Cellodextrin Transport in Yeast for Improved Biofuel Production

See allHide authors and affiliations

Science  01 Oct 2010:
Vol. 330, Issue 6000, pp. 84-86
DOI: 10.1126/science.1192838


Fungal degradation of plant biomass may provide insights for improving cellulosic biofuel production. We show that the model cellulolytic fungus Neurospora crassa relies on a high-affinity cellodextrin transport system for rapid growth on cellulose. Reconstitution of the N. crassa cellodextrin transport system in Saccharomyces cerevisiae promotes efficient growth of this yeast on cellodextrins. In simultaneous saccharification and fermentation experiments, the engineered yeast strains more rapidly convert cellulose to ethanol when compared with yeast lacking this system.

The bioethanol industry uses the yeast Saccharomyces cerevisiae to ferment sugars derived from cornstarch or sugarcane into ethanol (1). Plant cell walls provide an abundant alternate sugar source yet remain largely unused (2, 3). Present strategies for using plant biomass use large quantities of cellulase cocktails to release glucose from plant cell walls, posing economic and logistical challenges for implementation of these processes (4). Furthermore, S. cerevisiae cannot ferment the cellodextrins naturally released by cellulases (5, 6) and require cellulase cocktails supplemented by β-glucosidases to quantitatively produce fermentable glucose (7).

In contrast to S. cerevisiae, cellulolytic fungi grow well on cellodextrins. The model cellulolytic fungus Neurospora crassa, when grown on pure cellulose, increases transcription of seven major facilitator superfamily (MFS) sugar transporters as well as an intracellular β-glucosidase (8). A strain carrying a deletion for one of these predicted transporters, NCU08114, grew slowly on cellulose (fig. S1) (9), and strains lacking either NCU08114 or NCU00801 poorly consumed cellobiose, cellotriose, and cellotetraose (fig. S2). Orthologs of NCU08114 and NCU00801 are widely distributed in the fungal kingdom (Fig. 1A) (1013), and recent whole-genome transcriptional profiling studies show their importance to interactions between fungi and plants. For example, some cellulolytic fungi increase expression levels of NCU08114 orthologs while degrading plant wall material (10), and the Périgord black truffle increases the expression of a NCU00801 ortholog during symbiotic interactions with plant roots (11). In addition, certain yeasts that grow on cellobiose contain orthologs of NCU08114 and NCU00801 (12). All of these organisms also contain genes for intracellular β-glucosidases (10), suggesting that cellodextrin transport systems are widespread in nature and are essential for optimal growth of fungi on cellulose-derived sugars.

Fig. 1

Cellodextrin transport by N. crassa transport systems expressed in S. cerevisiae. (A) Maximum likelihood phylogenetic analysis of the cellobiose transporters NCU08114 and NCU00801. With the exception of S. cerevisiae HXT1 and K. lactis LACP (12), all genes encoding proteins shown are reported to increase in expression level when the fungus comes into contact with plant cell wall material or cellobiose (8, 10, 11, 13). S. cerevisiae HXT1, a low-affinity glucose transporter (19), was used as an outgroup. Accession numbers for each gene are listed in (9). (B) Cellobiose-mediated growth of yeast strains expressing the gene NCU00801 (named cdt-1, open circle), NCU08114 (named cdt-2, solid triangle), or no transporter (solid circle). All strains also express the intracellular β-glucosidase, NCU00130 (named gh1-1). A representative experiment is shown. Growth rates from three independent experiments are as follows: cdt-1, 0.0341 ± 0.0010 hours−1; cdt-2, 0.0131 ± 0.0008 hours−1; and no transporter, 0.0026 ± 0.0001 hours−1 (mean ± SD). (C) Growth of yeast strains on cellotriose and cellotetraose. Strains expressing the intracellular β-glucosidase, gh1-1, as well as the transporters listed in the legend, were grown with 0.5% (w/v) of cellotriose (G3) or cellotetraose (G4) serving as the sole carbon source. A representative experiment is shown. Growth rates from three independent experiments are as follows: cdt-1 cellotriose, 0.0332 ± 0.0004 hours−1; cdt-1 cellotetraose, 0.0263 ± 0.0020 hours−1; no transporter cellotriose, 0.0043 ± 0.0015 hours−1; cdt-2 cellotriose, 0.0178 ± 0.0005 hours−1; cdt-2 cellotetraose 0.0041 ± 0.0003 hours−1; and no transporter cellotetraose, 0.0031 ± 0.0008 hours−1 (mean ± SD). The data for yeast growth on cellotetraose with cdt-2 overlaps the data in the absence of transporter (asterisk).

Because cellobiose is not catabolized by S. cerevisiae (5, 6) and is not accumulated in the cytoplasm (fig. S3), we reasoned that expression of a functional cellodextrin transport system from N. crassa might allow S. cerevisiae to grow with cellobiose as the sole carbon source. Yeast strains expressing NCU00801 or NCU08114, together with the intracellular β-glucosidase NCU00130 (hereafter named GH1-1), grew on cellobiose at rates of about 30 and 12% of the rate of S. cerevisiae on glucose, respectively (Fig. 1B and fig. S4). Cellodextrins longer than cellobiose also support the growth of yeast expressing these transporters (Fig. 1C), indicating that cellodextrins longer than cellobiose are transported by NCU00801 and NCU08114, hereafter called CDT-1 and CDT-2, respectively. Growth cannot be explained by the extracellular hydrolysis of cellodextrins to glucose followed by transport because a strain expressing only the intracellular β-glucosidase GH1-1 grew slowly on cellodextrins (Fig. 1, B and C, and fig. S4). Furthermore, the activity of β-glucosidase GH1-1—which is able to hydrolyze cellobiose, cellotriose, and cellotetraose (fig. S5)—is negligible in culture supernatants and cannot explain the growth of yeast expressing the transporters (fig. S6).

