Solid-State Fermentation with Trichoderma reesei for

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1985, p /85/ $02.00/0 Copyright C 1985, American Society for Microbiology Vol. 49, No. 1 Solid-State Fermentation with Trichoderma reesei for Cellulase Production D. S. CHAHAL Bacteriology Research Centre, Institut Armand-Frappier, Laval, Quebec, Canada H7V 1B7 Received 7 May 1984/Accepted 16 October 1984 Cellulase yields of 250 to 430 IU/g of cellulose were recorded in a new approach to solid-state fermentation of wheat straw with Trichoderma reesei QMY-1. This is an increase of ca. 72% compared with the yields (160 to 250 IU/g of cellulose) in liquid-state fermentation reported in the literature. High cellulase activity (16 to 17 IU/ml) per unit volume of enzyme broth and high yields of cellulases were attributed to the growth of T. reesei on a hemicellulose fraction during its first phase and then on a cellulose fraction of wheat straw during its later phase for cellulase production, as well as to the close contact of hyphae with the substrate in solid-state fermentation. The cellulase system obtained by the solid-state fermentation of wheat straw contained cellulases (17.2 IU/ml), I-glucosidase (21.2 IU/ml), and xylanases (540 IU/ml). This cellulase system was capable of hydrolyzing 78 to 90% of delignified wheat straw (10% concentration) in 96 h, without the addition of complementary enzymes, l-glucosidase, and xylanases. Solid-state fermentation (SSF) is a process whereby an insoluble substrate is fermented with sufficient moisture but without free water. (The abbreviation SSF is also used for "simultaneous saccharification and fermentation." But SSF is retained here for "solid-state fermentation" because it is also commonly used [4, 8, 29], and it is an antonym to another state of fermentation, i.e., "liquid-state fermentation" [LSF].) In liquid-state fermentation (LSF), on the other hand, the substrate is solubilized or suspended as fine particles in a large volume of water. In most LSF, substrate concentrations ranging from 0.5 to 6% are used depending upon the density of the substrate. SSF requires no complex controls and has many advantages over LSF (8); however, it has its own inherent problems (4). A critical analysis of literature on enzymatic hydrolysis reveals that high cellulase activity per unit volume of fermentation broth is the most important factor in obtaining sugar concentrations of 20 to 30% from hydrolysis of cellulose for ethanol production from cellulosic materials (3). It has also been confirmed that cellulase activity per unit volume can be increased by increasing the cellulose concentration in the medium (16), but it is not possible to handle more than 6% cellulose in a conventional fermentor because of rheological problems. A maximum concentration of substrate which can be handled in the conventional fermentor is ca. 2% for wood pulp and 6% for crystalline cellulose. Therefore, to increase the cellulose concentration to over 6%, SSF seems to be the most attractive alternative (3, 4). Trichoderma reesei is well known as a cellulase-producing organism (1-3, 7, 16, 21, 24, 25). The literature on enzyme production indicates that various mutants of T. reesei are able to produce 160 to 250 IU/g of pure cellulose under LSF (Table 1). The highest cellulase yield, 290 IU/g of cellulose by mutant Rut-C30 as recorded by Tangnu et al. (25), has not been reported again. Therefore, we concluded that the cellulase potential of various mutants of T. reesei ranges between 160 and 250 IU/g of pure cellulose in LSF. Development of an economical process for cellulase production is hindered because of the high costs of substrate (pure cellulose) and of some of the chemicals, such as proteose peptone, and because of low yields of cellulases per unit of cellulose. To overcome these bottlenecks, we first 205 used a cheap source of cellulose which requires minimum pretreatment and purification, and we then increased the