Global warming and consequent climate change caused by fossil fuel-based economy call for sustainable bioenergy production in the modern society [1,2]. Biofuel production based on land plants as a biomass source requires several factors such as land and water for cultivation and thus cannot provide a suitable solution [3,4]. In particular, bioconversion from an edible feedstock such as corn (i.e., first-generation biomass) caused controversy because it competes with food production [5,6]. As an alternative, biofuels obtained from materials not used for food or produced from waste such as wood residues and agricultural waste (i.e., second-generation biomass) have been introduced [7-9]. However, such a wood/grass-based biomass is composed of recalcitrant lignocellulose, yielding a low saccharification efficiency. Third-generation biofuels are derived from marine macroalgae (i.e., seaweed), which have several advantages: they have a rapid growth rate and do not require arable land, water supply, and fertilizers for cultivation [10-12]. In addition, since macroalgae do not contain lignin, their hydrolysis process is more efficient than that of wood- or grass-based biomass [13].

Sargassum horneri is a species of brown macroalgae inhabiting coastal waters of East Asian countries [14,15]. It is an excellent source of carbohydrates, which are represented by 40%-60%w/w in plant dry weight, depending on seasonal variation and sea depth [16-18] and they include cellulose, laminarin, mannitol, alginate, and fucoidan. Nevertheless, S. horneri is poorly investigated as a biomass source for biotechnological purposes [16]. Recently, biomass of S. horneri, drifting from China toward Jeju Island of South Korea, was calculated to reach about 3,000 tons in the period from 2015 to 2018 [19,20]. Because it causes a serious damage to vessels and fisheries and being detrimental to local ecology, S. horneri is treated as marine waste [19]. Therefore, it is necessary to develop a technology to dispose of S. horneri as a “waste”.

Mannitol, a six-carbon sugar alcohol widely distributed in nature, is known as one of the main carbohydrate components of brown algae such as S. horneri and could be a useful carbon source for bioethanol production [21,22]. The mannitol metabolic pathway natively exists in S. cerevisiae, but it is inactivated and additional expression is therefore required for its activation [10,23]. To metabolize mannitol, mannitol transporter and mannitol-2-dehydrogenase are needed to be activated in S. cerevisiae (Fig. 1) [24]. Some mannitol transporters such as Hxt13 and Hxt17 uptake mannitol into a cell, while mannitol-2-dehydrogenases such as Man2 and Dsf1 intracellularly oxidize D-mannitol to D-fructose [10,23]. Enquist–Newman et al. [10] found that deregulation of both HXT17 and MAN2 genes showed better mannitol uptake than deregulation of HXT13 and DSF1 genes.

Figure 1. A proposed mannitol catabolic pathway of Saccharomyces cerevisiae. Overexpression of the native HXT17 and MAN2 genes enabled the engineered yeast strain JE 03 to metabolize mannitol as the sole carbon source. G-3-P, glyceraldehyde-3-phosphate.

In this study, ethanol production from S. horneri biomass was investigated under various conditions such as pretreatment temperature, type of hydrolytic enzymes, and biomass solid loading using mannitol-fermentable S. cerevisiae constructed via activation of mannitol catabolism genes, HXT17 and MAN2. Our study suggests an efficient protocol not only to remove S. horneri as a marine waste but also to utilize it as a potential biomass source for biofuel production.

2.1. Engineering of yeast strains

To develop mannitol-fermentable S. cerevisiae strains, yeast’s mannitol catabolism genes such as mannitol transporter and mannitol-2-dehydrogenase need to be deregulated [10]. To activate the mannitol catabolism pathway, we constructed HXT17- and MAN2-activated strains using the S. cerevisiae D452-2 strain(wild-type; WT) (Table 1). Specifically, the original promoters of HXT17 and MAN2 were replaced by the constitutive strong promoters of TDH3 and CCW12, respectively, using the CRISPR-Cas9 technology. For the promoter substitution, the D452-2 strain expressing the pRS41N-Cas9 plasmid was transformed with the respective donor DNA and the guide RNA plasmid listed in Table 1. The detailed protocols for the generation of donor DNA and a guide RNA plasmid were previously described [25].

