A thermo- and toxin-tolerant kefir yeast for biorefinery and biofuel production

A thermo- and toxin-tolerant kefir yeast for biorefinery and biofuel production

Jui-Jen Chang

, Cheng-Yu Ho

, Chi-Tang Mao

, Nathan Barham

, Yu-Rong Huang

, Feng-Ju Ho

, Yueh-Chin Wu

, Yu-Han Hou

, Ming-Che Shih

, Wen-Hsiung Li

*, Chieh-Chen Huang

*

1

Department of Medical Research, China Medical University Hospital, Taichung 40402, Taiwan

2

Department of Life Sciences, National Chung Hsing University, Taichung 40227, Taiwan

3

Molecular and Biological Agricultural Sciences Program, Taiwan International

Graduate Program, National Chung-Hsing University - Academia Sinica, Taipei 11529, Taiwan

4

Biodiversity Research Center, Academia Sinica, Taipei 11529, Taiwan

5

Graduate Institute of Biotechnology, National Chung-Hsing University, Taichung 40227, Taiwan

6

Agricultural Biotechnology Research Center, Academia Sinica, Taipei 11529, Taiwan

*Co-corresponding authors Chieh-Chen Huang

Tel:886-(4)-22840416 ex 405 Fax:886-(4)-22874740 Email addresses:

Jui-Jen Chang ([email protected]) Cheng-Yu Ho ([email protected]) Chi-Tang Mao ([email protected])

Nathan Barham ([email protected]) Yu-Rong Huang ([email protected]) Feng-Ju Ho ([email protected])

Yueh-Chin Wu ([email protected]) Yu-Han Hou ([email protected]) Ming-Che Shih ([email protected]) Wen-Hsiung Li ([email protected]) Chieh-Chen Huang ([email protected]) 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23 24

25

26 27

28

29 30

31

Abstract

Although biorefinery has become a common concept to convert biomass into biofuels and value-added chemicals for better cost-performance, good microbial hosts that can be used to implement the concept are still wanting. In this study, a Kluyveromyces marxianus yeast, named KY3, was isolated from a Taiwanese kefir microbial consortium. We showed that KY3 could grow on a broad spectrum of substrates, including hexose and pentose sugars. It is heat and toxin tolerant, can grow under a wide range of pH values (pH 2.5-9), and shows a high ethanol production rate at elevated temperatures. It also can produce value-added aromatic chemicals, such as 2- phenylethylethanol and 2-phenylethyl acetate, during the fermentative process. A genetic transformation was achieved in KY3 to express a rumen fungal β-glucosidase gene, and the transgenic host (KY3-NpaBGS) could efficiently convert cellobiose to ethanol. Furthermore, it was shown that a novel dual-microbe co-culture system of Bacillus subtilis and KY3-NpaBGS can be employed for bioethanol production from cellulosic material. Thus, KY3 has a high potential to be a good host for biorefinery.

Keywords: biorefinery, kefir yeast, Kluyveromyces marxianus, dual-microbe co- culture, cellulosic ethanol

1. Introduction

Plant biomass is one of the most abundant renewable resources on Earth [1] and is considered an essential building block for developing a sustainable society. However, plant biomass such as cellulose is tough to be broken down as it comprises polymerized crystalline structures, making the conversion processes costly for producing second- 32

33 34 35 36 37 38 39 40 41 42 43 44 45 46

47 48 49

50

51

52

53

54

generation biofuels (SGBs). Therefore, “biorefinery” becomes a common concept to solve this problem by integrating biomass conversion processes to produce SGBs and value-added chemical by-products [2]. This concept can potentially reduce the cost for biofuel production [3, 4].

For SGB production, the consolidated bioprocessing (CBP) that integrates cellulase production, cellulose hydrolysis and fermentation in one single process is preferred over other processes because of its operational simplicity and potential low cost [1]. However, for CBP there is currently no “superbug” that can simultaneously perform efficient degradation of lignocellulosic materials, biofuel conversion, and production of valuable by-products and also can tolerate toxins produced by the chemical pretreatment to release hexose and pentose sugars from lignocellulose [5, 6].

In the past, Saccharomyces cerevisiae has been considered as an ideal host for CBP development, mainly because of its high fermentative capability, its powerful genetic manipulation tools, and good knowledge of its biology [7, 8]. Successful applications of S. cerevisiae include fuel and chemical productions [2]. However, there have been problems of enzymatic function and secretion in expressing cellulase genes in S. cerevisiae because of inadequate posttranslational modifications [9-11]. In addition, a process with a higher temperature is usually preferred for increasing the efficiency of enzyme reaction and bio-production [1], but S. cerevisiae does not grow well at higher temperatures. For developing an efficient CBP, a host that possesses a broader range of culturing conditions, a broader spectrum of substrates, and good toxin- and thermo- tolerance is still lacking.

In nature, many microorganisms can function effectively only when they cooperate with other microorganisms; examples include interspecies hydrogen transfer 55

56 57 58

59 60 61 62 63 64 65

66 67 68 69 70 71 72 73 74 75 76

77

78

within methanogenic systems and degradation of trichloroethylene compounds [12]. For biomass utilization, microbial consortia, such as rumen bacterial consortia or termite gut consortia, have been investigated for lignocellulolytic purposes. In our previous studies, some major microbes in these consortia have been identified and have been re- constructed as bacterial consortia for applications in biofuel production [13-15]. As an artificial consortium, aerobic Bacillus strains and anaerobic Clostridium strains have been developed as a syntrophic co-culture system that can perform SFF for bio-fuel or bio-hydrogen production [15-17]. Although the co-culturing of both saccharification and fermentation eukaryotic organisms has been successfully used in the production of alcoholic beverages such as Japanese sake [18], cellulolytic fungi do not grow as fast as bacteria and the saccharification efficiency are still too low for producing enough sugars for fermentative yeast.

To hunt for a suitable yeast candidate for CBP and for a possible bacterium-yeast co-culture system, we studied dairy kefir grains that include a combination of bacteria and yeasts that form a natural consortium. Dairy kefir yogurt, originally from the Caucasus Mountains, is a kind of mildly alcoholic milk fermented by a mixture of microbes [19]. The compatible microbial strains contain lactic acid bacteria (lactobacilli and lactococci), acetic acid bacteria, mycelial fungi and yeast species [20, 21]. Kefir grain yeasts produce not only alcohol but also higher aromatic compounds to provide kefir’s distinctive characteristics [22]. In this study, we isolated a Kluyveromyces marxianus strain, named KY3, from the microbial consortium of a Taiwanese kefir grain. Our experimental characterizations suggested that KY3 has a high potential for CBP because it has a broad spectrum of substrates, is heat and toxin tolerant, and can live under a wide range of pH values (2.5-9). Moreover, we found it can readily cooperate with other cellulolytic bacteria and can produce valuable by-products.

