FERMENTATION OF GLUCOSE AND XYLOSE IN CELLULOSIC BIOMASS USING GENETICALLY MODIFIED SACCHAROMYCES CEREVISIAE AND A SIMULTANEOUS SACCHARIFICATION AND

CO-FERMENTATION PROCESS

DESCRIPTION,

Technical field

The present invention relates to simultaneous saccharification and co-fermentation of cellulosic biomass substrates hydrolysable to produce glucose and xylose using a genetically 0 engineered Saccharomyces cerevisiae

Background of the invention

Bioethanol is a fuel that not only reduces negative environmental impacts from the transportation sector including reductions in greenhouse gas emissions but it can also be 5 manufactured from a wide range of available renewable feedstocks using established industrial processes.

When producing bioethanol from lignocellulosic material there are two factors that are of outmost importance for the price of the ethanol produced; to effectively use all sugar in the 0 biomass and to produce a slurry after the fermentation that contains high concentrations of ethanol. Since the raw material cost contributes greatly to the final ethanol price, an efficient use of all the sugars available will decrease the production cost. It is thus very important that both the hexoses and pentoses in the raw material can be hydrolysed without any larger losses and that they then can be co-fermented to ethanol. Refining the ethanol produced will 5 be done using distillation. A final slurry with higher ethanol concentration will thus demand much less energy for this down steam step than a slurry with a lower ethanol concentration.

If using a modified Saccharomyces cerevisiae strain capable of fermenting xylose, in fermenting a medium containing glucose and xytose, such as the TMB3400 mentioned below 0 expression of ethanol will only proceed to a certain yield.

Summary of the invention

The aim of the present invention is thus to improve the yield of expression of ethanol when fermenting a biomass comprising glucose and xylose. According to the present invention this

will be possible by using a simultaneous saccharification fermentation process.

Detailed description of the present invention

The present invention thus relates to a method for the manufacture of ethanol by fermenting a xylose and glucose containing biomass using a strain of Saccharomyces cerevisiae, wherein the method encompasses

a) using a strain of Saccharomyces cerevisiae being capable of fermenting pentoses, including xylose, to ethanol,

b) subjecting a slurry of biomass and said Saccharomyces cerevisiae strain to a simultaneous saccharification and co-fermentation of glucose and xylose, and

c) isolating ethanol expressed from the fermented slurry obtained.

In the present disclosure corn stover is used as an example of raw material. Corn stover is an abundant agricultural by-product with a low commercial value, thus it is a good raw material for bio-ethanot production. As most other agricultural by-products, corn stover consists of mainly two types of sugars: glucose and xylose. Both these sugars can be hydrolysed at high yield using steam pretreatment and subsequent enzymatic hydrolysis and the glucose can then, using Saccharomyces cerevisiae (a yeast strain used widely within the industry), be fermented to ethanol at high yield, preferably in a combined hydrolysis and fermentation (SSF- simultaneous saccharification and fermentation), Today, no industrial available organism exists that can co-ferment glucose and xylose at high yield and high productivity the same time as it is tolerant to the harsh industrial conditions. The pretreatment indicated above is not necessary per se, but a common enzymatic hydrolysis may be used. It is however, time-consuming if the biomass raw material is a fresh wooden material. Even a chemical hydrolysis is possible as such.

Saccharomyces cerevisiae has been used for a long time in ethanol producing industry. It has also been shown to be a very good and tolerant yeast for fermentation of lignocellulosic hydrolysate to ethanol. However, S. cerevisiae does not, in its natural occurring state, ferment xylose. Thus, a lot of work has been put into genetically modifying it and other yeast strains to

ferment xylose.

The intent of this work was to evaluate the yeast strain TMB3400, a xylose-utilization

S. cerevisiae strain, generated by random mutagenesis at the Department of Applied Microbiology, Lund University in Lund, Sweden on industrial process applicable condition.

In previous studies SSF has been shown to be superior to separate hydrolysis and fermentation (SHF) at elevated dry matter concentration when the ethanol yield is the main priority. Combining two process-steps also results in a lower capital cost and the fact that the ethanol concentration is high during the SSF makes the risk of contamination lower. On the other hand, the temperature optimum for the yeast and the enzymes used differ, which means that the conditions used can't be the optimal for both the enzymes and the yeast. A further advantage using the present method is the low formation of xylitol, which formation using a batch fermentation of a medium containing xylose and glucose as such can be rather high, thereby preventing ethanol formation.