Table 1 Strains, plasmids, and primers used in the present study

	Characteristics	References
Strains
D452-2	MATα leu2 ura3 his3	[44]
SR8	D452-2 with an optimized xylose-metabolic pathway	[45, 46]
JE 01	D452-2 PTDH3-HXT17	This study
JE 02	SR8 PCCW12-MAN2	This study
JE 03	SR8 PCCW12-MAN2, PTDH3-HXT17	This study
Plasdmis
pRS41N-Cas9	A single-copy plasmid containing Cas9 and natMX marker	[47]
pRS42H-INT#1	A multicopy plasmid containing an hphMX marker and a gRNA targeting an intergenic region (Int#1)	[48]
pRS42K-HXT17.1	A multicopy plasmid containing the kanMX4 marker and a gRNA targeting the upstream of the HXT17 gene (HXT17.1)	This study
Guide RNAs
INT#1	GAAAGTGATCATTAAGAACA	[48]
HXT17.1	TAATGCACATATAGGCACAT	This study
Primers
Kim1089	TGCACATATAGGCACATGTTTTAGAGCTAGAAATAGCAAG	HXT17.1_F
Kim1090	TGTGCCTATATGTGCATTAGATCATTTATCTTTCACTGCG	HXT17.1_R
Kim1091	GGTATGCATTTTTGTGCATAGAGTGGCGGGGTATTAAGAACTATTTTCGAGGACCTTGTC	PTDH3-HXT17_F
Kim1092	TGAATATCTCTATCACTTTCAGTGGATGATTGCATTTTGTTTGTTTATGTGTGTTTATTC	PTDH3-HXT17_R
Kim1201	GCTAAAATGACCGTAGGATG	MAN2_F
Kim1202	TAGTCATTACTATCGAGGGC	MAN2_R
Kim1203	AAATTAATCTTCTGTCATTCGCTTAAACACTATATCAATAAAAAATGACAAAATCAGACGAAAC	PCCW12-MAN2_F
Kim1515	CTGTTGTTTCGTCTGATTTTGTCATTTTTTATTGATATAGTGTTTAAGCGAATGACAG	PCCW12-MAN2_R
Kim1516	CGCTATTAAGGAAATTTTAGACCAAGTGTGATAATCATGTAATTAGTTATGTCACGCTTAC	MAN2-TCYC1_F
Kim1204	GGAGGGCGTGAATGTAAGCGTGACATAACTAATTACATGATTATCACACTTGGTCTAAAATTTC	MAN2-TCYC1_R

2.2. Yeast cultivation

The obtained yeast colonies were precultured in YPD medium (1% yeast extract, 2% peptone, and 2% D-glucose) for 24 h at 30°C and 250 rpm. Cell pellet of the preculture was subsequently collected by centrifugation, washed with sterilized water, and resuspended to the optical density of either 0.1 or 0.5 at 600 nm (OD600), as the measure of cell density. Finally, the cell suspension was inoculated into YPD medium, YPM medium (1% yeast extract, 2% peptone, and 2% D-mannitol) or a pretreated algal biomass solution and further cultivated at 30°C at either 130 or 250 rpm. All experiments were performed in triplicates.

2.3. Pretreatment of S. horneri

S. horneri was collected from the coast of Jeju Island, South Korea. The collected samples were washed with distilled water 4–5 times to remove salt and subsequently dried at 70°C for 24

h. The dried sample was then milled in a grinder with a pore size of 180−245 μm and stored at −80°C until use. The sample powder was soaked in 4% (v/v) H₂SO₄ with different solid loadings (16%, 20%, and 24%, w/v) and pretreated at 90°C, 105°C, and 121°C for 30 min using an autoclave. The pretreated samples were neutralized to pH 5.5−6.0 by adding 1.5 g CaCO₃.

2.4. Simultaneous saccharification and fermentation (SSF) of S. horneri

For the enzymatic hydrolysis of the pretreated sample, Cellic CTec2 (40 filter paper unit/g biomass) and Viscozyme L (100 fungal β-glucanase unit/g pretreated algae) were added (Novozymes, Krogshoejvej, Denmark). SSF was carried out using the S. cerevisiae WT or the engineered strains (JE 01, JE 02 or JE 03) in 100-mL Erlenmeyer flasks at 30°C and 250 rpm.