79 80 81 82 83 84 85 86 87 88 89 90

91

92

93

94

95

96

97

98

99

100

101

102

103

2. Material and Methods

2.1 Yeast strains isolation and culturing

Inoculation of kefir grains in milk was conducted at 18°C for 24 h to produce kefir yogurt. The microbial community became stable after repeated sub-culturing, and the main microorganisms were amplified. The cultures were then diluted and spread on media plates. After 10-fold serial dilutions, an aliquot of each dilution was spread onto Lactobacilli MRS Agar (Difco NO. 288210, USA) [23] and YPAD plates (10 g L

yeast extract, 20 g L

peptone, 24 mg L

adenine hemisulfate, 20 g L

glucose and 20 g L

agar), and the plates were incubated at 20°C under either aerobic or strict anaerobic conditions to screen the microorganisms. Reference yeast strains such as Kluyveromyces marxianus CBS600, K. marxianus CBS6432, K. lactis GG799, Saccharomyces cerevisiae BY 4741, S. cerevisiae YJM145, S. cerevisiae YJM189, Pichia pastoris GTS115, P. stipites CBS 6054, Lactobacillus, E. coli, and Bacillus subtilis 168, were used for comparison in different assays.

2.2 PCR–DGGE Analysis

The denaturing gradient gel electrophoresis (DGGE) technique, which separates sequences by G-C content, was used to profile microbial communities. For the analysis of bacterial diversity, the EUB968F and UNIV1392R primer pair [24] was used to amplify a partial sequence of the 16S rRNA gene to create a DNA fragment suitable for DGGE analysis. For analysis of yeast diversity, PCR amplification of the 26S rRNA gene was conducted using the universal NL1 and LS2 primer pair [25]. DGGE analysis of PCR products followed the method of [26], using the DCode system (Bio-Rad Laboratories, Hercules, CA, USA). Optimal separation of the bacterial and yeast DNA fragments in a sample was achieved using a 35–65% urea–formamide denaturant 104

105

106 107 108 109 110 111 112 113 114 115 116 117

118

119

120

121

122

123

124

125

126

127

gradient gel. The PCR products were extracted and purified from the gel bands using the Centrilutor microelectroeluter system (Millipore, Rockville, MD, USA). After re- amplification under the reaction conditions described above, the resulting PCR products were cloned for sequencing. BLAST searches were conducted using the NCBI web site to identify the species from which each band was derived. The primer pairs used for PCR were listed in the Supplementary file.

2.3 Gene sequencing and phylogeny reconstruction

The genomic DNA of the microbes isolated from kefir grains was extracted and subjected to PCR analysis with the primer pairs 16Sf/16Sr [27], NS-1/NS-8 [28] and NL1/LS2 [25] for amplification of 16S rRNA, 18S rRNA and 26S rRNA fragments, respectively. After initial heating at 94°C for 3 min, 30 cycles of heating at 94 °C (30 s), 55 °C (30 s) and 72 °C (1 min) were performed for both the 16S and 18S PCR reactions.

For amplification of 26S rRNA, the reaction mixtures were heated to 95°C for 5 min, then subjected to 30 cycles of heating at 94°C for 60 s, annealing at 52°C for 45 s and extension at 72°C for 2 min, followed by a 7 min extension at 72°C. PCR products from bacterial isolates were cloned into the yT&A cloning vector and subjected to DNA sequencing. The sequences obtained were aligned with sequences of various members of the bacterial phylum, whose gene sequences were obtained from GenBank by a BLAST search. A yeast phylogeny was reconstructed by the neighbor-joining method [29]. The primer pairs used for PCR were listed in the Supplementary file.

2.4 Metabolite determination

Ethanol productivity was analyzed using high-performance liquid chromatography (Waters, Waters 600E system controller, USA) with a Refractive Index detector (RI) (Shodex RI-201H Refractive Index Detector, USA) and an ICsep ION-300 128

129 130 131 132 133

134

135 136 137 138 139 140 141 142 143 144 145 146 147

148

149

150

151

column (Transgenomic inc. USA). 0.0085N sulfuric acid was used as the mobile phase.

Running conditions included heating the column to 50°C and the flow rate at 0.3 ml/min. Each liquid sample was sterilized by filtering through a 0.22-μm filter, and the injection volume was 50 μl. The 2-phenylethanol and 2-phenylethyl acetate productivities were analyzed using gas chromatography (Thermo Finnigan TRACE GC / POLARISQ) with a DB-1701 column (0.25 μm film thickness, 60 m x 0.25 mm i.d.). The running conditions included heating the column to 80-300°C with a heating rate of 5°C per min, an injection temperature of 230°C and a detection temperature of 300°C. The reducing sugar was detected using the 3,5-dinitrosalicylic acid (DNS) method [30]. The standard curve was made using 2-fold serial dilutions of a pure chemical sample. Each fermentation experiment and subsequent analysis was repeated three times.

2.5 Substrate spectrum and inhibitor tolerance assays

The media for the substrate spectrum assay were prepared in synthetic complete (SC) medium agar, and a 2% carbon source, such as glucose, galactose, fructose, mannose, raffinose, sucrose, glycerol, lactose, xylose, cellobiose, arabinose, and CM cellulose, was added for yeast growth. The media for the pH tolerance and toxin tolerance tests were prepared in YP medium (10 g L

yeast extract and 20 g L

peptone). Fifty milliliters of the pre-culture of the bacterium being tested was inoculated into 5 ml of YP medium containing 2% glucose at different pH levels (pH 2-9) and cultured at 30°C. The furfural tolerance was measured under different toxin concentrations (12, 24, 36 and 48 mM 2-furaldehyde). Cell densities were measured by determining the OD value of each sample at 600 nm using a spectrophotometer (Ultrospec 2100 pro; Amersham Bioscience). The experiment was repeated three times for each sample.

152 153 154 155 156 157 158 159 160 161 162 163

164

165

166

167

168

169

170

171

172

173

174

175

176

2.6 Heterologous β-glucosidase gene transformation and activity assay

To express a heterologous β-glucosidase gene in K. marxianus KY3, a commercial expression system that has been developed in K. lactis GG799 (K. lactis Protein Expression Kit, New England Biolabs) was used. A β-glucosidase gene in the rumen fungus Neocallimastix sp. W5 was amplified via specific primers (Supplementary file) and inserted into the multiple cloning sites (MCS) of the plasmid pKLac2; ampicillin- resistant E. coli clones were then selected. The transformation method was modified according to the previous studies [31, 32]. After recovery in 1 ml of YPAD for 45 min, the harvested cells were spread on YPG plates (1% Bacto Difco-Yeast Extract, 2%

Bacto Difco-Peptone and 2% Merck-galactose) for a two-day incubation. The correct insertion was confirmed by PCR (with a temperature profile of 95 °C for 10 min, 35 cycles of 95 °C for 1 min, 45 °C for 1 min and 72 °C for 2 min) using a TaKaRa ExTaq system (5 U μl

Taq, 10X Ex Taq Buffer and 10 mM dNTP) and specific primer pairs for verification (Supplementary file).