cellulase yields per unit of cellulose. At present the cheapest cellulose sources are lignocelluloses (crop residues, wood, and wood residues). Peitersen (18) obtained a filter paper activity equivalent to 0.28 IU/ml by growing T. reesei QM9123 on alkali-treated barley straw. Tangnu et al. (25) also reported very low activity of cellulases, i.e., 0.12 and 0.28 IU/ml on 1 and 2% acid-treated corn stover, respectively. Cellulase activity increased to 2.0 IU/ml when washed; acid and base-treated corn stover was used at a 2% concentration. Recently, high cellulase activity, 3.7 IU/ml (168 IU/g of cellulose), was reported by Chahal et al. (5) by growing T. reesei (Rut-C30) on 2.2% cellulose from washed, steamtreated wood. Cellulase yields of 168 IU/g of crude cellulose were as good as those obtained on pure cellulose by others (Table 1). Encouraged by these results, we envisaged a new approach to producing cellulases on cheap cellulose sources and to increasing cellulase yields per unit of cellulose. The new approach for the production of a cellulase system with high hydrolytic potential was to grow T. reesei on lignocellulose in SSF, similar to the Koji process of Toyama (27) except that wheat bran, wheat germ, or rice bran, the expensive additives, were not used. Moreover, in the new approach, the lignocelluloses were not delignified since almost all hemicelluloses are removed during delignification. Rather than delignification and removal of hemicelluloses, the lignocelluloses were treated with a small quantity of NaOH to solubilize some of the hemicelluloses and lignin to expose cellulose. The treated lignocelluloses were not washed, and all of the solubilized hemicelluloses and lignin were retained in the medium. MATERIALS AND METHODS Microorganisms. T. reesei QM9414, kindly supplied by Mary Mandels, U.S. Army Natick Development Center, Natick, Mass., was continuously maintained on delignified wheat straw (WS) agar medium in petri plates. The WS agar medium was specially designed for this purpose. The nutrients in the medium are described below. In one of the colonies, a sector showing high hydrolytic activity was no-

2 206 CHAHAL ticed. The mutation occurred due to continuous subculturing on WS medium (usually cultures of T. reesei are maintained on potato-dextrose agar). The new mutant was transferred to similar medium, and in five further transfers a stable mutant was obtained. This new mutant was tentatively named QMY-1. Another hyper-cellulase-producing mutant of T. reesei, NRRL (Rut-C30), received from J. J. Ellis, Northern Regional Research Centre, Peoria, Ill., was also used for comparison. It is referred to below as Rut-C30. Substrate. WS, ground to 20-mesh powder, was used as a source of cellulose. WS contains (percent dry weight): cellulose, 40; hemicelluloses, 29.2; lignin, 13.6; protein, 3.6; and other materials, 13.6 (23). Therefore, WS is composed of ca. 70% insoluble carbohydrates suitable for the growth of T. reesei and for cellulase production. Aspen pulp was prepared by a chemical-thermomechanical process (10), and the pulp thus prepared was called chemithermomechanical pulp (CTMP). During this process, the wood was pulverized to fine fibers which still retained most of the hemicelluloses and lignin. The chemical composition of CTMP was as follows (percent dry weight): cellulose, 63 to 66; hemicelluloses, 15 to 18; and lignin, 9 to 11. Pretreatment. Powdered substrate (5 g; 20 mesh) was dispensed into each Erlenmeyer flask of 250-ml capacity. Each substrate was treated with NaOH (4% [wt/wt]) with 33.3% moisture at 121 C (WS for 0.5 h and CTMP for 1 h). The treated substrate was not washed. All of the solubilized hemicelluloses and lignin were retained in the fermentation medium. After the addition of nutrients, the ph was adjusted TABLE 1. Cellulase production potential of new mutants of T. reesei in LSF Cellulose Cellulase Mutanta (yr developed) concn (IU/g of Renfce- (%) ~~cellulose) QM6a (parent strain; ) QM9414 (1971) MCG77 (1977) NG14 (1977) Rut-C30 (1979) L27 (1981) CL-847 (1983) 5 (+1) D1-6 (1983) a Strains QM6a, QM9414, MCG77, and NG14 were from the U.S. Army Natick Research and Development Laboratories; Rut-C30 was from Cook College, Rutgers, N.J.; L27 was from Cetus Corp.; CL-847 was from France; and D1-6 was from Delhi. b Range, 160 to 250. e +1, 1% glucose. APPL. ENVIRON. MICROBIOL. to ca. 5.8 with H2SO4. Water solubles were obtained by suspending treated WS in water (1:10) and filtering through four layers of cheese cloth. The WS solubles contained 2.5% solids (hemicelluloses, lignin, and other cell solubles). Nutrients. Nutrients described by Mandels and Weber (15) for cellulase production were supplied in concentrated form, but proteose peptone was replaced with yeast extract (Difco Laboratories, Detroit, Mich.). The quantity of nutrients required for each substrate was determined at the rate of their carbohydrate content. The required amount of concentrated nutrient salt solution (5 ml for WS and 5.7 ml for CTMP) was added to 5 g of substrate. The concentrated salt solution contained the following, dissolved in 200 ml of water: KH2PO4, 28 g; (NH4)2SO4, 19.6 g; urea, 4.2 g; MgSO4 * 7H20, 4.2 g; CoCl2, 4.2 g; FeSO4 * 7H20, 70 mg; MnSO4 7H20, mg; ZnSO4 * 7H20, 19.6 mg; CaC12, 28 mg; and yeast extract, 7 g. As referred to below, a full concentration of the nutrients means the complete required quantity of nutrients as mentioned above, whereas a onehalf concentration is one-half of that quantity. All of the flasks were autoclaved at 121 C for 20 min, after the nutrient salt solution was mixed with the substrate. Moisture. The moisture content of the substrates after pretreatment and the addition of nutrients and inoculum was 80% (wt/wt) in SSF. Sterilized water was added for LSF, to obtain the desired concentration of the substrate in the fermentation medium. Inocula. Inocula of mutants QMY-1 and Rut-C30 were produced on the modified medium as described above but containing 1.5% glucose, with the nutrient salt solution diluted accordingly. For inoculation of each flask containing 5 g of substrate, 5 ml of 2-day-old culture was used. The inoculum was spread on the surface of the substrate. Culture conditions. All of the SSF cultures were incubated at 30 C in a humidified incubator (about 80% relative humidity), whereas the LSF cultures were incubated at the same temperature on a shaker at 150 rpm. Extraction of the cellulase system. The culture of SSF from each flask (originally 5 g of substrate) was mixed well with more water to bring the final weight of the mixture (mycelium plus unutilized lignin, cellulose, and hemicelluloses) to 100 g. Tween 80 was added at a rate of 0.1%. The mixture was shaken for 0.5 h and centrifuged. The supernatant was used for enzyme determination. It was estimated that about 7 to 10% cellulases remained adsorbed on the residues (mycelium and unutilized cellulose, hemicelluloses, and lignin) when the residues were suspended in water and Tween 80 as before and the supernatant was tested for cellulase titer. Analyses. Cellulase titer was calculated in international units of enzyme activity (glucose released per min) on filter paper for 60 min, by the method of Mandels et al. (11). P-Glucosidase titer was measured in international units of glucose released per min with 1% salicin solution for 30 min. Xylanase titer was measured in international units of xylose released per min with 1% xylan for 10 min (9). The cellulase yield per gram of cellulose was calculated by dividing total international units of cellulase titer in 100 ml of enzyme broth by grams of cellulose present in the substrate supplied in the flask. Sugars were estimated by using a Beckman 344 gradient high-pressure liquid chromatograph with an Altex 156 refractive index detector and a Spherogel 7.5% carbohydrate column with a flow rate of 0.5 ml/min in the mobile phase of water at 80 C. The sugar samples were appropriately diluted before injection.

