Evidence for Transceptor Function of Cellodextrin Transporters in Neurospora crassa*

Background: Mechanisms filamentous fungi use to sense cellulose in the environment remain unclear. Results: Neurospora crassa cellodextrin transporters, CDT-1 and CDT-2, are required to induce the expression and secretion of cellulases, using a mechanism distinct from that in Trichoderma reesei. Conclusion: CDT-1 and/or CDT-2 may act as transporting receptors (transceptors). Significance: CDT-1- and/or CDT-2-mediated cellodextrin signaling may be conserved in industrially useful fungi. Neurospora crassa colonizes burnt grasslands and metabolizes both cellulose and hemicellulose from plant cell walls. When switched from a favored carbon source to cellulose, N. crassa dramatically up-regulates expression and secretion of genes encoding lignocellulolytic enzymes. However, the means by which N. crassa and other filamentous fungi sense the presence of cellulose in the environment remains unclear. Previously, we have shown that a N. crassa mutant carrying deletions of three β-glucosidase enzymes (Δ3βG) lacks β-glucosidase activity, but efficiently induces cellulase gene expression and cellulolytic activity in the presence of cellobiose as the sole carbon source. These observations indicate that cellobiose, or a modified version of cellobiose, functions as an inducer of lignocellulolytic gene expression and activity in N. crassa. Here, we show that in N. crassa, two cellodextrin transporters, CDT-1 and CDT-2, contribute to cellulose sensing. A N. crassa mutant carrying deletions for both transporters is unable to induce cellulase gene expression in response to crystalline cellulose. Furthermore, a mutant lacking genes encoding both the β-glucosidase enzymes and cellodextrin transporters (Δ3βGΔ2T) does not induce cellulase gene expression in response to cellobiose. Point mutations that severely reduce cellobiose transport by either CDT-1 or CDT-2 when expressed individually do not greatly impact cellobiose induction of cellulase gene expression. These data suggest that the N. crassa cellodextrin transporters act as “transceptors” with dual functions - cellodextrin transport and receptor signaling that results in downstream activation of cellulolytic gene expression. Similar mechanisms of transceptor activity likely occur in related ascomycetes used for industrial cellulase production.


Neurospora crassa colonizes burnt grasslands and metabolizes both cellulose and hemicellulose from plant cell walls.
When switched from a favored carbon source to cellulose, N. crassa dramatically up-regulates expression and secretion of genes encoding lignocellulolytic enzymes. However, the means by which N. crassa and other filamentous fungi sense the presence of cellulose in the environment remains unclear. Previously, we have shown that a N. crassa mutant carrying deletions of three ␤-glucosidase enzymes (⌬3␤G) lacks ␤-glucosidase activity, but efficiently induces cellulase gene expression and cellulolytic activity in the presence of cellobiose as the sole carbon source. These observations indicate that cellobiose, or a modified version of cellobiose, functions as an inducer of lignocellulolytic gene expression and activity in N. crassa. Here, we show that in N. crassa, two cellodextrin transporters, CDT-1 and CDT-2, contribute to cellulose sensing. A N. crassa mutant carrying deletions for both transporters is unable to induce cellulase gene expression in response to crystalline cellulose. Furthermore, a mutant lacking genes encoding both the ␤-glucosidase enzymes and cellodextrin transporters (⌬3␤G⌬2T) does not induce cellulase gene expression in response to cellobiose. Point mutations that severely reduce cellobiose transport by either CDT-1 or CDT-2 when expressed individually do not greatly impact cellobiose induction of cellulase gene expression. These data suggest that the N. crassa cellodextrin transporters act as "transceptors" with dual functions -cellodextrin transport and receptor signaling that results in downstream activation of cellulolytic gene expression. Similar mechanisms of transceptor activity likely occur in related ascomycetes used for industrial cellulase production.