To quantify the β-glucosidase activity, KY3 and its transformants were cultured with YPG medium, and S. cerevisiae BY4741, P. pastoris GTS115, and their transformants were cultured under the method of [33, 34]. A 10 μL yeast culture supernatant was added to deep-well microtiter plates with each well containing 90 μL of 50 mM p-nitrophenyl-β-D-glucopyranoside (pNPG) (Sigma-Aldrich, St Louis, MO, USA), 0.05 M acetate buffer pH 5.0 and incubated at 30°C for 10 mins. The release of p-nitrophenol was determined by a fluorescent intensity reader with absorbance at 410 nm. The protein concentration was determined by the Bradford method.

3. Results and Discussion

3.1 Predominant microbes in kefir 177

178 179 180 181 182 183 184 185 186 187 188 189 190

191 192 193 194 195 196 197 198

199

200

Kefir consortia with the co-existence of prokaryotic and eukaryotic microbes represent attractive models. In this study, a culture-independent approach (DGGE) was used to analyze the members of the microbial community of Taiwanese kefir grains.

Total genomic DNA was extracted from kefir grains, and the prokaryotic and eukaryotic microbial compositions were analyzed using either universal 16S rRNA or 26S rRNA gene-targeted DGGE-PCR amplification (data not shown). After the DNA from the bands was directly eluted and the nucleotide sequences determined, Lactobacillus kefiranofaciens was found to represent the major species of prokaryotic microbes. For the eukaryotic microbes, although Kluyveromyces marxianus and Saccharomyces exiguus were both detected in the system, K. marxianus was much more prevalent. The fact that the DGGE patterns were stable during the sub-culturing period suggests that the microbial consortium system was stable (data not shown). The co-existence of Lactobacillus and fungi in a kefir system is a common phenomenon [20], and K.

marxianus is also known to co-exist with other yeasts in kefir systems such as the Tibetan kefir [20, 21, 35]. However, previous studies have not described the dominance of K. marxianus.

Based on the ribosomal gene DGGE data of kefir microbes, different media suitable for Lactobacillus and yeast, such as MRS and YPAD, were used for the isolation. Two prokaryotes, L. gallinarum and L. kefiranofaciens, and a dominant kefir yeast, named K. marxianus KY3 or simply KY3, were isolated from the kefir yogurt.

Using partial 26S rRNA and 18S rRNA sequences and the neighbor-joining method, the 26S and 18S rRNA trees of Kluyveromyces strains were reconstructed. For the 26S partial rRNA sequences, most of the strains were identical because the 500 base pair region is highly conserved (data not shown). However, the tree of the partial 18S rRNA sequences of KY3 and all known K. marxianus strains shows that KY3 is most similar 201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

to the Canadian cheese whey Kluyverromyces isolates CW3-8, CW3-4 and CW4-3, whose partial 18S rRNA sequences are identical (Fig. 1A). A further BLAST analysis showed that the KY3 sequence is most similar to that of CW3-8. The alignment of CW3-8 and KY3 shows that CW3-8 is identical to the K. marxianus consensus sequences, whereas KY3 differs from CW3-8 and the consensus sequences by 10 nucleotides and 8 indels (insertions and deletions) among the 1051 aligned positions (data not shown). Moreover, the 18S rRNA tree shows that KY3 has a longer branch length than those of the other K. marxianus strains (Fig. 1A). Therefore, we regard the Taiwanese kefir isolate K. marxianus KY3 as a new unique strain.

Furthermore, the metabolites of K. marxianus KY3 grown in the YPAD medium were examined by mass spectrometry. Interestingly, a shikimate-derived metabolite, N- acetyltyramine, was identified as one of the fermentative metabolites (peak 4) of KY3 (Fig. 1B). Previously, it has only been reported to be produced by an antarctic strain Microbispora aerata and a pathogenic strain Mycobacterium tuberculosis [36, 37]. N- acetyltyramine, as a soybean phytoalexin, has been shown to be toxic to the plant pathogen Cladosporium sphaerospermum [38]. This data revealed that K. marxianus KY3 has the potential to produce N-acetyltyramine as the pathogenic-fungal antibiotics and this is the first time that N-acetyltyramine was found to be produced by a yeast strain. This anti-fungal ability may help keep the consortium of kefir grains stable with K. marxianus KY3 as the dominant species. Also, it might confer K. marxianus KY3 the ability to avoid contamination in an industrial fermentation process.

3.2 Adaptation to various conditions

K. marxianus KY3 was tested for its adaptation to various conditions, including utilization of different substrates, pH variation and toxin- and thermo-tolerance. To test 226

227 228 229 230 231 232 233 234

235 236 237 238 239 240 241 242 243 244 245 246

247

248

249

the thermo-tolerance of K. marxianus KY3, which was originally isolated from the kefir grains at 18°C, it was serially diluted and cultured at 30°C, 37°C, 40°C, 42°C and 46°C in YPAD medium supplemented with 2% glucose. Four familiar yeast lab-strains, K.

lactis GG799, S. cerevisiae BY 4741, S. cerevisiae YJM145, and S. cerevisiae YJM789, the latter two of which are thermo-tolerant strains, were cultured as the reference strains [39, 40]. After culturing for 48 hours, all four yeast strains were shown to grow at temperatures lower than 37°C, but K. marxianus KY3 and the two thermo-tolerant S.

cerevisiae strains could also grow at 40°C. However, only K. marxianus KY3 grew well at 42°C and substantially well at 46°C (Fig.2 A). The data indicated that K. marxianus KY3 is more thermo-tolerant than the reference strains and the optimal growth temperature is around 40°C-42°C. The thermo-tolerance phenomenon was also been reported in related strains, such as K. marxianus DMKU3-1042, K. marxianus IMB strains, K. marxianus var. lactis T1 and K. marxianus var. bulgaricus T3 [41-43]. The thermo-tolerance of these strains might be due to a highly efficient superoxide dismutase (SOD) expression at high temperature [42]. A thermo-tolerant host has advantages for developing bioprocesses, including decreasing cooling energy, higher saccharification and fermentation efficiencies [44, 45], as well as convenience for continuous ethanol removal and a lower risk of contamination.

Different carbon sources were employed to test the substrate utilization ability by S. cerevisiae 4741, K. lactis GG799, and K. marxianus KY3. All three yeast strains were able to utilize glucose, galactose, fructose, mannose, raffinose, and sucrose for growth. K. marxianus KY3 and K. lactis GG799 could also utilize glycerol and lactose well, but only K. marxianus KY3 was able to utilize xylose, cellobiose and arabinose (Fig 2B). Previous studies also revealed that K. marxianus DMKU3-1042 has a broad range of carbon source utilization [43]. The wide substrate utilization may enable K.

250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267

268

269

270

271

272

273

274

marxianus to utilize different biomass sources, such as whey waste, corn silage, brewery waste, condensed molasses fermentation solubles (CMS) and rice straw [16, 33, 46-48].

Since the challenges of acidic and toxic conditions are general problems in a lignocellulosic bioprocess [5, 6], these yeast strains were tested for their acid- and toxin- tolerance. The low pH of hydrolysates can affect the efficiency of ethanol production.

