Technical Field The present invention relates to the production of enzymes. More particularly, it relates to the production of thermostable enzymes and ethanol by co-cultures of microorganisms.

Background Art There is an active interest in the production of industrial feedstock chemicals or fuels from biomass.

Starch is a major component of agricultural crops and of corn processing waste and it is a preferred substrate for chemical and enzyme production because of its chemical composition and its higher density than other forms of biomass which facilitates prolonged storage, and decreases transportation and pretreatment costs.

Starch is known to be a valuable starting material for the enzymatic production of sugar, such as glucose, which may be converted by yeast to ethanol. The main amylolytic or starch converting enzymes used for the industrial production of glucose, maltose and maltosac- charide from starch are α-amylase, β-amylase, glucoamylase and pullulanase.

The known amylolytic enzymes, except for bacterial -amylases, are unstable at the elevated tempe ^' ratures preferred for the industrial conversion of starch.

It would be desirable to have both a method for the direct bioconversion of starch into ethanol and produc¬ tion of thermostable starch converting enzymes which could be used at the elevated temperature preferred by industry.

Disclosure of the Invention The primary objects of the present invention are to disclose a method for directly converting starch to ethanol and to disclose thermostable starch converting enzymes.

It is a further object to disclose a novel co-culture of microorganisms which can co-produce ethanol directly from starch and also produce thermostable starch convert¬ ing enzymes. We have discovered a simultaneous, enhanced, single step method for the production of both ethanol and thermo¬ stable starch converting enzymes which comprises culturing a co-culture of Clostridium thermosulfurogenes and Clostridium thermohydrosulfuricum under anaerobic condi- tions on a growth medium containing starch and essential minerals, vitamins and growth factors until detectable enzymatic activity and/or ethanol are present and then, if desired, isolating the enzymes and/or ethanol.

The thermostable starch converting enzymes produced by the co-culture are a β-amylase produced by C. thermosulfurogenes and a pullulanase and a glucoamylase produced by C. thermohydrosu1furicum.

The foregoing and other advantages and objects will be apparent from the description which follows:

Best Mode of Carrying Out the Invention

In the preferred practice of the present invention, a co-culture of C. thermosulfurogenes and C. thermohydro- sulfuricum is grown under anaerobic conditions at about 60°C on a substrate of starch in a medium comprising a source of assimilable nitrogen, essential vitamins and

minerals and growth factors until detectable enzymatic activity or ethanol is present and then, if desired, isolating the enzymes and/or ethanol. The enyzymes produced are a thermostable β-amylase, a thermostable glucoamylase and a thermostable pullulanase.

Organisms and growth conditions. C. thermohydrosulfuricum strain 39E ATCC 33223 and C. thermosulfurogenes strain 4B ATCC 33743 were isolated from Octopus Spring in Yellowstone National Park. Anaerobic culture techniques were employed for media preparation and cell cultivation. C. thermo¬ hydrosulfuricum and C. thermosulfurogenes were routinely grown at 65°C and 60°C, respectively, in 26-ml anaerobic pressure tubes that contained 10 ml of TYE medium with 0.5% glucose or soluble starch, and a N ₉ /C0„ (95:5) gas head space. Culture media were autoclaved for 45 min to assure destruction of the extremely heat resistant spores of thermoanaerobes. For starch metabolism studies in anaerobic pressure tubes and small fermentors, TYE medium was modified to contain the double strengths of vitamin solution, ammonium chloride, magnesium chloride and trace mineral solution. In co-culture studies on 5% starch in a large fermentor, TYE medium was modified to contain 4 times the vitamin solution, magnesium chloride and trace mineral solution, and 2 times ammonium chloride. Co-cultures and mono-cultures of C. thermosulfuro¬ genes were incubated at 60°C, and mono-cultures of C. thermohydrosulfuricum at 65°C, throughout this study. A 5% inoculum was applied for all mono-cultures. Co-culture experiments were initiated by injection of a 2.5% inoculum of each organism which were separately grown until the late logarithmic growth phase. The inoculum was prepared in sealed vessels that contained the same medium as in the experiments.

Starch metabolism time course studies on 2.5% starch were conducted in a fermentor that contained 650 ml of the modified TYE medium. The fermentors were mixed at

200 pm, gassed continuously with N ₂ /CO„ (95:5) at a flow rate of about 0.4 WM, and pH-controlled with 1.7 N ammonium hydroxide solution.