Liquid biofuels produced from lignocellulosic biomass are an environmentally clean and renewable source of energy that could displace a significant fraction of the current demand for petroleum (1)(2)(3)(4). However, the costs associated with conversion of insoluble polysaccharides in plant cell walls to easily fermentable sugars represent significant barriers to the production of cost-competitive biofuels (4 -6). Filamentous fungi have the capacity to secrete large amounts of lignocellulosic enzymes that release fermentable sugars from plant cell walls, and this ability has been exploited by industry to produce cellulases in quantities exceeding 100 g/liter of culture (7).
The best inducers of plant cell wall-degrading enzyme expression by filamentous fungi are insoluble substrates that include cellulose, hemicellulose, or mixtures of plant polymers. Since these naturally inducing substances cannot enter fungal cells, it is generally believed that oligosaccharides released from the polymers and their derivatives function as the actual molecules that trigger enzyme induction (8). Early work in Trichoderma reesei using [U- 14 C]cellobiose confirmed the presence of a cellobiose uptake system (9). This uptake system was specific for ␤-linked diglucosides, could be inhibited by glucose and uptake was stimulated following treatment with sophorose. Despite characterization of this cellobiose uptake system, no specific transporter or transporters were described to enable the diglucoside uptake. More recently, studies in T. reesei identified four different transporters involved in cellulase production in response to a media containing lactose or cellobiose. In the first study, deletion of one transporter (Trire2: 3405) was sufficient to prevent both lactose uptake and cellulase induction in T. reesei (10). However a second study using a different T. reesei strain found that deletion of the genes encoding either Trire2:77517 or Trire2:79202 showed a decrease in both lactose uptake and cellulase induction (11). While both studies identified different transporters involved in cellulase induction via lactose, the mechanisms of cellulase induction by sophorose (another soluble inducer of cellulases in T. reesei) and by crystalline cellulose (Avicel) remained unknown. In a third study using sophorose, cellobiose, or crystalline cellulose as inducers, a dual cellobiose/glucose transporter named Stp1 (Trire2:47710) was shown to be involved in the T. reesei carbon catabolite repression response (12). Intriguingly, although dele-tion of Trire2:3405 abolished lactose uptake (10), the deletion strain still imported sophorose and cellobiose. However, deletion of Trire2:3405 (Crt1) abolished the accompanying induction of cellulases (12). These data suggest that, similar to nutrient sensing in Saccharomyces cerevisiae (13,14), T. reesei uses multiple transporters that may function in signal transduction (so called transceptors) to induce cellulase gene expression.
The filamentous fungus N. crassa, has been used as a model laboratory organism for nearly 100 years (15). In nature, N. crassa is only associated with plant material, primarily killed by exposure to fire (16 -19). N. crassa is an effective degrader of lignocellulose (20). When grown on crystalline cellulose, N. crassa increases the transcription of seven major facilitator superfamily (MFS) 3 sugar transporters (21). Two of these transporters, CDT-1 and CDT-2, were capable of transporting cellodextrins when expressed in Saccharomyces cerevisiae (22). In this study, we examine the hypothesis that CDT-1 and CDT-2 function as transporters and also are required for signaling for cellulase induction by crystalline cellulose. We show that the presence of at least one of the transporters CDT-1 or CDT-2 is required for cellulase gene induction and that deletion of both cdt-1 and cdt-2 results in a N. crassa strain that is unable to sense the presence of crystalline cellulose. Furthermore, by using strains lacking the genes encoding three ␤-glucosidase enzymes (⌬3␤G) (23), we show that additional absence of both cdt-1 and cdt-2 results in a strain of N. crassa that is unable to sense the presence of cellobiose.
Shake Flask Studies-Conidia from deletion strains were inoculated at a concentration equal to 10 6 conidia per milliliter in 50 ml of Vogel's salts (26) with 2% w/v sucrose, 2% Avicel PH 101 (Sigma), or 2% cellobiose (Sigma) in a 250 ml Erlenmeyer flask and grown under constant light at 200 rpm. Images were taken at 48, 72, and 96 h.