3 VOL. 49, 1985 RESULTS AND DISCUSSION Cellulase production in LSF and the effect of different concentrations of WS. The highest cellulase titer (1.65 IU/ml) and cellulase yield (412 IU/g of cellulose) in LSF were obtained with mutant QMY-1 in 1% WS (0.4% cellulose) slurry after 7 days (Table 2). When the concentration of WS was increased to 5% (2% cellulose), the enzyme production time was increased from 7 to 11 days. The enzyme activity increased to 6.0 IU/ml, but the cellulase yield dropped to 300 IU/g of cellulose (Table 2). The drop in cellulase yields might have been due to poor mass transfer in the thick slurry of 5% WS. The decrease in cellulase yield by mutant QM9414 with increases in the concentration of cellulose is also evident from the work of Sternberg (24), Gallo et al. (7), Ryu and Mandels (21), and Tangnu et al. (25), as presented in Table 1. TABLE 2. Cellulase production on WS by QMY-1 Concn of Time of Cellulase Type of Concn of III.b Cellulase fermentation WS M cellulose incuba- titer yield (nutrient concn) [wt/wt]) (% tion (IU/ml) (IU/g of [wt/wt]) (days) (l/n cellulose) LSF (full concn) SSF (full 20a 8b concn) C' SSF (one-half concn) c g of WS + 20 g of water (no free water) = 20%o solids in each flask. b 5 g of WS contains 2 g of cellulose = 8% cellulose in each flask. c Cultures were stirred once after 11 days of incubation and were further incubated for 11 days without stirring. Cellulase production in SSF. (i) On WS. SSF was carried out with a full concentration of nutrients in one set of experiments and with a one-half concentration in another set to evaluate the effect of different concentrations of salts in the medium, since some microorganisms are unable to grow in the high osmotic pressure of the medium. T. reesei QMY-1 was quite tolerant to the high concentrations of the nutrients, as indicated in Table 2. It produced the highest enzyme titer (8.6 IU/ml) and cellulase yield (430 IU/g of cellulose) in SSF on a full concentration of nutrients, after 22 days. The highest cellulase yield, 412 IU/g of cellulose, obtained in LSF, was because that medium contained the lowest cellulose concentration (0.4%), as discussed earlier. However, in the present study, a cellulase yield of 275 to 300 IU/g of cellulose in LSF on 2% cellulose concentrations was considered for comparisons between LSF and SSF. When the nutrients were supplied in a one-half concentration, the cellulase titer dropped to 6.7 IU/ml, cellulase yield dropped to 335 IU/g of cellulose, and there was no increase in enzyme yields after an incubation of more than 14 days. But when the cultures were stirred after 11 days of growth and further incubated for 11 days without any stirring (total SOLID-STATE FERMENTATION WITH T. REESEI 207 of 22 days of incubation), the cellulase titer decreased somewhat for the full concentration of nutrients, whereas it increased considerably (8.0 IU/ml) for a one-half concentration of nutrients. This indicates that half of the quantity of required nutrients was sufficient to get an optimum cellulase titer as well as an optimum cellulase yield. This finding could contribute to a reduction in the cost of enzyme production. Recently, more expensive media have been developed to increase cellulase production by new mutants of T. reesei in LSF on pure cellulose (17, 28). Even then, the highest yields obtained are quite low, i.e., 140 IU/g of cellulose for mutant D1-6 (17) and 229 IU/g of cellulose for mutant CL-847 (28) (Table 1). (ii) On WS and CTMP. Cellulase production by strain QMY-1 was compared with that of strain Rut-C30, a hypercellulase-producing mutant, on two different lignocellulosic substrates, WS and CTMP. QMY-1 produced its highest cellulase titer (over 8 IU/ml) and yield (over 400 IU/g of cellulose) on treated WS as compared with that of Rut-C30, a 6.2 IU/ml cellulase titer and a yield of 310 IU/g of cellulose. Cellulase production by both of the mutants decreased considerably on untreated WS (Table 3). The mutant Rut- TABLE 3. Cellulase production by mutants QMY-1 and Rut-C30 on WS and CTMP Mutant QMY-1 Rut-C30 Substrate'~(days) Cellulase Cellulase Substateadays) Cellulase titer yield yed Cellulase titer yield yil (IU/mI) (IU/g of (IU/mI) (IU/g of (/m cellulose) (Um cellulose) Treated WS Untreated WS Treated CTMP Untreated CTMP a Cellulose content of wheat straw, 40%; cellulose content of CTMP, 66%. There was 5 g (dry weight) of each substrate in each flask. Substrates were treated with 4% NaOH (wt/wt) at 121'C for 0.5 h (WS) or 1 h (CTMP) with a 1:2 (solid to liquid) ratio.

