Production of neoagarooligosaccharides by probiotic yeast Saccharomyces cerevisiae var. boulardii engineered as a microbial cell factory

Background Saccharomyces cerevisiae var. boulardii is a representative probiotic yeast that has been widely used in the food and pharmaceutical industries. However, S. boulardii has not been studied as a microbial cell factory for producing useful substances. Agarose, a major component of red macroalgae, can be depolymerized into neoagarooligosaccharides (NAOSs) by an endo-type β-agarase. NAOSs, including neoagarotetraose (NeoDP4), are known to be health-benefiting substances owing to their prebiotic effect. Thus, NAOS production in the gut is required. In this study, the probiotic yeast S. boulardii was engineered to produce NAOSs by expressing an endo-type β-agarase, BpGH16A, derived from a human gut bacterium Bacteroides plebeius. Results In total, four different signal peptides were compared in S. boulardii for protein (BpGH16A) secretion for the first time. The SED1 signal peptide derived from Saccharomyces cerevisiae was selected as optimal for extracellular production of NeoDP4 from agarose. Expression of BpGH16A was performed in two ways using the plasmid vector system and the clustered regularly interspaced short palindromic repeat (CRISPR)-Cas9 system. The production of NeoDP4 by engineered S. boulardii was verified and quantified. NeoDP4 was produced by S. boulardii engineered using the plasmid vector system and CRISPR-Cas9 at 1.86 and 0.80 g/L in a 72-h fermentation, respectively. Conclusions This is the first report on NAOS production using the probiotic yeast S. boulardii. Our results suggest that S. boulardii can be considered a microbial cell factory to produce health-beneficial substances in the human gut. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1186/s12934-021-01644-w.


Background
The gut microbiota and intestinal immunity have been found to have a major impact on human health [1,2]. Microbial cell factories, including Saccharomyces cerevisiae, have been approved for human use and have been developed to produce therapeutic proteins [3,4]. Thus, if a microbial cell factory that can directly produce therapeutic proteins in the intestine is developed, it would be possible to study the effects of intestinal microbes on health more accurately, and the microbial cell factory could be used for the treatment of actual diseases. When designing microbial cell factories, one important consideration is the selection of an appropriate microbial host as an expression system [5], based on its safety and health benefits. Therefore, well-known probiotics, such as lactic acid bacteria and Saccharomyces cerevisiae var. boulardii, could be promising hosts as expression systems.
Lactic acid bacteria or intestinal bacteria can be used as hosts for microbial cell factories, but there may be some constraints. First, from a pharmacokinetic point of view, it may be necessary to control the inhabiting population and residence time in the gut [6]. In the case of S. boulardii, this yeast was washed out in the gut within 3-5 days after discontinuing oral administration [7,8]. However, in the case of intestinal bacteria, it is not possible to control their inhabiting population and residence time in the gut. Second, post-translational modifications including glycosylation and phosphorylation of eukaryotic proteins, are crucial for the expression of their biological activity [9]. Therefore, yeast has been used as an eukaryotic host to produce many therapeutic proteins [9]. Taken together, the eukaryotic probiotic S. boulardii is considered a promising host to use as an intestinal microbial cell factory [10,11].
S. boulardii, originally isolated from lychee and mangosteen, is a generally recognized as safe (GRAS) yeast [12,13]. S. boulardii is known to survive in the human gastrointestinal tract owing to its high tolerance to low pH and high temperatures [13,14]. Additionally, S. boulardii is the only probiotic yeast found to be effective in double-blind studies [14,15]. Previously, the metabolic engineering of S. boulardii and its use as a potential oral vaccine delivery vehicle were studied in mouse models [16,17]. However, to our knowledge, there have been no studies yet on the production of prebiotics which possess a beneficial effect on the health of a host by selective stimulation of the activity or growth of probiotic-like bacteria in the colon using an engineered S. boulardii [18]. In this study, we engineered the probiotic yeast S. boulardii to produce bioactive substances with prebiotic effects with the ultimate goal of developing a synbiotic system for humans.
When non-digestible diets reach the large intestine, they are utilized by the gut microbiota. Subsequently, non-digestible diets change the intestinal microflora and affect the overall health of the host [19,20]. As a non-digestible diet, agarose, a polysaccharide obtained from red macroalgae, is commonly used as dietary fiber by East Asians [21]. Neoagarooligosaccharides (NAOSs) derived from agarose were found to have various physiological and biological activities, including anti-obesity, anti-diabetic, anti-inflammatory, anti-viral, and antitumor activities [22][23][24][25][26][27]. Moreover, in vivo experiments confirmed that NAOSs have a prebiotic effect [28]. In particular, neoagarotetraose (NeoDP4), which contains various bioactive properties, such as anti-inflammatory [29] and anti-oxidative activity [30], has been found to be a potential prebiotic for modulating intestinal microbiota, thereby promoting the health of the host. In addition, Bifidobacterium, which is considered as a beneficial probiotic microorganism having therapeutic benefits and is one of the most commonly used probiotics in humans [31], significantly increased in mice treated with antibiotics supplemented with NeoDP4 [29]. NeoDP4 can be produced by endo-type β-agarase from agarose. Recently, an endo-type β-agarase, BpGH16A originating from human gut bacterium Bacteroides plebeius, has been reported [32,33]. As B. plebeius was isolated from human gut microbes that can be considered relatively safe, BpGH16A was chosen to be expressed in S. boulardii to enzymatically produce NAOSs, primarily NeoDP4, from agarose.
In this study, we introduced and expressed the gene for BpGH16A in the probiotic yeast S. boulardii using CRISPR-Cas9, and the production of prebiotic NeoDP4 from agarose by the engineered S. boulardii was verified and optimized (Fig. 1). To our knowledge, this is the first study to show that the probiotic S. boulardii can be used as a microbial cell factory for producing prebiotics.