This study showed that TMB3400 was co-fermenting glucose and xylose during simultaneous saccharification and fermentation (SSF) of steam pretreated corn stover with lower by-product formation than during fermentation of xylose in well-defined solutions. Furthermore, this work showed that TM33400 is very tolerant to harsh process conditions and that the process configuration 3SF is to be preferred when it comes to co-fermentation of glucose and xylose due to the increased fermentation rate of xylose by TMB3400 in the presence of the low amount of glucose constantly liberated in the hydrolysis. The upper limit of glucose in the hydrolysate should not exceed 3 g/l, preferably not 2 g/l, more preferably not 1 g/l, still more preferably 0,6 g/l.

Two studies were performed. In study A, TMB3400 and its mother-yeast USM2I was compared with ordinary baker's yeast in SSF at 5% WIS (water insoluble substances). USM2I was shown to ferment glucose as effectively as ordinary baker's yeast and TMB3400 fermented both glucose and xylose efficiently almost without any xylitol production.

In study B, TMB3400 was used in fed-batch SSF at final concentration of water insoluble substances (WIS) of 12 % resulting in an ethanol concentration in the fermented slurry of 17.5 g/l. This is the highest concentration in the slurry after an SSF ever published.

Raw material

All the investigations were made using corn stover as raw material. After collection, the stover was chopped, air-dried and then stored at room temperature. The sugar and lignin content of the raw material were determined according to National Renewable Energy Laboratory's (NREL) methods.

Steam pretreatment

Prior to pretreatment the material was milled and sieved and the fraction between 2 and 10 mm was recovered and used. Since the raw material used was very dry (95 % DM, owing to the dry storage) it had to be remoistened. This was done by presteaming with saturated steam at 10000 for 30 mm, after which the material was immersed in cold water and then allowed to drain off. The dried and re-wetted material may differ slightly from fresh, moist material since some collapsed pores may not regain their original shape, and the pore size has an impact on the subsequent enzymatic hydrolysis.

To adjust the dry matter content the material was pressed and the water pressed off was measured to assure that the drj matter content landed in the right region whereupon the material was thoroughly mixed to assure an even moisture distribution. To double-check the dry matter content, it was also 'measured by drying in 10500 till constant weight.

The rewetted and pressed com stover was impregnated with SO ₂ (used as a catalyst to enhance reaction speed) in plastic bags and the uptake was measured by weighing the material before and after impregnation. Adding 3% SO ₂ (w/w, based on the water content of the corn stover) resulted in an actual uptake of around 2% SO ₂ . The impregnated raw material was steam pretreated in a 10-1 reactor and was then collected for subsequent analysis and SSF-tests.

The composition of the pretreated solid fraction was determined in the same way as was the composition of the raw material., Analyses was made on the liquid fraction after pretreatment with respect to monomeric sugars, oligomeric sugars and inhibitors such as acetic acid, furfural and 5~hydroxymethyl-2-furfural (HMF).

Simultaneous saccharification and fermentation (SSF)

The SSF-tests were performed in 2 I fermentors (Labfors® Infers AG, Switzerland) with a working weight of 1.5 kg using the whole slurry from the pretreatment. The enzyme mixture used was a commercial cellulase mixture, Cefluclast I 51(65 FPU/g and 17 β-glucosidase IU/g} supplemented with the β-glucosidase preparation Novozyme 188 (376 β-glucosidase IU/g), both kindly donated by Novozymes A/S (Bagsvaerd, Denmark). The enzyme activity used in all experiments was 15 FPU and 25 IU/g water insoluble solids. The yeasts used was compressed baker's yeast, Saccharomyses cerevisiae from Jastbolaget AB, Rotebro (hexose fermenting) Sweden, USM21 (glucose fermenting) and TMB3400, a glucose and xylose fermenting yeast generated by random mutagenesis. All yeast was grown on the substrate used in the SSF.

The SSF-tests were performed under semi-sterile conditions. After the addition of the material to the fermentor, water was added to adjust the dry matter concentration, the pH was adjusted to 5.0 with 25% NH ₃ and the fermentor was sterilised. The equipment was left to cool down over night. Nutrients were added so that the concentrations in the fermentor were 0.5 g/l (NH ₄ ) ₂ HP0 ₄ 0.025 g/l MgSO ₄ x7H ₂ O and 1.0 g/l yeast extract. The experiments were run at 35°C for 96 hours with the pH maintained at 5.0 by 5% NH ₃ . Samples were withdrawn after 0, 2, 4, 6, 8, 24, 28, 32, 48, 72 and 96 hours, and analyzed for ethanol, sugars, glycerol, acetic acid, lactic acid and sugar degradation products.

