A new diet for yeast to improve biofuel production

Abstract

In 2010, our group announced the discovery of two cellodextrin transporter families from the cellulolytic fungus Neurospora crassa. Furthermore, we demonstrated the utility of these transporters in the production of lignocellulosic biofuels. This discovery was made possible by a decision to systematically study cell wall degradation by N. crassa. The identified transport pathway has opened up a new way of thinking about microbial fermentation of hexoses as well as pentoses derived from plant cell walls. Integrating this pathway with the endogenous metabolism and signaling networks of S. cerevisiae is now a major goal of our group.

Developing methods to convert the sugar-rich cell walls of plants to fuel is a major societal goal,¹ and has become the focus of intense research.² Present methods for conversion are limited by the evolved recalcitrance of plant cell walls to depolymerization into constituent sugars, and deficiencies in the traits of microbes used to ferment these sugars to fuels such as ethanol.³ It is our philosophy that fundamental breakthroughs addressing both of these limitations will be made through the study of model organisms. Hence, we study the mechanisms by which the filamentous fungus, Neurospora crassa, degrades and metabolizes plant cell walls. N. crassa has been studied since the 1920s,⁴ and features a wealth of resources including a high quality genome,⁵ facile methods for genetic manipulation,⁶ an almost complete set of gene knockouts,⁷ and an international community of researchers. In the wild, N. crassa is commonly found growing upon recently burned plant matter,⁸ and in 1977 was reported to secrete cellulases in response to specific inducers.⁹ However, until 2009 N. crassa was largely ignored as a model biomass degrader; research focused instead on the related fungus, Trichoderma reesei (Hypocrea jecorina). T. reesei is now favored by industry for its ability to secrete large quantities of cellulases used to deconstruct plant biomass.¹⁰

Biomass degradation by N. crassa was revisited in 2009, with the report of a systematic study of plant cell wall degradation by this fungus.¹¹ Transcriptomic, proteomic and phenotypic data showed: (1) that the core cellulases of N. crassa are similar to those of T. reesei; (2) that N. crassa transcriptionally upregulates ∼2% of genes in response to pure cellulose or plant cell walls; (3) that there is a large amount of functional redundancy to the cellulase system of N. crassa, as only a cellobiohydrolase(I) deletion strain had a significant deficiency in its capacity to degrade pure cellulose. Interestingly, the set of cellulose-induced genes identified in this study included 10 major facilitator superfamily transporters. The strong induction of these transporters during biomass degradation led to a hypothesis that they have a key role in this process. In addition to the transporters, we noticed that N. crassa induced the expression of a putative intracellular β-glucosidase, an enzyme that cleaves the β (1 → 4) glycosidic bond in short cellodextrin chains. Cellodextrins are β (1 → 4) linked oligosaccharides of glucose, and are the product of cellulose depolymerization by fungal cellulases.¹² We therefore reasoned that a likely substrate for some of these transporters were cellodextrins.

In 2010, we showed that two transporters, CDT-1 and CDT-2, indeed mediate the uptake of cellodextrins by N. crassa, and that following uptake cellodextrins are hydrolyzed intracellularly to glucose by a β-glucosidase.¹³ A number of transcriptomic studies have shown that orthologs of these transporters are transcriptionally upregulated in response to plant cell wall material or cellobiose in diverse fungi, suggesting they are fundamental to the strategies used by fungi to interact with plants.^14–16 The identification of these genes confirmed reports of cellodextrin permeases¹⁷ and intracellular β-glucosidases¹⁸ in T. reesei.

The importance of these transporters to biomass degradation by N. crassa and other fungi inspired us to engineer the transport pathway into the yeast, Saccharomyces cerevisiae. S. cerevisiae is the current organism of choice for fermenting glucose and sucrose to ethanol, but lacks the capacity to ferment cellodextrins.¹⁹ Strains engineered to express cdt-1 or cdt-2 along with an intracellular β-glucosidase (gh1-1) grew with cellodextrins as a sole carbon source, and fermented cellobiose to ethanol efficiently. Multiple reports preceded ours in describing cellodextrin consumption or fermentation by engineered strains of S. cerevisiae. As early as 1986, Kohchi and Toh-e reported the consumption of cellobiose by S. cerevisiae expressing a secreted β-glucosidase from Candida pelliculosa.²⁰ They did not measure its fermentative capacity, however. Then, in 1988 Machida et al. reported cellobiose fermentation by S. cerevisiae expressing secreted β-glucosidases from Saccharomycopsis fibuligera.²¹ Subsequent studies have repeated this feat both in lab^22–29 and industrial strains of S. cerevisiae,^30,31 and in other yeast species.³² All of these strains rely on extracellular hydrolysis of cellobiose to glucose by secreted β-glucosidases followed by uptake via the endogenous S. cerevisiae hexose transporters. In contrast, the first step in the pathway we described is the import of cellodextrins into the cytosol of S. cerevisiae, only then does hydrolysis to glucose occur. Other yeasts in nature have been shown to ferment cellobiose, but the molecular basis for this capability was not determined.^33–36