Therefore, the growth profiles of K. marxianus KY3 were measured at various pH. P.

stipites CBS 6054, which can assimilate pentose, and S. cerevisiae BY4741, which can assimilate hexose, were used as the reference strains (Fig 2C). All these yeast strains grew between pH 3 and pH 8, but showed very little growth at initial pH 2 and pH 10 conditions. At a low initial pH, P. stipitis CBS 6054 was more strongly inhibited than K. marxianus KY3 and S. cerevisiae BY4741. At initial pH 9, K. marxianus KY3 and P. stipitis CBS 6054 grew well, while the growth of S. cerevisiae BY4741 was significantly inhibited.

A kefir consortium with the co-existence of prokaryotic and eukaryotic microbes represents an attractive model for biorefinery, and it might also represent a co-evolution phenomenon between microbial cohabitants [49]. In this study, the microbial community analysis showed that the dominant yeast, Kluyveromyces marxianus, co- exists with the well known lactic acid producing strain, Lactobacillus kefiranofaciens.

The adaptive evolution due to its long-term existence in acidic environment might be the reason for its acidic tolerance. Three possible mechanisms were considered: First, acidic compounds, such as lactate and acetate, could be assimilated by K. marxianus. In a previous study, a pH control strategy, without addition of alkali, was demonstrated for nisin-producing system by L. lactis subsp. lactis (ATCC 11454), and the kefir grain isolated K. marxianus was employed for lactate assimilation [50]. Second, the firm and 275

276 277

278 279 280 281 282 283 284 285 286 287 288

289

290

291

292

293

294

295

296

297

298

299

solid structure of cell wall could reduce the environmental stress. The cell wall of Kluyveromyces is composed of a high content of β-1,6- and β-1,3-linked glucan chains of polysaccharides, so it is an immunomodulator. The cell wall thickness increased when growing on stress condition, such as 3 % ethanol, so it might also help defend other stresses [51]. Third, plasma membrane H(+)-ATPase could help the pH balance.

When Crabtree-negative yeasts, such as Kluyveromyces, are grown under carbon limitation, sugars mainly appear to be transported by proton-sugar co-transport. A proton-motive force is generated by the transfer of protons or electrons across an energy-transducing membrane and it can be generated by a variety of phenomena including the operation of electron transport chain, illumination of purple membrane, and the hydrolysis of ATP by proton ATPase. The cytochrome c oxidase, a proton- motive force-generating system, of Kluyveromyces can uphill transport under the condition where either the pH gradient or the electrical gradient was present [52]. It has been shown that the addition of ions, such as potassium, to a yeast cell suspension produces an immediate decrease in the level of ATP and an increase in ADP and Pi. An increased activity of a membrane H(+)-ATPase could pump out protons and cause an increase of the medium pH, when the medium pH was low [53].

Toxins produced during acid and steam pretreatment of lignocellulose cover a large range of substances, such as furfural and hydroxymethylfurfural from hemicellulose and cellulose, alcohols and aldehydes from lignin, and heavy metals from bioreactor [5, 6]. In this study, we just showed its tolerance of 2-furaldehyde as an example, but a large scale investigation has been in progress. The data showed that K.

marxianus KY3 could tolerate a concentration of 36 mM 2-furaldehyde, whereas S.

cerevisiae BY4741 could only tolerate a concentration of 24 mM (Fig 2D). The copper toxicity assay showed that KY3 has better tolerance (10 mM CuSO

) than S. cerevisiae 300

301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316

317

318

319

320

321

322

323

324

(8 mM CuSO

) (data not shown). These data indicate that K. marxianus KY3 has a broader pH range and a higher tolerance for toxins than S. cerevisiae BY4741. This is the first report that K. marxianus has the potential to grow under the adverse conditions produced by the cellulosic pretreatment process. Furthermore, n-butanol has been regarded as a potential bio-fuel, but it is known to be highly toxic to its producing hosts.

To study the potential of KY3 for large titers of n-butanol production, the tolerance of KY3 for 1-butanol and iso-butanol was examined. Under our culture condition, KY3 could tolerate a concentration of 1 % 1-butanol and 1.5 % iso-butanol. Since K.

marxianus KY3 can utilize a broad range of carbon sources, including both the hexose and pentose sugars that are the major contents in plant lignocellulose, it has a potential for biofuel industrial applications.

3.3 Biofuel production and biorefinery applications

A series of ethanol production experiments were tested for K. marxianus KY3 at 30°C, 37°C and 40°C in aerobic conditions with either glucose or xylose as the sole carbon substrate. After 12 hours aerobic culturing with 5% glucose, K. marxianus KY3 grew faster and produced more ethanol (1.5 g/L) at 37°C and 40°C than at 30°C (Fig 3A). However, it could produce more cell mass at 30°C than 37°C and 40°C, and this was the case when growing K. marxianus KY3 with 5% xylose for 108 hours of aerobic culturing. K. marxianus KY3 grew faster and produced more ethanol (0.18 g/L) at 37°C and 40°C than at 30°C (0.09 g/L) (Fig 3B). The data indicated that the optimal ethanol production condition of K. marxianus KY3 is around 37°C-40°C and it consumes glucose faster than xylose.

To test the potential for ethanol productivity, a high density cell inoculum approach with a condensed yeast culture (OD

= 11) was used for fermentation in YPAD medium containing 20% glucose at 37°C. The cultures were grown in an airtight 325

326 327 328 329 330 331 332 333 334 335

336 337 338 339 340 341 342 343 344 345 346

347

348

349

container under a semi-anaerobic condition, without aeration or gassing with oxygen- free nitrogen, and the remaining dissolved oxygen was monitored using an anaerobic detecting reagent. At 37°C the rate of ethanol conversion of K. marxianus KY3 was 5 % during the first 24 hours, which was higher than that of S. cerevisiae BY4741 (3 %), and was 3% during the first 12 hours, which was two-fold higher than the ethanol conversion rate of S. cerevisiae BY4741 (Fig 3C). In previous studies, the K. marxianus strains also have been reported to produce ethanol at temperatures above 40°C [41, 44].

On the other hand, an elevated working temperature used for bioprocessing by S.

cerevisiae may induce a negative effect on ethanol production and reduce cell viability.

Although the ultimate ethanol yield of S. cerevisiae BY4741 (6.8%) was higher than that of K. marxianus KY3 (6%) under our culturing condition, the high ethanol conversion rate of K. marxianus KY3 at higher temperature can be an advantage in bio- fuel industry.

In addition, this study also showed an example of the production of higher-alcohol and aromatic chemical byproducts by K. marxianus KY3. The kefir was found to be able to ferment raw milk and produce some “pleasant aroma” by the dominant yeasts.