Starch metabolism time course studies of co-cultures on 5% starch were conducted in a 14-liter fermentor that contained 8 liters of the modified TYE medium. For ethanol recovery, 2 flasks (6-liter) that contained 4 liters of water each, respectively, were connected to the fermentor in tandu and were.kept in ice baths during the fermentation period. A condensor was connected to the fermentor and the condensate was displaced into the first flask. The gas flow rate was 0.056 WM.

In mono-culture studies of C. thermohydrosu1furicum with β-amylase of C. thermosulfurogenes, the enzyme was injected (3 units/ml) after sterilization by membrane filtration. The β-amylase was prepared from a C. thermosulfurogenes culture which was grown in a 14-liter fermentor containing 10 liters of LPBB medium with 1% maltose and 0.02% ^* yeast extract. The- culture supernatant was obtained with a continuous-flow centrifuge system and was concentrated.

Quantification of growth, substrates, and products. Culture turbidities were determined by inserting anaerobic pressure tubes into a spectrophotometer. In all growth experiments, except for anaerobic pressure tube cultures, culture broths were diluted 5 times with double distilled water to measure optical density. To measure starch concentration, the culture broth samples were appropri¬ ately diluted with water and 50 ml of sulfuric acid solution (5M) was added to 0,5 ml samples. These samples were placed in steam bath for 3 hours and then neutralized by the addition of 35μl sodium hydroxide solution (10 N) . The reducing sugar content was measured by the dinitrosalicylic acid method. Starch concentration was estimated by fitting the reducing sugar values to the calibration curve for starch solutions which were treated by the same procedures as above.

Glucose was enzymatically determined by the hexo- kimase and glucose-6-phosphate dehydrogenase method. Reducing sugar was estimated using glucose as a standard by the dinitrosalicylic acid method. Total carbohydrate was assayed by the phenol-sulfuric acid method.

The ammonium content of the culture fluid was measured by known procedures and ethanol and acetate were measured by gas chromatography using a flame ionization detector. L-Lactic acid was assayed by standard pro- cedures using lactic dehydrogenase. Hydrogen was quanti¬ fied by thermal conductivity detection procedures. C0 _? production was calculated by the sum of ethanol and acetate production.

Enzyme assays. Cell-free culture broth and washed cell suspensions were used for measurements of β-amylase and other amylolytic enzyme activities, respectively. They were prepared by centrifugation of culture broths at 10,000 x g for 10 min to separate the cells and culture broth. 'The precipitated cells were suspended in the appropriate amount of double distilled water.

Amylase or β-amylase activity was assayed by ^' measurement of reducing power which was liberated during reaction on starch. The reaction mixture (5 ml) con¬ sisted of 2% soluble starch in 0.1 M sodium acetate buffer (pH 6.0) and the appropriately diluted enzyme.

After aerobic incubation at 60°C for 30 min, the reaction was stopped by cooling on ice. The released reducing sugar was measured by the dinitrosalicylic acid method. Pullulanase activity was assayed by analyzing the reducing power released from pullulan, using the dinitro¬ salicylic acid method. The reaction mixture consisted of 0.5 ml of 2% pullulan in 0.2 M" phosphate buffer (pH 7.0) and 0.5 ml of enzyme solution. The reaction was per¬ formed at 60°C for 30 min. , and was stopped by cooling on ice and adding 4 ml of cold dinitrosalicylic acid solu¬ tion. One unit of amylase, β-amylase or pullulanase is defined as the amount of enzyme which liberated 1

^• μmol of reducing sugar as a glucose standard per min under the described conditions.

Glucoamylase activity was measured after incubating the reaction at 60°C for 30 min. The reaction mixture consisted of 0.5 ml of 2% soluble starch solution in 0.2M sodium acetate buffer (pH 4.8) and 0.5 ml of enzyme solution. The reaction was stopped by cooling on ice, and then it was boiled in a steam bath for 10 min. The released glucose was quantified using the hexokinase and glucose-6-phosphate dehydrogenase method. One unit of glucoamylase is defined as the amount of enzyme that liberated one ymole of glucose per min under the conditions.