Transcriptional Studies-Conidia from strains were inoculated at a concentration equal to 10 6 conidia per milliliter in 50 FIGURE 1. Cellulase gene expression levels in WT and cellodextrin transporter deletion strains after induction with Avicel. A, gene expression of the cellulases cbh-1 and gh5-1 after 4 h induction with 1% Avicel in WT, ⌬cdt-1, ⌬cdt-2, and ⌬cdt-1⌬cdt-2 strains. Gene expression levels of cbh-1 and gh5-1 were normalized to 1 when induced with 1% sucrose. Expression levels of actin (NCU04173) were used as an endogenous control in all samples. Each strain was grown in triplicate, and error bars indicate 1 S.D. B, growth phenotype of WT, ⌬cdt-1, ⌬cdt-2, and ⌬cdt-1⌬cdt-2 strains when grown in sucrose or Avicel. Strains were grown for 48 h on 2% sucrose or 96 h on 2% Avicel. ml Vogel's salts (26) with 2% w/v sucrose in a 250 ml Erlenmeyer flask and grown under constant light at 200 rpm for 16 h. Biomass was then spun at 1,000 ϫ g for 10 min and washed in Vogel's salts twice to remove any excess sucrose. Biomass was then added to a new flask containing 50 ml of Vogel's salts supplemented with 1% w/v sucrose, 1% w/v Avicel PH 101 (Sigma) or cellobiose (Sigma) at the indicated concentration. Cultures were induced for 4 h under constant light at 200 rpm (23). The culture biomass was then harvested by filtration over a Whatman glass microfiber filter (GF/F) on a Büchner funnel and washed with 50 ml of Vogel's salts. The biomass was flash frozen in liquid nitrogen and stored at Ϫ80°C. Three independent biological replicates (flasks) were evaluated for each time point.
RNA Isolation-Total RNA from frozen samples was isolated using Zirconia/silica beads (0.5 mm diameter; Biospec) and a Mini-Beadbeater-96 (Biospec) with 1 ml TRIzol reagent (Invitrogen) according to the manufacturer's instructions. The total RNA was further purified by digestion with TURBO DNA-free (Ambion) and an RNeasy kit (Qiagen). RNA concentration and integrity was checked by Nanodrop and agarose gel electrophoresis, respectively.
RT-PCR-Quantitative RT-PCR was performed using the EXPRESS One-Step SYBR GreenER Kit (Invitrogen) and the StepOnePlus Real-Time PCR System (Applied Biosystems). Reactions were performed in triplicate with a total reaction volume of 10 l including 300 nM each forward and reverse primers and 75 ng of template RNA. Data analysis was performed using the StepOne Software (Applied Biosystems) with the Relative Quantitation/Comparative CT (⌬⌬CT) setting. Data were normalized to the endogenous control actin (NCU04173) with expression on sucrose as the reference sample.
Glucose and Cellobiose Concentration during Growth on Crystalline Cellulose-Conidia from the wild type strain were inoculated in duplicate at a concentration equal to 10 6 conidia per milliliter in 50 ml of Vogel's salts (26) with 2% w/v Avicel PH 101 (Sigma) in a 250 ml Erlenmeyer flask and grown under constant light at 200 rpm for 7 days. Samples were taken daily and spun at 1,000 ϫ g for 10 min to remove biomass and excess Avicel. The concentration of glucose and cellobiose were measured on a DIONEX ICS-3000 HPLC (Dionex Corporation) using a CarboPac PA20 Analytical Column (3 ϫ 150 mm) and a CarboPac PA20 guard column (3 ϫ 30 mm) at 30°C. Following injection of 25 l of diluted samples, elution was performed with 100 mM KOH (isocratic) at 0.4 ml/min. Sugars were detected using pulsed amperometric detection, Four-Potential Carbohydrate Waveform and Peaks were analyzed using the Chromeleon software package.