Our mass spectrometry assay data revealed that K. marxianus KY3 could produce flavor compounds, 2-phenylethyl alcohol (peak 1) and 2-phenylethyl acetate (peak 2), during the fermentation (Fig 1C). To quantify the productivity of flavor compounds, K.

marxianus KY3 and the other two flavor producing strains, K. marxianus CBS600 [54], and K. marxianus CBS6432 [55], were used for phenylalanine conversion testing using a HPLC assay. All three K. marxianus strains were able to produce 2-phenylethyl alcohol or 2-phenylethyl acetate from phenylalanine (2.4 g/l) containing medium after 192 h incubation at 30°C under aerobic or slightly anaerobic conditions. In aerobic culturing, a slightly higher level of 2-phenylethyl alcohol production (0.4686 g /ml) was 350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

found for KY3 than for the other two strains (Table 1). In slightly anaerobic culturing, 0.964 g 2-phenylethanol/ml and 0.435 g 2-phenylethyl acetate/ml were produced by strain KY3; the latter level was significantly higher than those of the two other K.

marxianus strains (Table 1). In addition, K. marxianus KY3 was tested by culturing with raw milk, which contains lactose, galactose, and phenylalanine, at 25°C, 30°C, 37°C, and 42°C under anaerobic conditions. The data showed that the productivity of 2- phenylethyl acetate from raw milk at 37°C were enhanced by ~2 fold of the productivity at 25 °C (Fig. 3D). Thus, raising the temperature seems to be a good strategy for enhancing natural flavor production by K. marxianus KY3.

Microbial biosynthesis of flavor molecules of high economic value from raw substrates might be an ideal process for the natural flavor compounds industry. Among the major metabolites of yeast fermentation, 2-phenylethyl alcohol (fruitlike flavor) and 2-phenylethyl acetate (roses, honeylike flavor), which are the most commonly used aromatic chemicals in perfumery and cosmetics and mainly produced through chemical synthesis, are known natural flavor molecules of high economic value [56]. Previous studies mentioned that K. marxianus could convert L-phenylalanine to a flavor compound, 2-phenylethyl alcohol, via the Ehrlich pathway [54]. The stronger Ehrlich pathway in K. marxianus KY3 suggests that it also has the potential to produce other higher alcohols, such as 1-propanol, 1-butanol, and isobutanol, for biofuel and biorefinery. Furthermore, compared to other yeast, K. marxianus KY3 showed a higher growth rate and a higher cell mass growth, especially at 30°C. This is a benefit for accumulating more metabolites or other target compounds for biorefinery applications [57].

3.4 Potential of enzyme production and CBP 375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

Because K. marxianus KY3 can utilize both hexose and pentose sugars, it was considered for testing as a CBP host in cellulosic ethanol production. Although K.

marxianus KY3 can use cellobiose for growth (Fig. 2B), its cellobiose digestion is poor.

In our study, we transformed a highly efficient β-glucosidase gene (NpaBGS) from the rumen fungus Neocallimastix sp. W5 into K. marxianus KY3 to hydrolyze cellobiose to glucose [33].

Although prior research has reported successful cloning of genes into K.

marxianus [45, 58], maintaining expression of the episomal vectors by antibiotics or nutrient selection is not a convenient strategy for industrial scale applications. In this study, we applied a method developed for transformation and recombination in K. lactis [31], taking advantage of the complementation of multi-copy integration of the LAC4 promoter domain, which possess a high sequencing conservation between K. lactis and KY3 (data not shown). The plasmid pKLAC2-NpaBGS was constructed to carry the NpaBGS gene and then integrated into the K. marxianus KY3 genome using a unique antibiotics-free method in which the acetamidase expressed from pKLAC2 permits the transformed cells to utilize acetamide as a sole nitrogen source [31]. As the transformant, KY3-NpaBGS, appeared to be stable even after growth in YPAD medium for more than 10 generations of culturing without selection pressure, we assumed that a stable integration of NpaBGS gene was successfully established in K. marxianus KY3.

Quantitative β-glucosidase activity assays (using pNPG as the substrate) were conducted on the supernatants of S. cerevisiae transformant (SC-NpaBGS), KY3- NpaBGS transformant and P. pastoris transformant (Pichia-NpaBGS) [59]. We found that the NpaBGS enzyme could not be secreted by SC-NpaBGS (data not shown). In addition, culturing Pichia-NpaBGS for one, three, and five days showed an activity lower than culturing KY3-NpaBGS for one day (Fig. 4A). Our data showed that the 399

400 401 402 403 404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

NpaBGS enzyme gene could be successfully transformed and expressed in K.

marxianus KY3. In a previous study, K. marxianus has been found to efficiently secrete high molecular weight proteins via proper post-translational modifications [60]. It has thus been considered as a promising flexible cell factory in production of recombinant proteins [59, 60].

To demonstrate the improvement of cellobiose digestion for ethanol production, KY3, KY3–NpaBGS and K. marxianus SSSJ-0, which is a native β-glucosidase possessing strain isolated from paper mill waste water [61], were employed for the ethanol conversion assay using cellobiose as the carbon source. The data showed that only KY3–NpaBGS was able to digest cellobiose in YP medium, producing ~ 1.0 g/L ethanol at 37C (Fig. 4B). In comparison, the reference strain, K. marxianus SSSJ-0, did not show any detectable ethanol conversion from 20 g/L cellobiose in the medium, although it could use cellobiose for cell growth.

To consider a suitable yeast candidate for CBP, both the ethanol productivity and the cellulolytic enzyme secretory productivity would be the key parameters. We now provided a summary discussion. Currently, defined single-cell specific secretory systems include methylotrophic species Candida boidinii, Pichia pastoris, Hansenula polymorpha, and Kluyveromyces lactis, and the dimorphic species Arxula adeninivorans and Yarrowia lipolytica. Although some of them can grow at a higher temperature or use xylose as carbon source, most of them can only produce low quantities of ethanol (<2%) compared with the fission yeast Schizosaccharomyces pombe and budding yeast S. cerevisiae ( ～ 10 ％ ). Although some thermotolerant S.

cerevisiae strains have been screened and represent potentials for bioethanol production, thermotolerant K. marxianus, which also possess high ethanol productivity and the cellulolytic enzyme secretory productivity, has been considered an alternative host for 424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

the CBP development [62]. Furthermore, our host KY3 may provide more benefits for biorefinery, such as a high grow rate, a broad spectrum of substrates, and production of value-added aromatic chemicals [59, 63].

3.5 Potential of a dual-microbe co-culturing strategy

A CBP process for SGB production needs to digest feedstock more complex than cellobiose, such as polysaccharides or β-glucans, for direct ethanol fermentation.

Because K. marxianus KY3 was isolated from a dairy kefir grain consortium of prokaryotic and eukaryotic microbes, we considered a co-culturing strategy to improve the efficiency of simultaneous saccharification and fermentation. Lactobacillus kefiranofaciens, Bacillus subtilis, and Escherichia coli were separately employed as a partner bacterial strain with KY3-NpaBGS in the co-culture test. Direct fermentation was conducted at 37°C in YP medium containing 20 g/L β-glucan as the sole carbon source. The data indicate that these systems were able to convert the β-glucan to ethanol when co-cultured with B. subtilis or L. kefiranofaciens, but not with E. coli. The ethanol productivity of B. subtilis-KY3 combination (7.3 g/L) was higher than the Lactobacillus-KY3 combination (2.0 g/L) (Fig. 4C). In our study, the partner B. subtilis might assist K. marxianus KY3 in cellulolytic hydrolysis with a better β-glucosidase and endo-glucanase secretion to contribute to the conversion of cellobiose and β-glucan [64].