Results Starch metabolism by Monocultures. The starch transformation time courses for C. thermohydrosulfuricum and C. thermosulfurogenes were compared when they were cultured in complex medium without shaking and pH-control. For both species, the synthesis of amylolytic enzymes (i.e., β-amylase, glucoamylase and pullulanase) was closely related to growth and hence these amylases are produced as primary metabolites. Under these experimental conditions, starch was not completely utilized by either species during the time course. The incomplete utilization of starch by C. thermo- hydrosulfuricum appears related to growth inhibition caused by high hydrogen or proton production because pullulanase and glucoamylase activity completely hydrolyze starch. In contrast, the inability of C. thermosulfuro- genes to completely utilize starch was a function the accumulation of limit dextrins caused by the absence of pullulanase or a debranching activity.

Under these conditions, the exponential growth rate in doublings per hour of the two species on starch versus glucose were: 1.5 versus 1.2 for C. thermosulfurogenes; and 4.1 versus 1.9 for C. thermohydrosu1furicum. The

slower growth rate of C. thermohydrosulfuricum on starch is considered to be a reflection of limited glucoamylase activity and lack of β-amylase activity. Neither species produce significant amounts of α-amylase activity.

Comparison of starch metabolism by mono- and co- cultures. Experiments were initiated to show that starch transformation rates would be enhanced in co-culture versus mono-culture fermentation. The growth of mono- and co-cultures of these thermo anaerobes on starch in sealed pressure tubes was compared. In mono culture, C. thermohydrosulfuricum grew slowly but without a significant lag phase; whereas, C. thermosul¬ furogenes displayed a lag phase but growth was more rapid upon initiation. In co-culture, growth initiated without a lag, and was very rapid in the logarithmic growth phase. Hence, growth rate and total cell concentration iii co-culture fermentation was ^' higher than "in mono-culture fermentations. The starch metabolism parameters of mono- and co- culture fermentations in sealed pressure tubes was com¬ pared. Total growth was enhanced by co-culture fermenta¬ tions at high starch concentrations. Starch was complete¬ ly utilized by the co-cultures at low but not high starch concentrations. The production of β-amylase was not significantly different in mono- versus co-cultures of C. thermosulfurogenes. It is notable that all three amylases (i.e., β-amylase, glucoamylase and pullulanase) were simultaneously expressed in co-culture fermentations. The concomitant growth of both species in co-culture was indicated by: the simultaneous production of all three amylolytic activities; enhanced utilization of starch; and higher total levels and ratios of end products. The ratio of reducing sugar accumulation to starch remaining in mono- versus co-cultures at 3% starch indicates a higher conversion of starch into the low molecular weight oligosaccharides by the co-culture. The dextrose equiva-

lent (DE) values of the remaining carbohydrates were about 74 in co-culture versus 39 in mono-cultures. Therefore, the cessation of growth before complete utili¬ zation of the accumulated saccharides is suggested to be due to the accumulation of toxic hydrogen (for 39E) or protons (for both strains). The final ethanol concentra¬ tions achieved here are not inhibitory to growth of either C. thermohydrosulfuricum (8) or C. thermosulfuro¬ genes. In order to eliminate toxic effects of hydrogen or protons, mono- and co-culture experiments were performed in small fermentors that were pH controlled at 7.0 and continuously gassed with N^/CO- (95:5). PH was controlled with ammonium hydroxide solution (1.7M) because prelim- inary experiments demonstrated that growth was ammonium limited at greater than 1% starch in normal TYE medium.

The fermentation time course of C. thermosulfurogenes grown on 2.5% starch at pH 7.0 displayed an initial lag phase where a high amount of reducing sugars accumulated as a result of active, extracellular β-amylase action. The accumulated reducing sugars were rapidly consumed when growth initiated; and, both β-amylase production and end product formation occurred simultaneously during the logarithmic growth phase. Starch was incompletely used, however, which implied that the β-limit dextrins formed by β-amylase were not further metabolized due to lack of pullulanase activity. The decrease of ethanol concentration observed during the stationary growth phase was the result of evaporation caused by gassing. A starch metabolism time course of C. thermohydro¬ sulfuricum grown under conditions of continuous gassing and pH control at 7.0 displayed an unusual fermentation pattern because biphasic growth was observed, which was indicative of extensive cell lysis. Starch also was not completely utilized by this species under the growth conditions employed. Reducing sugar accumulation in¬ creased according to the culture time, probably in re-

sponse to the amylolytic enzymes produced during growth. In addition, glucose accumulated continuously during the entire culture period. This finding suggests that the rate of glucose production by amylase surpassed the glucose consumption rate of the organism.