GFP-tagged Transporters-Template gDNA from N. crassa WT strain (FGSC 2489) was extracted according to the method of Lee et al. (51). Strains tagged with GFP were created by adapting the tools described in Honda and Selker (28). The 10xGly::GFP::hph::LoxP fragment was subcloned from FJ457011 (28) into pRS426 using XhoI (NEB) and KpnI (NEB). The upstream region was amplified by PCR (ϳ1000 bp upstream and coding region) using primers containing XhoI (NEB) sites and ligated using T4 DNA Ligase (NEB). The downstream region was then amplified by PCR (ϳ1000 bp downstream of the coding region) using primers containing KpnI (NEB) sites and ligated using T4 DNA Ligase (NEB). The final vectors were sequenced using several primers to ensure that the proper sequence was obtained. The primers used to clone the upstream region are as follows NCU08114: Mutagenesis of CDT-1 and CDT-2 and Strain Construction-Alanine mutants were produced in previously described 2 yeast expression plasmids (22). These plasmids contained the CDT-1 or CDT-2 open reading frame fused to GFP and the S. cerevisiae PGK1 promoter and CYC transcriptional terminator within the pRS426 backbone. The QuikChange (Agilent Technologies) protocol was used to introduce alanine mutations, and results were confirmed by sequencing the plasmids. All plasmids were transformed into the S. cerevisiae strain YPH449 (MATa ura3-52 lys2-801_amber ade2-101_ochre trp1-⌬63 his3-⌬200 leu2-⌬1), which contained a previously described (22) plasmid containing the GH1-1 open reading frame and the S. cerevisiae PGK1 promoter within the pRS424 backbone.
Determination of Strain Growth Rates-Colonies corresponding to independent transformants of S. cerevisiae strain YPH449 containing GH1-1 and a given alanine mutant of CDT-1 or CDT-2 were resuspended in 400 l of YNP (with ammonium sulfate) plus 2% cellobiose and CSM (-uracil, -leucine) at a starting OD (600 nm) of ϳ0.2. Growth curves were recorded using a Bioscreen C (Oy Growth Curves Ab Ltd), and growth rates determined from the exponential phase of the curve. Transformation of N. crassa Strains-1 g of plasmid DNA was transformed into the ⌬mus-52 strain (FGSC 9719) by electroporation as described (31). The constructs were targeted to the native locus by homologous recombination and integration of the constructs at the native locus in heterokaryotic transformants was confirmed by GFP fluorescence and PCR with primers designed to the region upstream and downstream of the integration cassette. To recover homokaryotic strains, and repair the mus-52 gene, transformants were crossed with the ⌬3␤G strain (23) and hygromycin B (Sigma-Aldrich) resistant/GFP-positive/PEST (Glufosinate-ammonium; Sigma-Aldrich) sensitive progenies were screened for the absence of NCU00130, NCU04952, and NCU08755 as described previously (23).
Cellobiose Uptake Assays-Conidia from each strain were inoculated at a concentration equal to 10 6 conidia per milliliter in 3 ml of Vogel's salts with 2% w/v sucrose in a 24 well plate and grown under constant light at 200 rpm for 40 h. Biomass was transferred to 5 ml of Vogel's salts containing no carbon and was incubated at 450 rpm for 5 min. Biomass was again transferred to 3 ml of Vogel's salts containing 2% w/v cellobiose (Sigma-Aldrich) and was incubated under constant light at 200 rpm for 24 h to induce transporter expression. Following induction, biomass was washed three times with Vogel's salts containing no carbon to remove any excess cellobiose and transferred to a new 24-well plate with 3 ml of cellobiose uptake assay solution (5 mM MES pH 6.0, 100 mM NaCl, 100 M cellobiose) and vortexed at 900 rpm for 10 s. 300 l samples were taken at 2.5, 10, 20, and 30 min intervals. Samples were centrifuged at 14,000 rpm at 4°C for 10 min. 200 l of each sample were transferred to an HPLC vial containing 800 l of 0.1 M NaOH. The concentration of remaining cellobiose was measured on a DIONEX ICS-3000 HPLC (Dionex Corporation) as described above. The remaining biomass was washed with water, dried for 24 h at 95°C, and weighed with a Sartorius SE2 ultra-micro analytic balance.