In previous studies, K. marxianus had been reported as a complementary partner for bioprocessing [50, 65, 66]. The facultative anaerobic Lactobacillus species, which coexist in kefir, might prefer lactose in a special medium, such as the MRS broth; this preference limits its application [23]. On the other hand, Bacillus species show a high similarity with Lactobacillus species because they are members of the Bacilli genus and 449

450 451 452

453 454 455 456 457 458 459 460 461 462 463 464 465 466 467

468

469

470

471

472

both are gram-positive, rod-shaped and probiotic bacteria [67]. B. subtilis has also been shown to be advantageous as the host for co-cultural CBP, with advantages such as easy culturing and a high secretion ability. It has been reported for co-culturing with several organisms, including bacteria, fungus, and mammalian cells [16, 66, 68, 69]. Since the growth rates of bacterial strains are commonly faster than fungi in the natural environment, a modernized bacterium-yeast co-culture system might be a feasible scheme for effective SGB production.

Both K. marxianus and B. subtilis are generally recognized as safe (GRAS) organisms by the Food and Drug Administration (FDA), and have been reported for a high biotechnological application potential [70, 71]. Both K. marxianus and B. subtilis are known for secreting extracellular enzymes and successful genetic engineering [16, 59, 72]. Furthermore, the engineered KY3-engineered Bacillus co-culturing strategy has been applied to improve a cellulosic ethanol production system in our previous study [73]. The K. marxianus-B. subtilis co-culturing system may also be considered for a

"cell factory" for production of biofuel, recombinant proteins, amino acids, and valuable metabolites and other chemicals via a biorefinery concept.

Conclusion

In this study, the kefir yeast K. marxianus KY3 was isolated and shown to have a broad range of fermentation temperatures, broad carbon source utilization, wide pH adaption and toxin tolerance, suggesting that it has a high potential as a host for bio- refinery. To apply it in cellulosic ethanol industry, KY3 can tolerate low pH, furfural toxin, alcohols inhibition, and heavy metals from bioreactor. Since K. marxianus KY3 can utilize both hexose and pentose and can be genetically engineered to secret a fungal 473

474 475 476 477 478 479

480 481 482 483 484 485 486 487 488

489

490

491

492

493

494

495

496

β-glucosidase, it can be employed for cellulase production and cellulosic ethanol conversion. Furthermore, a B. subtilis-KY3 co-culture strategy was developed to explore cellulosic ethanol production. Moreover, K. marxianus KY3 can produce high- value byproducts, which may reduce the cost for SGB production.

Acknowledgements

This work was supported by grants (NSC 96-3114-P-001-004-Y, NSC 97-3114-P-001- 001, NSC 99-3113-B-001-001, NSC 100-3111-Y001-006, and NSC 99-2321-B-001- 041-MY2) from the National Science Council, Taiwan. We appreciate the valuable metabolite determination technical advice of Prof. Kow-Jen Duan, Tatung University, Taiwan, and the Metabolomics Core Laboratory, Agricultural Biotechnology Research Center, Academia Sinica, Taiwan.

References

[1] Lynd LR, van Zyl WH, McBride JE, Laser M. Consolidated bioprocessing of cellulosic biomass: an update. Curr Opin Biotechnol. 2005;16:577-83.

[2] Nielsen J, Larsson C, van Maris A, Pronk J. Metabolic engineering of yeast for production of fuels and chemicals. Curr Opin Biotechnol. 2013.

[3] Fatih Demirbas M. Biorefineries for biofuel upgrading: A critical review. Applied Energy. 2009;86, Supplement 1:S151-S61.

[4] Martinez-Hernandez E, Sadhukhan J, Campbell GM. Integration of bioethanol as an in-process material in biorefineries using mass pinch analysis. Applied Energy.

2013;104:517-26.

[5] Ingram LO, Aldrich HC, Borges AC, Causey TB, Martinez A, Morales F, et al.

Enteric bacterial catalysts for fuel ethanol production. Biotechnol Prog.

1999;15:855-66.

[6] Klinke HB, Ahring BK, Schmidt AS, Thomsen AB. Characterization of degradation products from alkaline wet oxidation of wheat straw. Bioresource technology. 2002;82:15-26.

[7] Botstein D, Chervitz SA, Cherry JM. Yeast as a model organism. Science.

1997;277:1259-60.

[8] Romanos MA, Scorer CA, Clare JJ. Foreign gene expression in yeast: a review.

Yeast. 1992;8:423-88.

497 498 499 500 501 502

503 504 505 506 507 508

509 510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

[9] Hitzeman RA, Hagie FE, Levine HL, Goeddel DV, Ammerer G, Hall BD. Expression of a human gene for interferon in yeast. Nature. 1981;293:717-22.

[10] Ilmen M, den Haan R, Brevnova E, McBride J, Wiswall E, Froehlich A, et al. High level secretion of cellobiohydrolases by Saccharomyces cerevisiae. Biotechnology for biofuels. 2011;4:30.

[11] Yamada R, Taniguchi N, Tanaka T, Ogino C, Fukuda H, Kondo A. Direct ethanol production from cellulosic materials using a diploid strain of Saccharomyces cerevisiae with optimized cellulase expression. Biotechnology for biofuels.

2011;4:8.

[12] Chapelle FH, O'Neill K, Bradley PM, Methe BA, Ciufo SA, Knobel LL, et al. A hydrogen-based subsurface microbial community dominated by methanogens.

Nature. 2002;415:312-5.

[13] Chang J-J, Lin J-J, Ho C-Y, Chin W-C, Huang C-C. Establishment of rumen-mimic bacterial consortia: A functional union for bio-hydrogen production from cellulosic bioresource. International Journal of Hydrogen Energy. 2010;35:13399-406.

[14] Ho C-Y, Chang J-J, Lin J-J, Chin T-Y, Mathew GM, Huang C-C. Establishment of functional rumen bacterial consortia (FRBC) for simultaneous biohydrogen and bioethanol production from lignocellulose. International Journal of Hydrogen Energy. 2011;36:12168-76.

[15] Mathew GM, Mathew DC, Lo SC, Alexios GM, Yang JC, Sashikumar JM, et al.

Synergistic collaboration of gut symbionts in Odontotermes formosanus for lignocellulosic degradation and bio-hydrogen production. Bioresource technology.

2012.

[16] Chang J-J, Chou C-H, Ho C-Y, Chen W-E, Lay J-J, Huang C-C. Syntrophic co- culture of aerobic Bacillus and anaerobic Clostridium for bio-fuels and bio- hydrogen production. International Journal of Hydrogen Energy. 2008;33:5137-46.

[17] Chou C-H, Han C-L, Chang J-J, Lay J-J. Co-culture of Clostridium beijerinckii L9, Clostridium butyricum M1 and Bacillus thermoamylovorans B5 for converting yeast waste into hydrogen. International Journal of Hydrogen Energy.

2011;36:13972-83.

[18] Abouzied MM, Reddy CA. Direct fermentation of potato starch to ethanol by cocultures of Aspergillus niger and Saccharomyces cerevisiae. Applied and environmental microbiology. 1986;52:1055-9.

[19] Dimitrellou D, Tsaousi K, Kourkoutas Y, Panas P, Kanellaki M, Koutinas AA.