2.5. Analytical methods

For determination and quantification of glucose, mannitol, and ethanol in both YPM medium and algal hydrolysate, a high-performance liquid chromatography system was employed (Agilent 1260 series; Agilent, Santa Clara, CA, USA) equipped with a Rezex-ROA Organic Acid H⁺ column (8%, 150 mm × 4.6 mm; Phenomenex, Torrance, CA, USA). The column was eluded with 0.005 N H₂SO₄ at 50°C and the flow rate was set to 0.6 mL/min. Statistical analysis of differences among mean values was performed by either Student’s t test using Microsoft Excel or one-way ANOVA test using IBM SPSS (Armonk, New York, USA).

3.1. Activation of mannitol metabolic genes

A previous study reported that native S. cerevisiae cannot metabolize mannitol but the overexpression of each one of mannitol transporters (Hxt13 and Hxt17) and mannitol-2-dehydrogenases (Man2 and Dsf1) enables it to do so (Fig. 1) [26]. In the present study, we developed a strategy to activate the mannitol metabolic genes in the genome instead of the overexpression using multi-copy plasmids. Specifically, the native promoters of the mannitol metabolic genes (HTX17 and MAN2) in the genome were substituted by a constitutive and strong promoter by seamless gene editing using Cas9 (Fig. 2).

Figure 2. Activation of mannitol metabolic genes by Cas9-based promoter substitution. (A) Replacement of the native HXT17 promoter by the strong constitutive TDH3 promoter for the HXT17 activation. (B) An alternative promoter substitution strategy for MAN2, which has a paralog. The MAN2 gene was amplified from the upstream region, and another PCR was performed for the coding region with overhanging sequences for in vivo assembly. Next, the CCW12 promoter, MAN2 gene, and the CYC1 terminator were assembled in vivo and integrated at an intergenic region of the genome (Int#1).

First, the HXT17 promoter substitution was performed to activate the mannitol transporter gene. In JE 01, an HXT17-activated strain, the HXT17 promoter was replaced by the TDH3 promoter. The guide RNA (gRNA, pRS42K-HXT17.1) was designed to target the upstream of the HXT17 gene, where the TDH3 promoter was inserted as a donor DNA, resulting in the JE 01 strain. For the MAN2 gene, to avoid the genomic complexity due to its paralog gene (DSF1), the MAN2 gene was selectively amplified by a secondary PCR, It was assembled with the CCW12 promoter and CYC1 terminator in vivo, and genome was integrated at an intergenic site (Int#1) by Cas9, resulting in the JE 02 strain. Finally, the JE 03 strain was constructed by activating the HXT17 gene, as described above, in the JE 02 strain, resulting in the JE 03 strain.

3.2. Growth of the engineered S. cerevisiae strains on mannitol as the sole carbon source

To confirm the activation of the mannitol metabolic pathway in the engineered S. cerevisiae strains, mannitol consumption and yeast growth were evaluated on YPM medium using the D452-2, JE 01, JE 02, and JE 03 strains (Fig. 3). Only JE 03 showed a significant consumption of mannitol and a positive growth rate in a mannitol-containing medium, indicating that both mannitol transporter and mannitol-2-dehydrogenase were required for yeast to utilize mannitol for its growth. To be specific, mannitol was exhausted in 24 h, achieving the maximum cell density (i.e., OD600) of 9.8, at which 6-7 g/L of ethanol were produced by JE 03, whereas, similarly to WT, neither JE 01 nor JE 02 showed any mannitol consumption and cellular growth. These results verified that S. cerevisiae was successfully engineered to metabolize mannitol as the sole carbon source.

Figure 3. Mannitol supports the growth of the engineered strain (JE03) as the sole carbon source in a complex medium. Mannitol consumption (A) and cell density (B) were monitored for Saccharomyces cerevisiae D452-2 (control) and the engineered strains, JE 01 (activated HXT17), JE 02 (activated MAN2), and JE 03 (activated HXT17 and MAN2). Fermentation was performed at 30°C and 250 rpm with an initial OD₆₀₀ of 0.1. Values represent the means of three independent experiments. Error bars indicate standard deviation.