Yeast cultivation Inoculum culture

A small amount of pure Baker'.s yeast culture (Jastbolaget, Rotebro, Sweden) from an agar plate was added to a 300 ml Erlenmeyer flask, which contained 100 ml of sterile medium with a pH of 55. The medium composition was as follows; Glucose: 16.6 g/l, (NhU) ₂ SO ₄ ; 7.5 g/l, KH ₂ PO ₄ : 3.5 g/l, MgSO ₄ : 075 g/l, Trace Metal Solution: 10 ml/I and Vitamin Solution: 1 ml/I. The culture was incubated at 30°C for 24 h. The Erlenmeyer flask was sealed with a cotton plug.

Aerobic batch cultivation

In order to produce cell mass before the fed-batch phase, batch cultivation on glucose, with a working volume of 0.9 I, was carried out at 30 ⁰ C under sterile conditions. A 2.5 1-bioreactor was used for the batch and fed-batch cultivation phases (BiostatA, B. Braun Biotech International, Melsungen, Germany). The medium contained the following components

Glucose: 22 g/l, (NH ₄ ) ₂ S0 ₄ : .3 g/l, KH ₂ PO ₄ : 11 g/l, MgSO ₄ : 2.6 g/l, trace metal solution: 34 ml/l and vitamin solution: 5 ml/I. The pH was maintained at 5.0 by automatic addition of 3 M NaOH. The stirrer speed was kept at 750 rpm and the aeration was maintained at 1 .4 1/mm. Adding 45 ml of inoculum culture started the cultivation. The dissolved oxygen concentration (DOT) was measured with a DOT-electrode, The DOT never decreased below 20 % during the batch and fed-batch phases.

Aerobic fed-batch cultivation

Subsequent to depletion of the ethanol produced during the glucose consumption phase, feeding of the pretreatment liquid was started. The liquid was enriched with 64 g glucose/1 and pH-adjusted to 4.7 with NaOH. 1.0 I of glucose enriched pretreatment liquid was added during 16 h. The feed rate was initially set to 0.040 l/h and was increased linearly to 0.10 l/h after 16 h. The pH was maintained at 5.0 by automatic addition of 3 M NaOH. The stirrer speed was kept at 700 rpm and the aeration was maintained at 1.5 I/mm.

Cell harvest

When the feeding was stopped the cultivation liquid containing the yeast was transferred from the fermentor into a sterile glass flask. The cultivation liquid was centrifuged (1,000 g) in 700 ml containers using a HERMLE Z 513 K (HERMLE Labortechnik, Wehingen, Germany). The supernatant was discarded and the pellets were transferred to a sterile glass flask. Sterile 0.9% NaCI-solution was added in order to obtain a cell suspension with a cell mass concentration of about 75 dw /I. The time lapse between the end of the fed-batch phase and the addition of the harvested cells to the 3SF-fermentations was less than 2 h.

Analysis

The amounts of monosaccharides and inhibitors were determined by HFLC. Glucose, arabinose, galactose and xylose were separated using an Aminex HPX-87-Pb column (Bio- Red, Hercules, USA) at 85°C and a flow rate of 0.5 ml/mm with water as eluent. Glucose, arabinose, lactic acid, glycerol, acetic acid, ethanol, HMF and furfural were separated on an Aminex HFX-87-H column at 65 ⁰ C using 5 mmol/l H ₂ SO ₄ as eluent at a flow rate of 0.5 ml/mm. The samples from the liquid fraction after pretreatment were neutralised using CaCO ₃ and Ba(OH) ₂ and diluted 3 times. Ba(OH) ₂ was used to precipitate sulphur CaCO ₃ was used for the final pH-adjustment. All samples were filtered through a 0.20-pm filter before analysis.

To measure the total amount of sugars (monomers and oligomers) in the liquid fraction after pretreatment, the samples were hydrolysed with 4% H ₂ SO ₄ at 121 ⁰ C for 1 h and then neutralised using Ca(CO) ₃ . They were then diluted six times and analysed with HPLC using an Aminex l-IPX-87-Pb column as described above.

The amount of acid-soluble lignin was determined using an absorption spectrophotometer at a wavelength of 205 nm with a 4 % H ₂ SO ₄ solution as a reference.

Result and discussion Raw material

The sugar content in the raw mate al used is presented in Table I The raw material used in the two studies was delivered from the same place in Italy but at different times thus the raw material content in the two batches differ slightly, as can be expected, but the contents are still in a normal range for corn stover.

Table I Comoosition of com stayer expressed as % of dry matter

Origin Glucan Xylan Arabinan Galactan

Study A 42.6 22.7 2.8 1.4

Steam pretreatment

Several pretreatment batches were performed for each study at a temperature (200°C) and a dwell time (5 mm) optimised in earlier work by the same authors. The batches were collected and mixed thoroughly before being analysed and thus pretreated material with the same composition and dry matter content were used throughout an entire study In Table 2 the WIS content in the slurry, the content of glucose and xylose in the solid part and the content of glucose and xylose liquid part is presented.