Strains and media
Escherichia coli DH5α was used to construct the plasmids. E. coli strains were grown in Luria-Bertani medium (10 g/L tryptone, 5 g/L yeast extract, and 10 g/L sodium chloride) containing 100 µg/mL ampicillin (Sigma-Aldrich, St. Louis, MO, USA) at 37 °C and 200 rpm. S. boulardii ATCC MYA-796 was used as the parental strain for producing NeoDP4. Yeast strains were grown at 37 °C on yeast synthetic complete (YSC) medium, which contained a 6.7 g/L yeast nitrogen base without amino acids (Becton, Dickinson and Company, Franklin Lakes, NJ, USA), 20 g/L glucose, and 1.92 g/L yeast synthetic drop-out medium supplements without uracil (Sigma-Aldrich), or YSC medium, which contained a 6.7 g/L yeast nitrogen base (Becton, Dickinson and Company) and 20 g/L glucose. For CRISPR-Cas9-based genome editing experiments, 100 µg/mL nourseothricin (NAT; Jena Bioscience, Jena, Germany) and 500 µg/ mL geneticin (AG Scientific, San Diego, CA, USA) were added to the medium if required for selection of the transformants.

Plasmid and strain construction
To use the auxotrophic marker for transformant selection, a strain with inactivated HIS3, TRP1, and URA3 was created first. Based on the SB-TU strain [17] in which TRP1 and URA3 were inactivated, HIS3 was further inactivated. The gene fragment of HIS3 [34] was amplified using the primer pair gHIS3_F and gHIS3_R ( Table 1). The resulting polymerase chain reaction (PCR) product was digested with SacI and NotI and ligated to pRS42H, resulting in plasmid p42H_gHIS3, designed for simultaneously inactivating HIS3 (Table 2). Repair DNA for HIS3 disruption was amplified using PCR by using primers dDNA_HIS3_F and dDNA_HIS3_R ( Table 1).
The Cas9-NAT plasmid conferring NAT resistance was used to select the transformants using an antibiotic marker after transformation [34]. Yeast transformation into SB-TU was carried out using the polyethyleneglycol (PEG)-LiAc method, as described previously [35]. Finally, S. boulardii strain SB-HTU, in which HIS3, TRP1, and URA3 were inactivated, was constructed ( Table 3).
Plasmids that were used to screen an optimal signal peptide for secretion of BpGH16A were constructed as follows. The gene BACPLE_01670, encoding BpGH16A, was cloned into the pRS426GPD plasmid. The BpGH16A gene fragment was amplified from B. plebeius DSM 17135 (DSMZ, Braunschweig, Germany) genomic DNA using PCR with different primer pairs depending on the type of signal peptide ( Table 1). The predicted signal peptide sequences at the N-terminus of BpGH16A were removed for signal peptide screening. In total, four different signal peptides, namely chicken lysozyme signal peptide (CL), α-mating factor signal peptide (α-MF) from S. cerevisiae, Sta1 signal peptide (STA1) from Saccharomyces diastaticus, and Sed1 signal peptide (SED1) from S. cerevisiae, were used [17,36,37]. Additionally, for construction of the control strain without any signal peptide, PCR was performed using the primer pairs 16A_W/OSP_F_SpeI and 16A_W/OSP_R_XhoI ( Table 1). The PCR products were double-digested by restriction enzymes determined during primer design and ligated with plasmid pRS426GPD digested using the same restriction enzymes, using T4 DNA ligase (New England Biolabs, Ipswich, MA, USA). The resulting plasmids were designated as p426_Bp_W/OSP, p426_Bp_CL, p426_Bp_αMF, p426_Bp_STA1, and p426_Bp_SED1 ( Table 2). Yeast transformation into SB-HTU was performed using the PEG-LiAc method. Finally, the experimental strains SB-HTU_16A_C, SB-HTU_16A_A, SB-HTU_16A_S, and SB-HTU_16A_D were prepared for signal peptide screening ( Table 3). As control strains, SB-HTU_E harboring neither BpGH16A nor signal peptide, and only pRS426GPD vector, and SB-HTU_W harboring the BpGH16A gene but no signal peptide were prepared.