3.3. Effect of pretreatment temperature on ethanol production

Bioconversion of biomass into ethanol normally requires four steps: pretreatment, saccharification, fermentation, and ethanol extraction (i.e., recovery) [27,28]. Pretreatment helps in releasing and extracting sugars such as mannitol from brown algae, which is closely associated with deconstruction of algal tissues. Thus, implementation of a pretreatment prior to enzymatic hydrolysis and/or fermentation processes enhances both sugar and ethanol yields from algae [27,29]. In particular, chemical pretreatments that include acids such as HCl and H₂SO₄, leading to the hydrolysis of polysaccharides such as cellulose and laminarin, have been widely used in brown algae [29-31]. In this study, algal biomass was hydrothermally pretreated with 4% (v/v) H₂SO₄ at different temperatures. SSF of the JE 03 strain showed the highest ethanol productivity by the 121°C-pretreated algal sample (Fig. 4).

Figure 4. Effect of pretreatment temperatures of algal biomass on ethanol productivity by the JE 03 strain. Acid-pretreatment of Sargassum horneri (16%, w/v) was performed at 90°C, 105°C, and 121°C. Simultaneous saccharification and fermentation of the hydrolysates were performed for 36 h at 30°C and 250 rpm with an initial OD₆₀₀ of 0.5. Mean values with different letters are significantly different (p < 0.05, one-way ANOVA/Tukey test).

Pretreatment is an important step in biomass conversion into a fuel or other chemicals that facilitates both enzymatic saccharification and microbial fermentation. Up to date, a variety of pretreatment methods have been developed for seaweed based on e.g., ultrasound, microwave or ionic liquid solvents [32,33]. In particular, microwave-assisted pretreatment has some advantages, enabling wide application for algal biomass [34]. Microwave-assisted pretreatment (i.e., hydrolysis) of algae, which is characterized by fast heating, is highly efficient in releasing sugars and minimizes the production of inhibitors detrimental to the subsequent fermentation [34,35]. Although the highest ethanol productivity in this study was obtained by the acidic pretreatment of S. horneri at 121°C, further optimization needs to be performed under variable conditions that include employment of other pretreatment techniques, temperatures, and acid concentrations.

3.4. Enzymatic saccharification of pretreated algae

Brown algae contain cellulose and laminarin as the major polysaccharides [36]. Cellulose, a linear glucan with β-1,4 linkages, is the most abundant carbon source on the planet and its enzymatic hydrolysis kinetics is well-known [2,37]. In many cases, cellulase mixtures composed of lytic polysaccharide monooxygenase, cellobiohydrolase (CBH), endoglucanase (EG), and β-glucosidase (BG) (e.g., Cellic CTec2) have been applied. For the hydrolysis of laminarin, a glucan with a β-1,3-linked backbone partially branched with β-1,6 bonds, a different class of enzyme that could cleave β-1,3 and/or β-1,6 linkages is required [38, 39]. In this study, we carried out enzymatic saccharification of pretreated algae using commercial enzyme mixtures, Cellic CTec2 and/or Viscozyme L. Cellic CTec2 is widely used for cellulose hydrolysis, while Viscozyme L is a multi-enzyme blend with a broader spectrum of carbohydrate substrates, including β-glucanase, cellulase, hemicellulase, and pectinase.

The pretreated algal samples were hydrolyzed either with only Cellic CTec2 or a mix of Cellic CTec2 and Viscozyme L (Fig. 5), and the released glucose and mannitol concentrations were monitored. During a 24-h incubation, when Cellic CTec2 only was used, 22.21 g/L glucose was obtained with a marginal amount of mannitol (< 3 g/L). When using the two-enzyme mixture, however, significantly higher concentrations of glucose and mannitol were obtained, 34.29 g/L and 19.56 g/L, respectively. The result suggested that Viscozyme L is able to hydrolyze laminarin of the brown algae, resulting in the release of glucose and mannitol.