4 208 CHAHAL TABLE 4. Cellulase production by different mutants of T. reesei on different cellulosic substrates in LSF Mutant' and Cellulase Cellulase substrateb titer" yield (5%o) (IU/ml) (lu/g of (5%) (lu/mi) ~~~~~~~~~~cellulose) QM-9414 Solka Floc SE SEWA Rut-C30 Solka Floc SE SEWA E.58 Solka Floc SE SEWA a Mutants were grown for 11 days at 28 C in LSF. E.58 is a mutant from Trichoderma harzianum (Forintek Culture Collection). bse, Steam-exploded wood (55% cellulose); SEWA, steam-exploded wood, water extracted and alkali treated (91% cellulose). C Data from reference 6. C30 failed to grow in a number of flasks containing WS or CTMP. This finding indicated that this mutant is not well adapted to such conditions of SSF. The mild alkali treatment of CTMP did not affect cellulase production by either mutant (Table 3). Because the highest cellulase titer and yield obtained on treated CTMP after 20 days in SSF were almost comparable to those obtained on untreated CTMP, CTMP seems to be a good substrate for cellulase production even without any further treatment. Further studies on cellulase production on CTMP are in progress. The cellulase titer and yield on CTMP with strains QMY-1 and Rut-C30 (Table 3) and on treated WS (Table 2) with strain QMY-1 in SSF were higher than those of mutants QM-9414, Rut-C30, and E.58 on steam-exploded wood, on steam-exploded and alkali-treated wood with water extracted, and even on pure cellulose (Solka Floc; Brown Co., Berlin, N.H.) in LSF, as reported by others (Table 4). The titer and yield of cellulase obtained with QMY-1 in SSF were also higher than those obtained by other workers who grew various mutants of T. reesei on pure cellulose in LSF (Table 1). The results clearly indicate that the new approach of retaining the hemicelluloses and lignin of the alkali-pretreated lignocelluloses (WS, CTMP) in SSF increased significantly the cellulase titer per unit volume and the cellulase yield per unit of cellulose. Role of hemicelluloses and lignin in cellulase production. The increase in cellulase titer was postulated to be due to the use of hemicelluloses during the initial growth of T. reesei and then to the use of cellulose during the later phase of growth for production of cellulases. To test this postulate, we grew T. reesei QMY-1 in LSF on pure cellulose (acellulose; Sigma Chemical Co., St. Louis, Mo.) in one set of experiments, and cellulose was fortified with a mixture of solubles obtained from WS. Delayed and slow synthesis of cellulases during the early phase for WS soluble-fortified cellulose was attributed to the presence of hemicelluloses, an easily metabolizable carbon source (Fig. 1). After hemicelluloses were used, cellulase synthesis increased considerably during the later phase of fermentation. This indicated that WS solubles which contained mostly hemicelluloses and APPL. ENVIRON. MICROBIOL. lignin were responsible for the high cellulase titer (3.4 IU/ml) and yield (340 IU/g of cellulose). However, further work on the role of hemicelluloses and lignin, individually and in combination, is in progress. Composition of the cellulase system. The cellulase system produced in SSF contained the following enzymatic activities (international units per milliliter): cellulase, 8.6; 3- glucosidase, 10.6; and xylanase, 270. The xylanase titer was quite variable (between 190 and 480 IU/ml); however, the ratio of cellulases and 3-glucosidase varied between 1:1 and 1:1.5 in various cellulase system preparations. These are enzyme activities when 5 g of WS fermented in SSF was suspended in ca. 100 ml of water to extract the enzymes. The enzyme titer could be doubled (17.2 IU/ml) by extracting the enzyme in 50 ml of water. The composition of the cellulase system indicated that there was no need to add extra 3-glucosidase or xylanase for the hydrolysis of pure cellulose or lignocelluloses. Hydrolytic potential of the cellulase system. (i) Cotton. The cellulases produced in SSF, when used in a substrate-to-enzyme ratio of 1 g:20 IU, showed saccharification of different concentrations of untreated cotton as high as that reported by Mandels et al. (13) (Table 5). There was also an indication that this enzyme system was able to hydrolyze a higher concentration (10%) of cotton without any reduction in the percentage of hydrolysis. The high-pressure liquid chromatography of the hydrolysate obtained from 10% concentrations of