To directly assay transporter function, the uptake of [3H]-cellobiose into yeast cells was measured. Both CDT-1 and CDT-2 are high-affinity cellobiose transporters, with Michaelis constant (Km) values of 4.0 ± 0.3 μM and 3.2 ± 0.2 μM (mean ± SD), respectively (fig. S7). The expression-normalized maximum velocity (Vmax) of CDT-1 is over twice that of CDT-2 (fig. S7), which might explain differences in the observed growth rates of the yeast strains (Fig. 1, B and C). Cellobiose transport by CDT-1 and CDT-2 is inhibited by excess cellotriose, and CDT-1 activity is also inhibited by cellotetraose (fig. S8). These data further support that CDT-1 and CDT-2 directly transport the longer cellodextrins in growth assays (Fig. 1C). Thus, the combinations of CDT-1 or CDT-2 with GH1-1 constitute fully functional cellodextrin transport systems.

We next tested whether, in addition to growth, a complete cellodextrin catabolism pathway would be useful in S. cerevisiae for lignocellulosic biofuel production. Fungal cellulases evolved to function in conjunction with the consumption of sugars released from plant cell walls and work most effectively in this context (14). Simultaneous saccharification and fermentation (SSF) mimics this natural synergy by the addition of fermenting microbes to the plant biomass depolymerization reaction, resulting in a more efficient process (15, 16). Current SSF schemes are limited by the need for quantitative conversion of cellulose to glucose outside yeast cells (7). A cellodextrin transport system could be particularly useful during SSF because it would remove the requirement for full hydrolysis of cellulose to glucose by extracellular β-glucosidases (17) and would reduce the risk of contamination by glucose-dependent organisms (18). Furthermore, both CDT-1 and CDT-2 have a higher apparent affinity for cellobiose (Km ≈ 3 to 4 μM) (fig. S7) when compared with secreted fungal β-glucosidases [Km ≈ 100 to 1000 μM (17)] and to S. cerevisiae hexose transporters’ apparent affinity for glucose [Km ≈ 1000 to 10,000 μM (19)]. Therefore, the cellodextrin transport systems should more effectively maintain soluble sugar levels below the concentration at which they inhibit fungal cellulases [inhibition constant (Ki) of cellobiose ≈ 19 to 410 μM (20)].

With little optimization, yeast expressing cdt-1 and gh1-1 fermented cellobiose with an ethanol yield of 0.441 ± 0.001 (grams of ethanol/grams of glucose ± SD), which is 86.3% of the theoretical value (Fig. 2A) (21). This yield is close to present industrial yields of ethanol from glucose of 90 to 93% (22). Yeasts expressing a cellodextrin transport system markedly improve the efficiency of SSF reactions by reducing the steady-state concentration of both cellobiose and glucose and by increasing the ethanol production rate (Fig. 2, B and C). The addition of a cellodextrin transport system to biofuel-producing strains of yeast (Fig. 3) overcomes a major bottleneck to fermentation of lignocellulosic feedstocks and probably will help to make cellulosic biofuels economically viable.

Fig. 2

Cellobiose fermentation, and simultaneous saccharification and fermentation of cellulose, by S. cerevisiae expressing the cellobiose transport system from N. crassa. (A) Cellobiose fermentation to ethanol. Ethanol was produced by yeast strains with CDT-1 (solid circle) or without CDT-1 (open circle) under anaerobic conditions. Also shown are cellobiose concentrations during the fermentation reaction using yeast strains with CDT-1 (solid triangle), or without CDT-1 (open triangle). (B) SSF using yeast strains with and without CDT-1, under anaerobic conditions. Shown are cellobiose (solid circle) and glucose (solid triangle) concentrations in the presence of a strain with CDT-1, and cellobiose (open circle) and glucose (open triangle) concentrations in the presence of a strain lacking CDT-1. 0.1 mg/ml cellobiose = 292 μM. (C) Ethanol produced during SSF by using a strain with CDT-1 (solid circle) or without CDT-1 (open circle). In (A) to (C), values are the mean of three biological replicates. Error bars are the SD between these replicates. All strains also express the intracellular β-glucosidase, gh1-1.

Fig. 3

Use of cellodextrin transport pathways from filamentous fungi in yeast during simultaneous saccharification and fermentation of cellulose. The cellodextrin (Cdex) transport pathway (black) includes a cellodextrin transporter (CDT) and intracellular β-glucosidase (βG). The sugar catabolism pathway present in standard yeast includes hexose transporters (HXT). In SSF, cellulases (GH) and extracellular β-glucosidase (βG) may both be used.

Supporting Online Material

Materials and Methods

Figs. S1 to S8


References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. We thank J. Doudna, M. Marletta, J. Taylor, T. Bruns, and C. Phillips for helpful discussions and comments on the manuscript; M. Toews for help with growth assays; S. Bauer and A. Ibanez for help with analytical methods; and C. Anderson for help with confocal microscopy. This work was supported by funding from the Energy Biosciences Institute to J.H.D.C. and N.L.G. The Regents of the University of California, the authors, and British Petroleum Technology Ventures (through the Energy Biosciences Institute) have submitted a patent for the use of cellodextrin transporters in fermenting organisms for the use of plant biomass.
View Abstract

Stay Connected to Science

Navigate This Article