Figure 5. Role of glucanase for mannitol release from pretreated algal biomass. Enzyme hydrolysis of pretreated Sargassum horneri (16%, w/v) at 121°C was performed by cellulase only (Cellic CTec2) and a mixture of cellulase and glucanase (Cellic CTec2 and Viscozyme L) at 30°C, and the released glucose (A) and mannitol (B) concentrations were monitored for 24 h.

3.5. Optimization of ethanol production by SSF of pretreated algal biomass

To increase the final ethanol concentration, the solid loading of algal biomass for the pretreatment was increased from 16%, to 20% and 24% (w/v). A mixture of Cellic CTec2 and Viscozyme L was used for saccharification and the JE 03 strain was used for fermentation (Table 2 and Fig. 6). The final ethanol concentrations were 16.31 g/L, 20.46 g/L, and 22.95 g/L at 16%, 20%, and 24% (w/v) solid loadings, respectively. However, ethanol productivity (0.680 g/L/h) was the highest at the lowest solid loading tested (16%). Ethanol yield from 16% biomass was 10.19% (w/w, based on biomass), which was higher than those

Table 2 Effect of biomass solid loading on ethanol production

Solid loading (%, w/v)	Ethanol (g/L)	Ethanol productivity (g/L/h)
16	16.3 ± 3.4	0.680 ± 0.140
20	20.4 ± 1.0	0.568 ± 0.027
24	22.9 ± 2.7	0.479 ± 0.056

Figure 6. Effect of solid loadings of pretreated algal biomass on cellulosic ethanol production. Pretreated Sargassum horneri (16%, 20%, and 24%, w/v) at 121°C were subjected to simultaneous saccharification and fermentation by the JE 03 strain with a mixture of cellulase and glucanase, and the concentration of glucose (A), mannitol (B), and ethanol (C) were monitored. Values represent the means of three independent experiments. Error bars indicate standard deviation.

Table 2. Effect of biomass solid loading on ethanol production obtained from other brown seaweeds, fermented using S. cerevisiae (Table 3) [40,41].

Table 3 Comparison of ethanol yields from various brown algae

Brown algae	Ethanol yield (%, w/w)	Reference
Laminaria japonica	5.90	[40]
Hizikia fusiformis	0.45	[40]
Undaria pinnatifida	4.14	[41]
Sargassum horneri	10.19	This study

Glucose repression is the phenomenon widely observed in microorganisms including S. cerevisiae, in which glucose suppresses the activities of enzymes involved in metabolism of alternate carbon sources. In this study, the presence of glucose in the pretreated biomass hydrolysates did not have a pronouncedly negative effect on mannitol utilization by JE03 strain (Fig. 6). In the presence of glucose, mannitol consumption rate was comparable to that for glucose, achieving the 100% consumption.

Performing SSF with a high solid loading is important to achieve an economically effective ethanol recovery process [42]. Ethanol productivity in our study decreased as the solid loading increased. A SSF process at a high solid loading is known to suffer from low ethanol yield and slow fermentation due to a poor access of microbial cells and hydrolytic enzymes to a substrate (i.e., mass transfer limitation) [43]. To tackle this problem, it would be helpful to utilize enzymes that rapidly hydrolyze biomass, thereby reducing the viscosity of media as well as to develop an efficient mechanical mixing system. Also, performing SSF with a higher initial cell density could be helpful for obtaining more efficient ethanol productivity from high biomass solid loading.

S. horneri, a brown seaweed widely distributed along the coastal regions of Korea, Japan, and China, is being considered as a marine waste. In this study, we have not only shown an efficient way to treat S. horneri but also suggested a method how the “waste” S. horneri could be utilized as biomass for producing bioethanol. Our study shows how polysaccharides from marine algae could be valorized as fuels or useful chemicals, providing an opportunity for the research to be expanded to the exploration of other brown algal species. Future works will be directed toward a metabolic engineering-based development of microbial systems in order to utilize other major carbohydrates present in brown algae such as alginate and fucoidan that were not examined in this study to complete the conversion of carbohydrates and allow for more efficient ethanol production.

This work was supported by Center for Women in Science, Engineering and Technology (WISET) grant funded by the Ministry of Science and ICT (MSIT), Korea under the team research program for female engineering students.