cotton showed mostly glucose (92.8%) and very little cellobiose (3.5%). It is therefore assumed that cellulases produced in SSF may be able to hydrolyze a higher concentration of cellulose to obtain high concentrations of glucose in the hydrolysate suitable for economical ethanol fermentation or for fermentation of any other product. (ii) Cellulose fiber. Hydrolysis of cellulose fiber with cellulases produced in SSF was faster than that of cotton. There was also an indication that cellulases produced in SSF gave a faster rate and higher percentage of hydrolysis than cellulases produced in LSF (Table 6). (iii) Delignified WS. WS delignified by the method of Toyama (26) was hydrolyzed with the cellulase system 0- LL. 4 I TIME (h) FIG. 1. Effect of hemicelluloses and lignin on cellulase production by mutant QMY-1 on 1% cellulose (sigma cell). (The ratio of cellulose to WS solubles in the medium was 1:0.25.) Symbols: 0, cellulose only (control); A, cellulose plus WS solubles containing hemicelluloses and lignin.

5 VOL. 49, 1985 TABLE 5. Hydrolysis of cotton with cellulases produced in SSF Substrate Substrate/enzyme % Saccharification" concn w] ratio (giu) 24 h 48 h 1 1: : : : : C 1: Saccharification = grams of reducing sugars x 0.9 x 100. grams of substrate b 15.4% Saccharification = 17.1 g of reducing sugars per liter (15.7 g of glucose [92.8%] g of cellobiose [3.5%] g of xylose [2.2%] = 16.9 g of sugars per liter, detected by high-pressure liquid chromatography). C Data derived by using a substrate concentration of 5% (dry weight) are from reference 13, with T. Reesei QM9414 cellulases. produced in SSF at a 10% (wt/vol) concentration. Hydrolysis was done at ph 6.7, the original ph of the enzyme solution, and also at standard ph 4.8 (Fig. 2 and 3). The rate of hydrolysis of cellulose into glucose was very high for the first 20 h of hydrolysis at both ph levels, and ca. 65% of total hydrolysis was recorded during this period. Almost all of the xylan and arabinan were hydrolyzed within the first 20 h of hydrolysis. Very little cellobiose accumulated in the hydrolysate. The maximum concentration of cellobiose accumulated was 7.75 g/liter (Fig. 2), which was too low to cause any significant inhibition of cellulose hydrolysis (12, 14). After 20 h of hydrolysis, the concentration of cellobiose further decreased to ca. 3 g/liter. C C0 (.) Co nx SOLID-STATE FERMENTATION WITH T. REESEI 209 TABLE 6. Comparison of hydrolysis of cellulose' fiber with cellulases produced in SSF and LSF System of Substrate/enzyme % Saccharification fermentaion ratio (g/iu) 24 h 48 h 72 h SSF 1: LSF 1: a Cellulose fiber (Sigma Chemical Co., St. Louis, Mo.), 5% concentration. The hydrolysate obtained with ph 6.7, after 96 h of hydrolysis, contained g of sugars per liter (glucose, g; cellobiose, 3.19 g; xylose, g; arabinose, 1.67 g), whereas the hydrolysate obtained with ph 4.8, after the same hydrolysis period, contained g of sugars per liter (glucose, g; cellobiose, 3.88 g; xylose, g; arabinose, 1.46 g), giving total saccharifications of 89.7 and 77.9%, respectively. Conclusions. The increase in cellulase yields, from a range of 160 to 250 IU/g of pure cellulose to a range of 250 to 430 IU/g of crude cellulose from WS and CTMP, was due to the use of a hemicellulose fraction during the initial growth of the organism and to the production of cellulases on a cellulose fraction of substrates during the later phase of growth of the organism. The presence of lignin could be another factor in the increase of cellulase yields. The role of lignin and hemicelluloses in enzyme production is being further evaluated. High cellulase activity of 8.6 IU/ml could be doubled (17.2 IU/ml) by extracting cellulases with half the quantity of water. Because high cellulase titer (over 15 IU/ml) is required to obtain high concentrations of glucose for economical ethanol fermentation (4), SSF seems to hold promise for obtaining a high cellulase titer per unit volume of enzyme broth. The cellulase system produced in SSF contained sufficient quantities of,b-glucosidase and xylanase for com _ 70 CP,, 60 c 0 50 t 40 0 x T I M E ( h) FIG. 2. Hydrolysis of delignified WS at ph 6.7. Symbols: A, total sugars; x, glucose; *, xylose; Ol, cellobiose; 0, arabinose. 20 io- 0 K..m 0 0 oi a T I 1----T- I I I I I iob) I TIME ( h ) FIG. 3. Hydrolysis of delignified WS at ph 4.8. Symbols are defined in the legend to Fig. 2.