. Identification of non-transporting CDT-1 and CDT-2 point mutants.
A, cellobiose mediated growth rates of S. cerevisiae transformed with CDT-1 or CDT-2 bearing alanine substitutions and GH1-1. Conserved amino acids in CDT-1 and CDT-2 were separately mutated to alanine. A construct containing a mutation at one position was transformed into S. cerevisiae alone with the ␤-glucosidase, GH1-1. These strains were grown with cellobiose as the sole carbon source. The growth rates (increase in the OD at 600 nm per hours) were measured and compared with strains with a wild type CDT and GH1-1 or no transporter and GH1-1. Each sample was performed in duplicate and error bars represent the standard deviation. B, sequence logos for CDT-1 and CDT-2 were retrieved from the NCBI BLAST server with an E value cutoff of 1E Ϫ130 and aligned using the MUSCLE algorithm (47). The alignment logos were created with WebLogo (48). The size of the residue letter is proportional to its frequency of appearance at the given position. C, homology models of CDT-1 and CDT-2. Homology models of CDT-1 and CDT-2 were built by Modeler v9.12 (49) using a xylose transporter XylE structure; PDB code 4GBZ (50) as the template. The black lines indicate the transmembrane region assigned by comparing with XylE structure. Structures were colored in rainbow with N terminus in blue and C terminus in red. The mutated residues are shown as sticks. The figure was prepared with the molecular visualization software PyMol (52).
Fluorescence Microscopy of GFP-tagged Transporters-Mycelia for microscopy were prepared in the same way as biomass was prepared for cellobiose uptake assays, without the final Vogel's salts washes. Confocal microscopy was performed on a Leica SD6000 microscope equipped with a Yokogawa CSU-X1 spinning disk head and 488-nm lasers controlled by Metamorph software (Molecular Devices). All images were acquired using a 100 ϫ 1.4 NA oil immersion objective.

Induction of Cellulase Gene Transcripts Is Blocked in Cellulose-induced Cultures of N. crassa Lacking Two Cellodextrin
Transporter Genes-A previous systems analysis identified several predicted transporters that increase in expression level when N. crassa is grown on crystalline cellulose (21). When expressed in S. cerevisiae, two of these transporters were capable of efficiently transporting cellooligosaccharides (22). We hypothesized that these up-regulated transporters are involved in sensing of crystalline cellulose and contribute to the induction and secretion of active cellulases. To examine this hypoth-esis, we initially tested whether the expression of two major cellulase genes (cbh-1, NCU07340; gh5-1, NCU00762) was induced in strains carrying deletions of the eight previously identified transporters (21) ("Experimental Procedures"). Because strains carrying single deletions were unable to produce a significant defect in induction (Fig. 1A), a mutant strain was constructed that lacked both cdt-1 and cdt-2. Unlike the single mutants, this strain (⌬cdt-1 ⌬cdt-2) showed a significant defect in the induction of both cbh-1 and gh5-1 following a 4-h transfer of N. crassa to 1% w/v Avicel (Fig. 1A). In addition to the induction phenotype, we also examined the growth phenotype of strains carrying both single deletions and double deletions of cdt-1 and/or cdt-2 on both sucrose and Avicel as compared with wild type (Fig. 1B). All strains were able to grow equally well on sucrose. However, a minor growth defect was seen in strain lacking cdt-2 after 96 h growth on 2% w/v Avicel. A severe growth defect was observed when a strain lacking both cdt-1 and cdt-2 was cultivated on Avicel. These data indicate that cdt-1 and cdt-2 are involved in both the induction of cellulases as well as the growth of N. crassa on crystalline cellulose. Deletion of both genes was sufficient to prevent N. crassa from sensing the presence of crystalline cellulose in its environment.