Amylase activities were measured in cell free cul¬ ture broth of C. thermohydrosulfuricum in order to deter¬ mine whether cell lysis occurred during growth. Gluco¬ amylase and pullulanase activities in supernatants were not detected in the early to mid exponential growth phase, but after that, the activities gradually increased. Supernatant glucoamylase and pullulanase activities after cultivation for 42 hours, were 0.22 and 0.27 units/ml, respectively, which implied extensive release of amylases into the culture fluid by cell lysis. This was confirmed by microscopic observation of lysed cells in the culture. Therefore, cell lysis as well as repression of amylase synthesis caused by accumulation of glucose during growth may account for relatively low glucoamylase and pullula- nase production by C. thermohydrosulfuricum in mono¬ culture fermentations.

A dramatic improvement in starch metabolism was observed when co-cultures were grown at pH 7.0 under conditions of continuous gassing. Starch was rapidly degraded during the fermentation time course and almost completely consumed at 28 hours. Co-culture fermenta¬ tions consumed starch more rapidly and completely than mono-cultures grown under similar conditions. Most importantly, the co-culture produced more cells, total amylases and ethanol. The activities of β-amylase, glucoamylase and pullulanase increased in relation to cell growth. Reducing sugar and glucose accumulation was limited during the co-culture fermentation time course. The co-culture also showed both an enhanced rate and yield of ethanol production. It is interesting to note that after depletion of sugar substrates, lactate was consumed. Balanced growth was achieved by both species

in this co-culture as evidenced by: the pattern of growth; complete starch consumption; the simultaneous production of three amylases; and, the ratio of end product formation. The most remarkable feature of the co-culture fer¬ mentation was the significant enhancement of glucoamylase ^" and pullulanase production. Growth of both species may have competitively removed low molecular substrates (e.g., glucose) liberated by the coordinate action of amylases, resulting in alleviation of catabolite repres¬ sion of enzyme synthesis. The amylase synthesis in these two species is repressed by glucose. Also, other sub¬ strates with o-l, 6 linkages formed by starch hydro¬ lysis which were not attacked by β-amylase, may have more efficiently induced the synthesis of glucoamylase and pullulanase. The glucoamylase and pullulanase activi¬ ties in cell free co-culture broth at 42 hours were 0.02 and 0.07 units/ml, respectively. This indicated that lysis of C. thermohydrosulfuricum still occurred in co-culture fermentations but not at such significant levels as in mono-culture fermentations.

Higher starch concentrations (5%) influenced the fermentation time course of co-cultures grown at pH 7.0, under continuous gassing. Under these conditions, starch was rapidly degraded but not completely consumed. End products were formed very rapidly in relation to the high growth rate until high concentrations of ethanol ^' (1.6% v/v) was accumulated. Growth of C. thermohydrosu1furicum and C. thermosulfurogenes wild type strains was inhibited by this concentration of ethanol. Glucose concentration was maintained at fairly low levels during the time course. The final yield of β-amylase but not gluco¬ amylase or pullulanase was enhanced. It is notable that β-amylase activity increased very rapidly after the stationary growth phase. The evaporated ethanol removed by gassing the co-culture was measured to determine total ethanol produced by in this co-culture fermentation.

About 84% of starch was fermented by the co-culture, and the product yield (i.e., mM ethanol produced per mM glucose as a starch consumed) was 1.7. The maximum specific production rate was 15.0 mmoles of ethanol per g-cell weight per hour.

Total Soluble End Products Formed by Co-culture Fermentations of C. thermosulfurogenes and C. thermohydrosulfuricum on 5% Soluble Starch Medium Starch Consumption End Product Formation (mM)

(mM as Glucose) Ethanol Acetate Lactate

252 F Feerrmmeennttoorr 2 26655 48 90

Reservoir 19

Total 284

The experiment was conducted in a 14 liter fermentor that contained 8 1 of modified TYE medium, and, which was gassed with N^/CO^ (95:5) and incubate with pH control at 60°C with mixing. Total ethanol values in the reservoir represent those recovered in the off-gas vapors.