6 210 CHAHAL plete hydrolysis of pure cellulose as well as lignocelluloses. It was also assumed that the cellulase system produced in SSF was rich in C1 factor as proposed much earlier by Reese et al. (20). It was reported by them (20) that the C1 factor was essential to hydrolyze the crystalline portion of cellulose, and this concept is still maintained by Reese (19). By the new approach, ca. 100 g of sugars per liter were obtained by hydrolyzing 100 g of delignified WS with the cellulase system produced with SSF. Moreover, SSF enables drastic reductions in the cost of enzyme because cellulase yields per unit of cellulose were increased by 72%, crude cellulose (WS) with a minimum of pretreatment and purification could be used as a carbon substrate, the quantity of nutrients could be reduced to one-half, and SSF does not require very complex control systems. However, SSF has its own inherent problems, including maintenance of ph, moisture level, aeration, agitation, etc., when large quantities of solid substrates are used (4). ACKNOWLEDGMENTS I thank M. Mandels and E. T. Reese, Materials Protection and Biotechnology Division, Science and Advanced Technology Laboratory, U.S. Army Natick Research and Development Laboratories, Natick, Mass., and V. Portelance, Bacteriology Research Center, Institut Armand-Frappier, Laval, Quebec, Canada, for critical reading of the manuscript and for their valuable suggestions. I am also very grateful to Johanne Lemay for her technical help and to P. S. Chahal for drawing the figures and for technical help. LITERATURE CITED 1. Allen, A. L., and R. E. Andreotti Cellulase production in continuous and fed-batch culture by Trichoderma reesei MCG 80. Biotechnol. Bioeng. Symp. 12: Blanch, H. W., and C. R. Wilke Cellulase production and kinetics, p H. E. Duckworth and E. A. Thompson (ed.), Proceedings of the International Symposium on Ethanol from Biomass. The Royal Society of Canada, Ottawa, Ontario. 3. Chahal, D. S Enzymatic hydrolysis of cellulose: state-ofthe-art, p. 74. National Research Council of Canada report no National Research Council of Canada, Ottawa. 4. Chahal, D. S Growth characteristics of microorganisms in solid state fermentation for upgrading protein values of lignocelluloses and cellulase production, p In H. W. Blanch and E. T. Popoutsakis (ed.), Foundations of biochemical engineering: kinetics and thermodynamics in biological systems. American Chemical Society, Washington, D.C. 5. Chahal, D. S., S. McGuire, H. Pikor, and G. Noble Production of cellulase complex by Trichoderma reesei Rut-C30 on lignocellulose and its hydrolytic potential. Biomass 2: Forintek Canada Corp Conversion of cellulose to ethanol using a two-stage process. ENFOR project no. C-181 (2). Forintek Canada Corp., Ottawa. 7. Gallo, B. J., R. Andreotti, C. Roche, D. Ryu, and M. Mandels Cellulase production by a new mutant strain of Trichoderma reesei MCG77. Biotechnol. Bioeng. Symp. 8: Hesseltine, C. W Solid state fermentations. Biotechnol. Bioeng. 14: Ishaque, M., and D. Kluepfel Production of xylanolytic enzymes by Streptomyces flavogriseus. Biotechnol. Lett. 3: Law, K. N., M. Lapointe, and J. L. Valade Production of CTMP from aspen, p International Mechanical Pulp- APPL. ENVIRON. MICROBIOL. ing Proceedings. TAPPI Proceedings. 