Deletion of cdt-1 and cdt-2 in a Strain Lacking Genes Encoding Three ␤-Glucosidase Enzymes (⌬3␤G) Results in a Strain
Unable to Respond to Cellobiose-Previously, we showed that a N. crassa strain lacking genes encoding three ␤-glucosidase enzymes (two extracellular and one intracellular; ⌬3␤G) responds to cellobiose in the same way that wild type N. crassa responds to crystalline cellulose (23). We therefore hypothesized that a strain lacking the genes encoding the three ␤-glucosidase enzymes as well as cdt-1 and cdt-2 (⌬3␤G⌬cdt-1⌬cdt-2) would be unable to induce cellulases in response to a physiologically relevant concentration of cellobiose. To examine this hypothesis, we first defined the physiologically relevant concentration of cellobiose required for cellulase induction in N. crassa.
We first measured the concentration of soluble glucose and cellobiose from cultures growing over 7 days on 2% w/v Avicel ( Fig. 2A). The glucose concentration reached a maximum of ϳ70 M on the second day and proceeded to decrease until day 6 when it reached its minimum concentration of ϳ20 M. On the other hand, the cellobiose concentration was much lower, reaching a maximum of about 1 M over the 7 day time course. In a second experiment, we used the ⌬3␤G background strain to examine the induction of two major cellulase genes (cbh-1, NCU07340; gh5-1, NCU00762) in response to decreasing concentrations of cellobiose. We confirmed that 0.2% w/v (5.8 mM) cellobiose produced maximal induction (23), but were also able to detect induction of cbh-1 and gh5-1 above that seen by starvation (derepression) with concentrations of cellobiose as low as 10 nM (Fig. 2B). Taken together, the results of these two experiments indicate that 1 M cellobiose should serve as a physiologically relevant concentration for examining cellulase gene induction in strains lacking cdt-1 and/or cdt-2 in the ⌬3␤G background strain.
In a 4-h transfer experiment using 1 M cellobiose, we found that strains lacking either cdt-1 or cdt-2 (in the ⌬3␤G background strain) showed an induction profile for the two major cellulase genes (cbh-1, NCU07340; gh5-1, NCU00762) similar to that observed in the ⌬3␤G background strain alone (Fig. 3A). However, when both cdt-1 and cdt-2 were deleted (⌬3␤G⌬cdt-1⌬cdt-2), the induction profile for both cbh-1 and gh5-1 matched the starvation/derepression profile (Fig. 3A). These data imply that this strain is unable to sense the presence of cellobiose. In addition to cellulase gene induction, we examined the growth of strains carrying both the single deletion and double deletions of cdt-1 and/or cdt-2 (Fig. 3B) in the ⌬3␤G background strain. The ⌬3␤G strain grows slowly on cellobiose, likely due to the expression of minor predicted ␤-glucosidase enzymes (23,32). Although additional deletion of either cdt-1 or cdt-2 did not exacerbate this growth defect, a ⌬3␤G strain carrying deletions of both cdt-1 and cdt-2 (⌬3␤G⌬cdt-1⌬cdt-2) showed almost no growth after 72 h on 2% cellobiose (Fig. 3B). Taken together, these results reveal that either cdt-1 or cdt-2 are required for the induction of cellulose gene expression when N. crassa is exposed to cellobiose as well as for its growth on cellobiose or Avicel.

Point Mutants in CDT-1 and CDT-2 Significantly Decrease
Cellobiose Uptake, but Still Allow Cellulase Gene Induction-When heterologously expressed in S. cerevisiae, CDT-1 and CDT-2 have been shown to be high-affinity transporters, with CDT-1 acting as a proton symporter and CDT-2 acting as a facilitator (33). To identify functionally important residues within the cellodextrin transporters, CDT-1 and CDT-2, several conserved amino acids were individually mutated to alanine. The functional consequence of each mutation was determined by monitoring the growth of S. cerevisiae strains expressing the intracellular ␤-glucosidase, GH1-1, and mutant CDT-1 or CDT-2 transporters, using cellobiose as a sole carbon source, as described previously (22) (Fig. 4A). From these libraries of mutant CDT-1 and CDT-2 (34), several completely abolished growth of S. cerevisiae on cellodextrins, implying that these residues were required for cellobiose transport. A comparison of these residues indicates that the sites we chose to mutate are highly conserved (Fig. 4B).