Fermentation efficiency of thermally dried immobilized kefir on casein as starter culture. Process Biochemistry. 2008;43:1323-9.

[20] Pintado ME, Da Silva JAL, Fernandes PB, Malcata FX, Hogg TA. Microbiological and rheological studies on Portuguese kefir grains. International Journal of Food Science & Technology. 1996;31:15-26.

[21] Zhou J, Liu X, Jiang H, Dong M. Analysis of the microflora in Tibetan kefir grains using denaturing gradient gel electrophoresis. Food Microbiol. 2009;26:770- 5.

[22] Zajšek K, Goršek A. Mathematical modelling of ethanol production by mixed kefir grains yeast population as a function of temperature variations. Biochemical Engineering Journal. 2010;49:7-12.

[23] Kullen MJ, Klaenhammer TR. Genetic modification of intestinal lactobacilli and bifidobacteria. Curr Issues Mol Biol. 2000;2:41-50.

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

[24] Nielsen AT, Liu WT, Filipe C, Grady L, Jr., Molin S, Stahl DA. Identification of a novel group of bacteria in sludge from a deteriorated biological phosphorus removal reactor. Applied and environmental microbiology. 1999;65:1251-8.

[25] Cocolin L, Bisson LF, Mills DA. Direct profiling of the yeast dynamics in wine fermentations. FEMS Microbiol Lett. 2000;189:81-7.

[26] Chang JJ, Chen WE, Shih SY, Yu SJ, Lay JJ, Wen FS, et al. Molecular detection of the clostridia in an anaerobic biohydrogen fermentation system by hydrogenase mRNA-targeted reverse transcription-PCR. Applied microbiology and biotechnology. 2006;70:598-604.

[27] Suzuki MT, Giovannoni SJ. Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. Applied and environmental microbiology. 1996;62:625-30.

[28] Kurtzman CP, Robnett CJ. Phylogenetic relationships among yeasts of the 'Saccharomyces complex' determined from multigene sequence analyses. FEMS Yeast Res. 2003;3:417-32.

[29] Saitou N, Nei M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 1987;4:406-25.

[30] Miller GL. Use of Dinitrosalicylic Acid Reagent for Determination of Reducing Sugar. Analytical Chemistry. 1959;31:426-8.

[31] Colussi PA, Taron CH. Kluyveromyces lactis LAC4 promoter variants that lack function in bacteria but retain full function in K. lactis. Applied and environmental microbiology. 2005;71:7092-8.

[32] Swinkels BW, van Ooyen AJ, Bonekamp FJ. The yeast Kluyveromyces lactis as an efficient host for heterologous gene expression. Antonie Van Leeuwenhoek.

1993;64:187-201.

[33] Chen HL, Chen YC, Lu MY, Chang JJ, Wang HT, Ke HM, et al. A highly efficient beta-glucosidase from the buffalo rumen fungus Neocallimastix patriciarum W5.

Biotechnology for biofuels. 2012;5:24.

[34] Kobayashi K, Kuwae S, Ohya T, Ohda T, Ohyama M, Ohi H, et al. High-level expression of recombinant human serum albumin from the methylotrophic yeast Pichia pastoris with minimal protease production and activation. J Biosci Bioeng.

2000;89:55-61.

[35] Dimitrellou D, Kourkoutas Y, Koutinas AA, Kanellaki M. Thermally-dried immobilized kefir on casein as starter culture in dried whey cheese production.

Food Microbiol. 2009;26:809-20.

[36] Ivanova V, Graefe U, Schlegel R, Schlegel R, Gusterova A, Kolarova M, et al.

Isolation and structure elucidation of tyramine and indole alkaloids from antarctic strain microbispora aerata imbas-11A. Biotechnology & Biotechnological Equipment. 2003;17:128-33.

[37] Sakurai H. Studies of metabolism in tuberculous lesions. II. Further studies on the N-acetyltyramine formation by Mycobacterium tuberculosis. Acta Tuberc Jpn.

1956;6:63-8.

[38] Garcez WS, Martins D, Garcez FR, Marques MR, Pereira AA, Oliveira LA, et al.

Effect of spores of saprophytic fungi on phytoalexin accumulation in seeds of frog- eye leaf spot and stem canker-resistant and -susceptible soybean (Glycine max L.) cultivars. J Agric Food Chem. 2000;48:3662-5.

[39] Stumpferl SW, Brand SE, Jiang JC, Korona B, Tiwari A, Dai J, et al. Natural genetic variation in yeast longevity. Genome Res. 2012;22:1963-73.

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

[40] Wei W, McCusker JH, Hyman RW, Jones T, Ning Y, Cao Z, et al. Genome sequencing and comparative analysis of Saccharomyces cerevisiae strain YJM789.

Proceedings of the National Academy of Sciences of the United States of America.

2007;104:12825-30.

[41] Banat IM, Nigam P, Marchant R. Isolation of thermotolerant, fermentative yeasts growing at 52°C and producing ethanol at 45°C and 50°C. World Journal of Microbiology and Biotechnology. 1992;8:259-63.

[42] Kujumdzieva A, Nedeva T, Morfova M, Savov V. ISOLATION AND CHARACTERISATION OF TWO THERMOTOLERANT KLUYVEROMYCES YEAST STRAINS. Journal of Culture Collections. 1998;2:44-50.

[43] Nonklang S, Abdel-Banat BM, Cha-aim K, Moonjai N, Hoshida H, Limtong S, et al. High-temperature ethanol fermentation and transformation with linear DNA in the thermotolerant yeast Kluyveromyces marxianus DMKU3-1042. Applied and environmental microbiology. 2008;74:7514-21.

[44] Brulc JM, Antonopoulos DA, Miller ME, Wilson MK, Yannarell AC, Dinsdale EA, et al. Gene-centric metagenomics of the fiber-adherent bovine rumen microbiome reveals forage specific glycoside hydrolases. Proceedings of the National Academy of Sciences of the United States of America. 2009;106:1948-53.

[45] Hong J, Wang Y, Kumagai H, Tamaki H. Construction of thermotolerant yeast expressing thermostable cellulase genes. J Biotechnol. 2007;130:114-23.

[46] Belem MA, Lee BH. Production of bioingredients from Kluyveromyces marxianus grown on whey: an alternative. Crit Rev Food Sci Nutr. 1998;38:565-98.

[47] Chang JJ, Chou CH, Hsu PC, Yu SJ, Chen WE, Lay JJ, et al. Flow-FISH analysis and isolation of clostridial strains in an anaerobic semi-solid bio-hydrogen producing system by hydrogenase gene target. Applied microbiology and biotechnology.

2007;74:1126-34.

[48] Hang YD, Woodams EE, Hang LE. Utilization of corn silage juice by Klyuveromyces marxianus. Bioresource technology. 2003;86:305-7.

[49] van de Guchte M, Penaud S, Grimaldi C, Barbe V, Bryson K, Nicolas P, et al. The complete genome sequence of Lactobacillus bulgaricus reveals extensive and ongoing reductive evolution. Proceedings of the National Academy of Sciences of the United States of America. 2006;103:9274-9.