There are two main distinctions between cellodextrin fermentation mediated by transport and one mediated by extracellular hydrolysis to glucose: (1) cellodextrin transporters have a relatively high affinity for cellodextrins; (2) cellodextrins are hydrolyzed intracellularly. Both may be important during the fermentation of plant cell wall derived sugars. The high-affinity of CDT-1 and CDT-2 for cellodextrins could be particularly important during the simultaneous saccharification and fermentation (SSF) of plant cell walls to fuel. SSF, in which fermenting microbes are included in the depolymerization reaction, increases the efficiency of conversion by relieving product inhibition upon cellulases.³⁷ With a K_M of ∼5 µM, cellodextrin transporters may be capable of reducing sugar concentrations below levels obtainable by the inclusion of extracellular β-glucosidases, which generally have a K_M of 100–1,000 µM.³⁸ This idea has not been rigorously tested and further studies applying and validating models of SSF using cellodextrin-transporting strains are necessary.^39,40

Relocating hydrolysis of cellodextrins from outside to inside the cell may seem like a trivial change, but it has a profound advantage. Specifically, transport of cellodextrins followed by intracellular hydrolysis to glucose facilitates the cofermentation of cellulose-derived glucose and hemicellulose-derived xylose.^41,42 The inability of S. cerevisiae to coferment glucose and xylose is a barrier to efficient lignocellulosic biofuel production. Both glucose and xylose enter S. cerevisiae through the endogenous S. cerevisiae hexose transporters, but glucose is vastly preferred to xylose.^43,44 Thus, xylose will not be taken up until all glucose is consumed. This sequential fermentation of glucose and xylose leads to lower ethanol yields and productivities.⁴⁵ Cellodextrin transport represents an alternate route for glucose to enter yeast, leaving the endogenous hexose transporters unoccupied and capable of transporting xylose. Strains containing the cellodextrin transport pathway from N. crassa and an efficient xylose fermentation pathway coferment cellobiose and xylose at high rates and with high yields.^41,42

Intracellular hydrolysis of cellodextrins does have associated problems, however. Yeast set metabolic and growth rates to match the perceived abundance of nutrients.^46,47 Normally, perceived and actual nutrient availabilities match, and metabolism is optimal. This may no longer be the case in strains engineered to hydrolyze cellodextrins intracellularly. S. cerevisiae detects the presence of extracellular glucose via three sensors: the 7-transmembrane receptor, Gpr1; and two non-transporting transceptors, Rgt2 and Snf3 (Fig. 1). The associated signaling pathways induce a robust transcriptional response to glucose, placing S. cerevisiae in an optimal state for high-capacity fermentation.⁴⁸ In strains engineered to hydrolyze cellodextrins intracellularly, glucose is never present at high concentrations outside of the cell (though some glucose may leak out through the endogenous S. cerevisiae hexose transporters). Therefore, the presence of glucose in the form of cellodextrins will not be signaled through Snf3, Rgt2 & Gpr1, placing the cell in a sub-optimal state. Additional signals emanate from intracellular glucose, and presumably these pathways will still respond following hydrolysis of transported cellodextrins as is seen following the transport and hydrolysis of maltose.⁴⁹ Presently, we do not know if strains engineered to transport and hydrolyze cellodextrins respond appropriately to the presence of glucose in the form of cellodextrins. This will be an important area of future experimentation, the results of which may necessitate the integration of cellodextrin metabolism into the signaling network of S. cerevisiae.

Figure 1.

Model of the glucose sensing network of S. cerevisiae in the context of a cellodextrin transport pathway. Glucose is sensed by both extracellular and intracellular systems. Extracellular glucose is sensed by the 7-transmembrane protein, Gpr1, as well as the non-transporting transceptors, Snf3 and Rgt2. Intracellular glucose, which can enter through the hexose transport system (Hxt), is sensed by a pathway that involves Hxk2. Signals emanating from Gpr1 and intracellular glucose converge to induce a genome-wide transcriptional response through the Ras/PKA pathway, which includes adenylate cyclase (Cyr1) and protein kinase A (Tpk1, Tpk2, Tpk3). Signals emanating from Snf3 and Rgt2 modulate the expression of various hexose transporters through Std1 and Mth1. A cellodextrin transport pathway would likely bypass Gpr1, Snf3 or Rgt2, but may activate the Ras/PKA pathway following transport and hydrolysis of cellodextrins. This figure is a simplified model adapted from Rolland et al.⁵⁰

The identification of a cellodextrin transport pathway in N. crassa has offered insights into the strategies used by filamentous fungi to degrade plant cell walls, and the successful recapitulation of this pathway in S. cerevisiae has provided new routes for the microbial synthesis of lignocellulosic fuels. Many questions must be answered before this pathway is fully integrated with the endogenous metabolism of S. cerevisiae, and additional development will be necessary to produce a strain useful in industrial processes.