Influence of β-amylase on C. thermohydrosul¬ furicum starch fermentations. These experiments were performed to verify whether the dramatic improvement of starch metabolism in co-cultures was caused by the co¬ ordinate action of amylases alone or by both enzymatic and metaolic cooperations. The starch fermentation time course of C^ thermosulfurogenes when grown in the presence of added β-amylase from C^ thermosulfurogenes at pH 7.0 under conditions of continuous gassing was evaluated. Under these conditions, starch was not completely hydro- lyzed, similar to mono-cultures. However, growth and end product formation were fairly improved in comparison to the mono-culture of C_ _; _ thermohydrosulfuricum. Notably, amylase production was not improved. Glucose accumulated in large quantities after the mid logarithmic growth phase and this may have resulted in catabolite repression of glucoamylase and pullulanase synthesis. Therefore,

these results imply that improvement of starch metabolism by co-culture fermentation was achieved by the coordinate action of amylolytic enzymes as well as metabolic coopera¬ tion between the two species. The foregoing data clearly show that co-culture fermentations comprised of C. thermohydrosulfuricum and C. thermosulfurogenes dramatically enhance the rate and yield of transforming starch into both amylases and ethanol. Furthermore, ethanol and three thermostable amylases - glucoamylase, pullulanase and β-amylase are produced in high yields by a single step fermentation system.

The following points summarize the general signifi¬ cance of the results in relation to metabolic regulation and application of amylase activities that were expressed in the co-culture starch fermentation.

1. The rate of starch saccharification was enhanced in co-culture as a consequence of sharing amylase activities to improve rate limiting steps..

2. Total starch saccharification activity was en¬ hanced in co-cultures as key modulators of amylase synthesis (e.g., glucose) were decreased.

3. The total starch metabolism physiology (i.e., growth, enzyme production and fermentation product formation) of the thermoanaerobic species was enhanced in co-culture.

4. The rate and yield of ethanol production from starch was increased in co-culture. 5. Co-culture starch fermentations of these species may have utility in continuous production and recovery of ethanol at high temperature and re¬ duced pressure, or in the production of thermo¬ stable amylases for industrial use. The amylolytic enzymes such a β-amylase, pullula¬ nase and glucoamylase which are needed in starch trans¬ formation industries are generally not thermostable at

operating conditions used for starch conversion into soluble sugars. Therefore, simultaneous and enhanced production of these thermostable amylases by co-culture starch fermentations provide an inexpensive method for producing industrial amylases needed in both starch saccharification and ethanol production processes. In addition, ethanol can be recovered for its by-product value. The cell bound pullulanase and glucoamylase may be practically applicable for glucose production from starch or maltodextrins; whereas, β-amylase may be useful for maltose production.

The co-culture fermentation has particular advantages for industrial ethanol production provided the ethanol can be economically removed by continuous reduced pressure distillation. Starch is less expensive than glucose as a feedstock substrate and the co-culture directly produces its own amylase. The utility costs assumed for the co-culture in relation to: not needing aeration; limited mixing; use of metabolic heat; and, the potential to recover ethanol continuously at process temperatures by reduced pressure distillation imply substantial savings. Notably, the productivity rate for starch transformation to ethanol (0.7 g ethanol/g cell/h) is close to the values reported for glucose transformation to ethanol by Saccharomyces cerevisial (1 g ethanol/g cell/h). The productivity rate for starch conversion to ethanol by thermoanaerobes should be further enhanced by use of mutants that display higher ethanol tolerance and yield. It will be apparent to those skilled in the art that a number of modifications and changes may be made without departing from the spirit and scope of the present inven¬ tion. For example, although in the experimental work the substrate was purified starch other less expensive and more abundant substrates can be used. In addition, pH's and other conditions can be modified within limitations. It will be appreciated by those skilled in the art that the scope of the invention should not be limited to the

specific strains of organisms described because genetic variants produced by mutation or genetic recombination techniques and having the enzyme producing capabilities and identifying characteristics of the specifically described organisms can be used. Therefore, the invention is not to be limited except by the claims which follow: