source (Galazka et al. 2010; Ha et al. 2011).
Comparison of cellobiose fermentation characteristics of engineered S. cerevisiae expressing CDT-1 and CDT-2
CDT-1 and CDT-2 from N. crassa were further studied for the performance of cellobiose fermentation when they are co-expressed with GH1-1 in S. cerevisiae. Two strains (DCDT-1G and DCDT-2G) overexpressing gh1-1 and cdt-1 or cdt-2 under the control of a strong constitutive promoter were developed. The resulting strains were then evaluated for their fermentation capability using cellobiose as a sole carbon source under various fermentation conditions.
Cellobiose fermentation under oxygen-limited conditions
Fermentation parameters of DCDT-1G and DCDT-2G under various media conditions and in oxygen-limited or anaerobic conditions
0.42 ± 0.00
0.46 ± 000
0.73 ± 0.02
0.31 ± 0.00
0.14 ± 0.00
0.17 ± 0.01
0.39 ± 0.01
0.53 ± 0.01
0.39 ± 0.55
0.19 ± 0.00
0.11 ± 0.00
15.3 ± 0.65
SCC20 w/ CSMe
0.39 ± 0.00
0.13 ± 0.00
0.63 ± 0.00
0.23 ± 0.02
0.34 ± 0.00
10.5 ± 0.62
SCC20 w/o CSMf
0.37 ± 0.00
0.13 ± 0.00
0.60 ± 0.00
0.23 ± 0.03
0.03 ± 0.00
12.3 ± 0.09
0.43 ± 0.01
0.35 ± 0.01
0.83 ± 0.00
0.36 ± 0.00
0.09 ± 0.00
0.56 ± 0.16
0.42 ± 0.01
0.41 ± 0.02
0.35 ± 0.50
0.28 ± 0.00
0.13 ± 0.00
11.7 ± 0.46
SCC20 w/ CSM
0.49 ± 0.01
0.17 ± 0.00
0.71 ± 0.00
0.48 ± 0.00
0.05 ± 0.00
13.1 ± 0.06
SCC20 w/o CSM
0.47 ± 0.00
0.16 ± 0.00
0.61 ± 0.00
0.45 ± 0.05
0.05 ± 0.01
12.5 ± 0.07
Because our data suggests that CDT-2 is a simple facilitator, we hypothesized that CDT-2 may work better when the cellobiose concentration is high. Therefore, we also examined cellobiose fermentation rates using 80 g/L of cellobiose (Fig. 2b). However, similar fermentation profiles were observed as compared when 40 g/L of cellobiose was used. The DCDT-1G still showed higher cellobiose consumption rate and produced ethanol with a higher yield than the DCDT-2G strain. The DCDT-1G consumed all cellobiose within 36 h and produced ethanol with a yield of 0.39 g ethanol/g cellobiose (Table 2) whereas the DCDT-2G failed to utilize all of the cellobiose. The DCDT-2G strains consumed only 60 g/L of cellobiose in 62 h and stopped using cellobiose when 20 g/L cellobiose and 25 g/L cellodextrin were left in the medium. Also, less biomass formation and ethanol production were observed in DCDT-2G strain.
Cellobiose fermentation under anaerobic conditions
Taken together, the better cellobiose fermentation capability of the DCDT-1G strain, when compared to the DCDT-2G strain, was observed regardless of the fermentation conditions, resulting in both a higher yield and productivity for ethanol (Table 2). While CDT-2 may have energetic benefits, the expression levels and kinetic properties of CDT-2 in S. cerevisiae might not be optimum for cellobiose fermentation. These results suggest CDT-1 is a more effective cellobiose transporter than CDT-2 for engineering S. cerevisiae to ferment cellobiose.
Traditionally, additional β-glucosidase prepared by extra fungal or bacterial fermentation has been added to allow efficient hydrolysis of cellulose to glucose. However, this may increase the enzyme cost and risk of contamination by bacteria during the saccharification step. To overcome this problem, fungal cellobiose transporters and intracellular β-glucosidase were expressed in S. cerevisiae for direct fermentation of cellodextrins (Galazka et al. 2010). However, the detailed mechanisms and characteristics of these cellobiose transporters have not been investigated systematically. Among the cellobiose transporters discovered previously, CDT-1 or CDT-2 along with GH1-1 expressing strains were able to grow on cellobiose as a sole carbon source and both transporters had a similar affinity (Km ≈ 3–4 μM) for cellobiose. However, the rates of cellobiose consumption were very different (Galazka et al. 2010). Hence, this study was undertaken to verify the difference between CDT-1 and CDT-2 in terms of energy (proton gradient) dependence and cellobiose fermentation characteristics under various cultivation conditions when they are overexpressed in S. cerevisiae.
Generally, transport of a molecule across the plasma membrane can be mediated by two types of transporters: passive and active. Passive transport by a uniporter is driven only by the concentration gradient of the substrate, while active transport by a symporter (or antiporter) is also driven by the concentration of a second substrate. This second substrate is maintained out of equilibrium by the cell through an energy consuming process (Cussler et al. 1989). Some native transporters present in yeast, such as glucose transporters, transport glucose via facilitated diffusion (Postma et al. 1989). On the other hand, yeast disaccharide transporters such as maltose and sucrose transporters transport substrates via active transport. Specifically, the maltose and sucrose transporters require the presence of a proton gradient to move their substrates across the membrane (proton symporter) (Santos et al. 1982; Weusthuis et al. 1993).
From the results of cellobiose transport assay, we determined that CDT-1 is a proton symporter and CDT-2 is a uniporter. This result suggests that the CDT-1 expressing strain may require more energy than the CDT-2 expressing strain during cellobiose fermentation. Consequently, we hypothesized that this difference in energy requirement will affect fermentation rates of these engineered strains.
As YP media is composed of rich nutrients that can allow generation of excess energy, we cannot excluded that the considerable difference in fermentation rates between the two strains may be because CDT-1 was reported to have a twofold higher VMax than CDT-2 in a previous kinetic study (Galazka et al. 2010). Indeed, a faster fermentation rate was observed by DCDT-1G in YP medium with 40 g/L (or 80 g/L) cellobiose under oxygen-limited conditions (Fig. 2). In addition, we anticipated that the DCDT-2G might show an increased fermentation rate with an increased initial concentration of cellobiose (40 to 80 g/L); however, slower cellobiose consumption, less production of ethanol, and much higher accumulation of cellodextrin were observed in the DCDT-2G fermentation with 80 g/L of cellobiose (Fig. 2). We reason that increased levels of intracellular cellobiose leads to transglycosylation of cellobiose in cellodextrins by GH1-1 rather than degradation to glucose. The transglycosylation reaction of GH1-1 occurs under kinetically controlled conditions; therefore, an increased level of intracellular cellobiose will favor formation of cellodextrin accumulation intracellularly (Bhat et al. 1993; Park et al. 2005), along with some secretion via the transporters. It is also possible that, although GH1-1 was expressed intracellularly, a very small amount of GH1-1 might be secreted out via nonconventional protein secretion pathway (Nombela et al. 2006; Giuliani et al. 2011). A previous study on fungal β-glucosidase suggests a small amount of GH1-1 can have transglycosylation activity with high concentration of the substrate (Bohlin et al. 2013). Thus, a small amount of GH1-1 secretion could result in significant cellodextrin accumulation. This explains why DCDT-1G accumulated less cellodextrin than DCDT-2G. Because CDT-1 was reported to have higher VMax for cellodextrin (cellotriose and cellotetraose) than CDT-2 (Galazka et al. 2010), DCDT-1G can re-assimilate cellodextrin faster than DCDT-2G.
We speculated that a similar or even better fermentation rate by the DCDT-2G rather than the DCDT-1G strain would occur if the fermentation was performed with minimal medium because nutrients for energy generation can be limited in minimal media conditions. However, the DCDT-2G strain showed the same pattern of fermentation profile as in YP media (Fig. 3), and the DCDT-1G strain was still more efficient for fermenting cellobiose than the DCDT2-G strain. Under minimal media condition, there was negligible amount cellodextrin accumulation (<1 g/L; Fig. 3). This result is consistent with the previous study on cellodextrin transporters (Galazka et al. 2010) where no secretion of GH1-1 was confirmed under minimal media. It is possible that the leakage of GH1-1 occurs under certain culture conditions, such as complex media. Regardless of the cellodextrin accumulation mechanism, CDT-1 is more effective in transporting cellobiose in the presence of oxygen regardless of the nutrient availability.
In addition to nutrient availability, the presence of oxygen was also expected to influence the fermentation profile. Under the anaerobic condition, the TCA cycle will be fully inactivated and ATP generation is limited, and we reasoned that CDT-1 expressing cells may not produce enough energy to maintain active transport of cellobiose (Spiro and Guest 1991). However, overall, there was no significant difference in the fermentation profile pattern of the two strains under anaerobic conditions. Again, the CDT-1 expressing strain showed faster cellobiose fermentation as compared to the CDT-2 expressing strain. These results show that CDT-1 mediates more efficient cellobiose transport, which in turn results in a higher yield and productivity than CDT-2 regardless of nutrient and oxygen availability. As the previous study indicates, both CDT-1 and CDT-2 had a high affinity to cellobiose, but CDT-1 had a higher VMax. Superior cellobiose fermentation by strains expressing CDT-1 may be due to the VMax difference in CDT-1 and CDT-2.
In conclusion, cellodextrin transporters (cdt-1 and cdt-2) and intracellular ß-glucosidase (gh1-1) from N. crassa were introduced into S. cerevisiae. By introducing each cdt-1 and cdt-2 paired with ß-glucosidase, both engineered strains (DCDT-1G and DCDT-2G) were able to utilize cellobiose as a sole carbon source and produce ethanol. The DCDT-1G strain showed a more efficient cellobiose fermentation as compared to the DCDT-2G strain regardless of oxygen levels and media conditions. This is likely to due to the VMax difference between CDT-1 and CDT-2. It will be interesting to determine if the transport rate of CDT-2 may be improved through various engineering techniques and if this improvement may subsequently provide the dual advantages of higher energy levels and better fermentation parameters.
This work was supported by funding from the Energy Biosciences Institute (EBI) to Yong-Su Jin and Jamie Cate.