[50] Shimizu H, Mizuguchi T, Tanaka E, Shioya S. Nisin production by a mixed- culture system consisting of Lactococcus lactis and Kluyveromyces marxianus.

Applied and environmental microbiology. 1999;65:3134-41.

[51] Backhaus K, Heilmann CJ, Sorgo AG, Purschke G, de Koster CG, Klis FM, et al. A systematic study of the cell wall composition of Kluyveromyces lactis. Yeast.

2010;27:647-60.

[52] Van Leeuwen CC, Postma E, Van den Broek PJ, Van Steveninck J. Proton-motive force-driven D-galactose transport in plasma membrane vesicles from the yeast Kluyveromyces marxianus. The Journal of biological chemistry. 1991;266:12146- 51.

[53] Miranda M, Ramirez J, Pena A, Coria R. Molecular cloning of the plasma membrane H(+)-ATPase from Kluyveromyces lactis: a single nucleotide substitution in the gene confers ethidium bromide resistance and deficiency in K+

uptake. Journal of bacteriology. 1995;177:2360-7.

[54] Etschmann MM, Schrader J. An aqueous-organic two-phase bioprocess for efficient production of the natural aroma chemicals 2-phenylethanol and 2- 625

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

654

655

656

657

658

659

660

661

662

663

664

665

666

667

668

669

670

671

672

phenylethylacetate with yeast. Applied microbiology and biotechnology.

2006;71:440-3.

[55] CE F, PJ B, G G. Screening of yeasts producing 2-phenylethylalcohol.

Biotechnology Techniques. 1997;11:523-5.

[56] Longo MA, Sanromán MA. Production of food aroma compounds: microbial and enzymatic methodologies. Food Technol Biotechnol. 2006;44:335-53.

[57] Hong M-E, Choi SP, Park Y-I, Kim Y-K, Chang WS, Kim BW, et al. Astaxanthin production by a highly photosensitive Haematococcus mutant. Process Biochemistry. 2012;47:1972-9.

[58] Siekstele R, Bartkeviciute D, Sasnauskas K. Cloning, targeted disruption and heterologous expression of the Kluyveromyces marxianus endopolygalacturonase gene (EPG1). Yeast. 1999;15:311-22.

[59] Chang JJ, Ho CY, Ho FJ, Tsai TY, Ke HM, Wang CH, et al. PGASO: A synthetic biology tool for engineering a cellulolytic yeast. Biotechnology for biofuels.

2012;5:53.

[60] Raimondi S, Zanni E, Amaretti A, Palleschi C, Uccelletti D, Rossi M. Thermal adaptability of Kluyveromyces marxianus in recombinant protein production.

Microbial cell factories. 2013;12:34.

[61] Levine RP. Kluyveromyces strains metabolizing cellulosic and hemicellulosic materials. United States Patent. 2008;US8304219.

[62] Hu N, Yuan B, Sun J, Wang SA, Li FL. Thermotolerant Kluyveromyces marxianus and Saccharomyces cerevisiae strains representing potentials for bioethanol production from Jerusalem artichoke by consolidated bioprocessing.

Applied microbiology and biotechnology. 2012;95:1359-68.

[63] Chang JJ, Ho FJ, Ho CY, Wu YC, Hou YH, Huang CC, et al. Assembling a cellulase cocktail and a cellodextrin transporter into a yeast host for CBP ethanol production. Biotechnology for biofuels. 2013;6:19.

[64] Beukes N, Pletschke BI. Effect of sulfur-containing compounds on Bacillus cellulosome-associated 'CMCase' and 'Avicelase' activities. FEMS Microbiol Lett.

2006;264:226-31.

[65] Nissen P, Nielsen D, Arneborg N. The relative glucose uptake abilities of non- Saccharomyces yeasts play a role in their coexistence with Saccharomyces cerevisiae in mixed cultures. Applied microbiology and biotechnology.

2004;64:543-50.

[66] Waites MJ, Morgan NL, Rockey JS, Higton G. Industrial Microbiology: An Introduction: Wiley-Blackwell; 2001.

[67] Kivisto AT, Karp MT. Halophilic anaerobic fermentative bacteria. J Biotechnol.

2011;152:114-24.

[68] Bielecki J, Youngman P, Connelly P, Portnoy DA. Bacillus subtilis expressing a haemolysin gene from Listeria monocytogenes can grow in mammalian cells.

Nature. 1990;345:175-6.

[69] Machado A, Anzai M, Fischman O. Bacillus subtilis induces morphological changes in Fonsecaea pedrosoi in vitro resulting in more resistant fungal forms in vivo. Journal of Venomous Animals and Toxins including Tropical Diseases.

2010;16:592-8.

[70] Fonseca GG, Heinzle E, Wittmann C, Gombert AK. The yeast Kluyveromyces marxianus and its biotechnological potential. Applied microbiology and biotechnology. 2008;79:339-54.

673

674

675

676

677

678

679

680

681

682

683

684

685

686

687

688

689

690

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

713

714

715

716

717

718

719

720

[71] Schallmey M, Singh A, Ward OP. Developments in the use of Bacillus species for industrial production. Can J Microbiol. 2004;50:1-17.

[72] Tsuge K, Itaya M. Recombinational transfer of 100-kilobase genomic DNA to plasmid in Bacillus subtilis 168. Journal of bacteriology. 2001;183:5453-8.

[73] Ho CY, Chang J J, Lee SC, Chin TY, Li WH, Huang CC. Development of cellulosic ethanol production process via co-culturing of artificial cellulosomal Bacillus and kefir yeast. Applied Energy. 2012; 100: 27–32.

Tables

Table 1. Yields of 2-phenylethyl alcohol and 2-phenylethyl acetate of yeast strains.

Figure captions

Figure 1. The phylogenetic study and metabolite assay of kefir grains. (A) Phylogeny of the partial 18S rRNA sequences of all available K. marxianus strains constructed using the neighbor-joining method. (B) Mass spectrometric assay of metabolites.

Figure 2. Adaption of of K. marxianus KY3 to various conditions. (A) Different growth temperatures adaption; (B) Different carbon sources utilization; (C) Various pH values;

(D) Toxin tolerance in cultures containing various concentrations of 2-furaldehyde (0 mM – 48 mM) in a medium with 2% glucose (w/v).

Figure 3. Potential of bioethanol production and biorefinery applications. (A) The ethanol production test with 5% glucose at different temperatures. (B) The ethanol production test with 5% xylose at different temperatures. (C) The ethanol productivity assay with 20% glucose by an initial cell density of OD = 11 at 37C. (D) Phenylalanine conversion testing.

721 722 723 724 725 726 727 728

729 730 731 732 733 734

735

736 737 738 739 740

741

742

743

744

745

746

Figure 4. Potential of cellulosic ethanol conversion of engineered K. marxianus KY3.

(A) Quantitative β-glucosidase activity assays of the supernatants of NpaBGS transformants using pNPG as the substrate. (B) Rates of ethanol conversion from cellobiose for different yeast hosts. (C) Ethanol production with beta-glucan as a carbon source by the co-culture strategy.

747

748

749

750

751