11. Mandels, M., R. Andreotti, and C. Roche Measurement of saccharifying cellulase. Biotechnol. Bioeng. Symp. 6: Mandels, M., S. Dorval, and J. Medeiros Saccharification of cellulose with Trichoderma cellulase, p In W. W. Shuster (ed.), Proceedings of the Annual Symposium on Fuel from Biomass. Rensselaer Polytechnic Institute, Troy, N.Y. 13. Mandels, M., J. E. Medeiros, R. E. Andreotti, and F. H. Bisset Evaluation of cellulase culture filtrates under use conditions. Biotechnol. Bioeng. 23: Mandels, M., and E. T. Reese Inhibition of cellulases and 1-glucosidase, p In E. T. Reese (ed.), Advances in enzymatic hydrolysis of cellulose and related materials. Mac- Millan Co., New York. 15. Mandels, M., and J. Weber The production of cellulases. Adv. Chem. Ser. 95: Nystrom, J. M., and P. H. DiLuca Enhanced production of Trichoderma cellulase on high levels of cellulose in submerged culture, p Proceedings of the Bioconversion Symposium. Indian Institute of Technology, New Delhi, India. 17. Panda, T., V. S. Bisaria, and T. K. Ghose Studies on mixed fungal culture for cellulase and hemicellulase production. I. Optimization of medium for the mixed culture of Trichoderma reesei D1-6 and Aspergillus wentii Pt Biotechnol. Lett. 5: Peitersen, N Production of cellulase and protein from barley straw by Trichoderma viride. Biotechnol. Bioeng. 17: Reese, E. T History of the cellulase program at the U.S. Army Natick Development Center. Biotechnology Bioeng. Symp. 6: Reese, E. T., R. G. H. Siu, and H. S. Levinson The biological degradation of soluble cellulose derivatives and its relationship to the mechanism of cellulose hydrolysis. J. Bacteriol. 59: Ryu, D. D., and M. Mandels Cellulases: biosynthesis and applications. Enzyme Microb. Technol. 2: Shoemaker, S. P., J. C. Raymond, and R. Bruner Cellulases: diversity among improved Trichoderma strains, p In A. Hollaender (ed.), Trends in the biology of fermentations for fuels and chemicals. Plenum Publishing Corp., New York. 23. Sloneker, J. H Agricultural residues including feedlot wastes. Biotechnol. Bioeng. Symp. 6: Sternberg, D Production of cellulase by Trichoderma. Biotechnol. Bioeng. Symp. 6: Tangnu, S. K., H. W. Blanch, and C. R. Wilke Enhanced production of cellulase, hemicellulase, and,-glucosidase by Trichoderma reesei (Rut-C30). Biotechnol. Bioeng. 23: Toyama, N Utilization of cellulosic wastes by Trichoderma viride, p In G. Terui (ed.), Proceedings of the IV International Fermentation Symposium: Fermentation Technology Today. Society of Fermentation Technology, Japan, Osaka. 27. Toyama, N Feasibility of sugar production from agricultural and urban cellulosic wastes with Trichoderma viride cellulase. Biotechnol. Bioeng. Symp. 6: Warzywoda, M., V. Ferre, and J. Pourquie Development of a culture medium for large scale production of cellulolytic enzyme by Trichoderma reesei. Biotechnol. Bioeng. 25: Zadrazil, F., and H. Brunnert Solid state fermentation of lignocellulose containing plant residues with Sporotrichum pulverulentum Nov. and Dichomitus squalens (Karst.) Reid. Eur. J. Appl. Microbiol. Biotechnol. 16:45-51.