The fact that deletion of cdt-1 and cdt-2 eliminated nearly all growth of N. crassa on cellobiose or Avicel (Fig. 3B) allowed us to probe the transport properties of these transporters in their native host. The wild type cellodextrin transporters and point mutants were GFP-tagged and individually transformed into N. crassa, replacing the wild type copy of cdt-1 or cdt-2 at its resident locus with its GFP-tagged form ("Experimental Procedures"). Only those strains in which mutant CDT-1 or CDT-2 localized properly to the membrane (Fig. 5A) were further tested for cellobiose uptake (Fig. 5B) and cellulase induction (Fig. 6, A and B). Comparing cellobiose uptake in the ⌬3␤G background strain to that in the strain lacking both cdt-1 and cdt-2 confirmed that CDT-1 and CDT-2 are the major transporters responsible for cellobiose uptake in N. crassa (Fig. 5B). Strains carrying a deletion of each transporter individually showed a slight decrease in cellobiose uptake, but as shown earlier (Fig. 3) did not significantly affect the cellulase gene induction by cellobiose. GFP-tagged CDT-1 or CDT-2 transporters also did not significantly alter cellobiose uptake or the ability to induce cellulases in response to cellobiose or Avicel as compared with wild type (Fig. 5, A and B). Thus, the presence of either cdt-1 or cdt-2, with or without a GFP tag, was sufficient to allow induction of cellulose genes in response to either crystalline cellulose or physiologically relevant concentrations of cellobiose.
As predicted by the S. cerevisiae growth assays (Fig. 4A), expression of three of the four defective transporters in N. crassa resulted in strains in which cellobiose uptake was almost completely abolished (Fig. 5B). We hypothesized that if N. crassa senses cellobiose intracellularly, abolishing transport should also abolish cellulase gene induction in a similar fashion to what was observed in the strain lacking both transporters (⌬3␤G⌬cdt-1⌬cdt-2) (Fig. 3A). We therefore examined the ability of cellobiose or Avicel to induce cellulases in N. crassa strains expressing the defective transporters. To our surprise, three of the four point mutants still promoted induction of cbh-1 and gh5-1 in response to either cellobiose or Avicel at levels indistinguishable from wild type (Fig. 6). Particularly noteworthy, the ⌬3␤G⌬cdt-1 cdt-2 (E417A)::GFP strain had a 9-fold reduced rate of cellobiose import (Fig. 5B), yet had an indistinguishable cellulase induction profile when compared with WT cdt-2 tagged with GFP (Fig. 6, strain ⌬3␤G⌬cdt-1 cdt-2::GFP). The fourth point mutant, ⌬3␤G⌬cdt-1 cdt-2 (E476A)::GFP showed a decrease in induction of cbh-1 and gh5-1 in response to either cellobiose or Avicel as compared with wild-type with GFP-tagged cdt-2, although the induction of cbh-1 and gh5-1was still far above that observed for starvation/derepression, as shown by expression levels in the ⌬3␤G⌬cdt-1 ⌬cdt-2 mutant (Fig. 6).

DISCUSSION
In this study, we identified two cellodextrin transporters in the model cellulolytic fungus N. crassa (cdt-1 and cdt-2) that are essential for the induction of cellulase gene expression in response to cellulose. Furthermore, in a strain lacking three ␤-glucosidase genes, one or both transporters are required for sensing the cellulose-derived product cellobiose. Deletion of either cdt-1 or cdt-2 does not affect the ability of cellulose or cellobiose to induce cellulase gene expression or activity, indicating that both transporters contribute to cellobiose sensing. However, based on the growth defect for a strain lacking cdt-2 (22) (Fig. 1B), we conclude that cdt-2 is crucial for sustained growth on cellulose.