Table 2. WIS in slurry Content in the solid part (%) Content in the liquid part (g/l)

(%) GlucanXylan Glucose Xylose

Study A 13 .7 59 .4 9 .2 7 36

SSF

In study A, Xylose-utilizing Saccharomyces cerevisiae TMB3400, generated by random mutagenesis, and its mother-yeast USM21 was compared with ordinary baker's yeast in SSF at 5% WIS. All tests were run at duplicates and all duplicates showed very good conformance. SSF with ordinary baker's yeast resulted in an ethanol yield based on the glucose content in the raw material of 78 and 79% respectively (14.3 and 14.4 g/l). Some lactic acid formation was observed in both duplicates after 72 h but since the fermentation was finished by then, it had no effect on the outcome of the test. The xylitol production was moderate giving 3.4 and 3.9 g/l respectively and the xylose level at the end of the fermentation was 8.5 g/l for both duplicates.

SSF with USM21 resulted in an ethanol yield based on the glucose content in the raw material of 81 and 71 % respectively (14.7 and 13.0 g/l). The difference in ethanol production was a result of lactic acid formation observed after 32 h in the second SSF why the ethanol production ceased at that point. The xylitol production was again moderate and somewhat lower than with baker's yeast giving 3.0 and 2.7 g/l respectively and the xylose level at the end of the fermentation was 8,7 and 9.0 g/l.

SSF with TMB3400 resulted in an ethanol yield based on the glucose AND the xylose content in the raw material of 64 and 62 % respectively (18.2 and 17.5 g/l). After 96 h there were 2.1 and 1.5 g/l xylose left in the two duplicates so approximately 7 g/l xylose had then been fermented to an extra 3.2 g/l ethanol → 46 g ethanol/g xylose (compared with 0.51 g/g theoretically). Wahlborn et al., who genetically engineered TMB3400, reported an ethanol yield of 0.25 g ethanol/g xylose and Kuyper et al. reported an ethanol yield of 0.42 g ethanol/g xylose using S. cerevisiae where a heterogenous xylose isomerase (EC 5,3.1.5) is functionally expressed. However both of these yields were obtained in well-defined xylose containing solutions. Thus, TMB3440 is fermenting xylose at better yields when the fermentation is conducted in SSF where several undefined pretreatment byproducts are present. Furthermore, a small amount of glucose is constantly liberated during the enzymatic hydrolysis of the fibrous material that could be affecting the fermentation of xylose. This phenomena (glucose addition affecting the xylose fermentation) was indicated already 1985 by Jeffries. However, Jeffries. did only observe the effect of glucose addition on xylose fermentation under aerobic conditions and not with an acid hydrolysate.

Only a small formation of lactic acid was observed in the last duplicate after 96 h indicating that SSF also suppresses lactic acid bacteria. The xylitol production was surprisingly low giving 1.1 and 1.0 g/l respectively. This can be a result of co-factor regeneration during xylose-to-ethanol fermentation giving the yeast no reason to produce xylitol.

Figure Legends

Figure 1 denoted SSF 5 % with cultivated BY shows the result of fermenting 5 % SSF with bakers yeast with regard to outcome of ethanol, glucose, xylose, galactose, arabinose, xylitol, lactic acid, glycerol, acetic acid, HMF and furfural, respectively.

Figure 2 denoted SSF 5 % with cultivated USM21 shows the result of fermenting 5 % SSF with USM21 with regard to outcome of ethanol, glucose, xylose, galactose, arabinose, xylitol, lactic acid, glycerol, acetic acid, HMF and furfural, respectively.

Figure 3 denoted SSF 5 % with cultivated TMB3400 shows the result of fermenting 5 % SSF with TMB3400 with regard to outcome of ethanol, glucose, xylose, galactose, arabinose, xylitol, lactic acid, glycerol, acetic acid, HMF and furfural, respectively.

Figure 4 denoted SSF 5 % with cultivated BY (2) shows the result of fermenting 5 % SSF with bakers yeast with regard to outcome of ethanol, glucose, xylose, galactose, arabinose, xylitol, lactic acid, glycerol, acetic acid, HMF and furfural, respectively.

Figure 5 denoted SSF 5 % with cultivated USM21 (2) shows the result of fermenting 5 % SSF with USM21 with regard to outcome of ethanol, glucose, xylose, galactose, arabinose, xylitol, lactic acid, glycerol, acetic acid, HMF and furfural, respectively.

Figure 6 denoted SSF 5 % with cultivated TMB3400 shows the result of fermenting 5 % SSF with TMB3400 with regard to outcome of ethanol, glucose, xylose, galactose, arabinose, xylitol, lactic acid, glycerol, acetic acid, HMF and furfural, respectively