Recently, two studies in T. reesei identified transporters involved in lactose uptake and strains lacking these lactose transporters no longer induce cellulases in response to lactose. However, because these strains are still able to induce cellulases in response to sophorose (10) or crystalline cellulose (11), these results imply that cellulase signaling via lactose either occurs through an alternative transporter or an alternative mechanism unrelated to the one used during growth on plant cell wall substrates. In a recent study using sophorose, cellobiose or crystalline cellulose as inducers, T. reesei was shown to use a "trans-  JANUARY 31, 2014 • VOLUME 289 • NUMBER 5 ceptor" Stp1, a homologue of the receptors Rgt2 and Snf3 in S. cerevisiae, to regulate the carbon catabolite repression (12). Nutrient transceptors, transporter-like proteins with a receptor function, have been identified and characterized in several systems including E. coli (glucose-6-phosphate sensing (35)), S. cerevisiae (nitrogen (36), phosphate (37), and glucose sensing (38)), Drosophila (amino acid sensing (39)), Arabidopsis (nitrate sensing (40)), as well as in the fungus Ustilago maydis (41) and Candida albicans (42) for the control of filamentation. Notably, in T. reesei, the transceptor Stp1 seems to act independently of the sophorose induction mechanism. By contrast, the T. reesei protein Crt1 (Trire2:3405) may act directly as a receptor for cellobiose and sophorose, as its deletion does not impact cellobiose or sophorose import, but abolishes cellulase induction (12). While the studies performed in T. reesei utilize single transporter deletion strains to examine transporters involved in cellulase signaling, the data presented in this study utilizes strains containing several deletions as well as site directed mutants to determine how transport of cellobiose versus signaling though transceptors mediates the induction of cellulases.

Evidence for Transceptor Function of Cellodextrin Transporters
In contrast to T. reesei, N. crassa uses an alternative mechanism for sensing cellulose. On both crystalline cellulose and cellobiose (the major soluble end product of cellulases) the two transporters CDT-1 and CDT-2 are solely responsible for the ability to of N. crassa to sense and grow on cellulose in plant cell walls. Both CDT-1 and CDT-2 seem to act as "transceptors" with the ability for robust cellodextrin transport (21) (Fig. 5B) and cellulase gene induction by cellobiose (Fig. 6). While further characterization must be completed to map the structural determinants in CDT-1 and CDT-2 to fully separate nutrient sensing from transport, our studies looking at two point mutations in each transporter have helped to begin to separate transport from signaling and lead to the model presented in Fig. 7. Based on our homology models to the xylose transporter XylE (Fig. 4C) homologous point mutants in CDT-1 (E476A) and CDT-2 (E417A) fall within an intracellular unstructured loop region and may block the import of cellobiose by preventing a required conformational change in the protein. The other point mutants in CDT-1 (D307A) and CDT-2 (S224A) fall within an intracellular ␣-helix and an intracellular transition between an unstructured loop and ␣-helix, respectively. Their role in blocking the import of cellobiose is less easy to predict due to their further distance from the transmembrane region.
The data presented in this study significantly adds to the model for the induction of cellulases in filamentous fungi, however several questions remain. Although the mutations identified in this study begin to define regions of CDT-1 and CDT-2 required for transport versus nutrient signaling, support for a role for CDT-1 and CDT-2 as transceptors will additionally require the identification of the mechanism of intracellular signaling that ultimately results in the activation of the transcription factor CLR-1, the major transcriptional regulator required for induction of cellulase genes, and CLR-2, which also regulates cellulase gene expression in N. crassa (Fig. 7) (43,44). Given the widespread presence of cellodextrin transporters in diverse fungi (22), such signaling networks may also function in a wide range of filamentous fungi of ecological or industrial interest for biofuels applications, such as the related thermotolerant fungus Myceliophthora thermophila (45,46).