To integrate the BpGH16A gene into the genome of S. boulardii for the stable expression of BpGH16A, a guide RNA plasmid p42K_CS5 was constructed ( Table 2). The plasmid was generated using reverse PCR of the pRS42K plasmid containing a guide RNA sequence using the primer pairs gRNA_CS5_F and gRNA_CS5_R ( Table 1). The 20-bp targeting sequence of the guide RNA binds to the front of the PAM sequence (NGG) in the empty locus on chromosome XV (CS5). The BpGH16A and SED1 signal peptide were incorporated into this locus via homologous recombination without affecting the function of other genes. For homologous recombination, the plasmid p426_Bp_SED1 was amplified using the primer pairs dDNA-CS5-F and dDNA-CS5-R as donor DNA for genome integration using CRISPR-Cas9 (Table 1). To overcome the inefficiencies associated with genome integration, the PCR product constructed using the primer pairs dDNA-CS5-F and dDNA-CS5-R was amplified using PCR by using primer Fig. 1 Schematic representation of NAOS production by engineered S. boulardii. BpGH16A expression was performed using two different systems. The first was a vector system with an auxotrophic marker, and the second was the clustered regularly interspaced short palindromic repeat (CRISPR)-Cas9 system. NAOSs, neoagarooligosaccharides; BpGH16A, endo-type β-agarase originating from human gut bacterium Bacteroides plebeius   Genomic integration was verified with yeast colony PCR by using the primer pairs Conf-CS5-F and Conf-CS5-R (Table 1).

Fermentation experiments
To produce NAOSs in the engineered S. boulardii, fermentation was performed with 2.5 g/L agarose at 37 °C and 200 rpm in 125-mL flasks for 72 h. During fermentation, agarose, low gelling temperature (Sigma-Aldrich) was used to prevent congealing of agarose. First, strains SB-HTU_16A_C, SB-HTU_16A_A, SB-HTU_16A_S, and SB-HTU_16A_D were grown in YSC medium at 37 °C and 200 rpm. Pre-cultured cells were centrifuged at 10,170 × g for 10 min and washed twice using sterilized distilled water. The harvested cells were inoculated into 20 mL of YSC medium containing 20 g/L glucose and 2.5 g/L agarose in 50 mM potassium hydrogen phthalate (KHP) buffer (pH 5.5). The initial cell density was adjusted to an optical density at 600 nm (OD 600 ) of 1.0. As a control, strains SB-HTU_16A_E and SB-HTU_16A_W were fermented under the same conditions. To examine NAOSs production by S. boulardii SB_16A_D engineered by CRISPR-Cas9, fermentation was performed in YSC medium. Pre-cultured cells in YSC medium were inoculated in 20 mL of YSC medium containing 20 g/L glucose and 2.5 g/L agarose in 50 mM KHP buffer (pH 5.5). The initial cell density was adjusted to OD 600 = 1.0. Wild-type S. boulardii ATCC MYA-796 was used as the control strain. Fermentation experiments were performed in triplicates.

Analyses of cell growth and NAOS production using high-performance liquid chromatography
Cell growth was monitored by measuring OD 600 using a UV-visible spectrophotometer (Bio-Rad, Hercules, CA, USA). To analyze and quantify the reaction products of agarose by the engineered S. boulardii, including NeoDP4, glucose, acetic acid, and ethanol, high-performance liquid chromatography (HPLC) analysis was performed. The HPLC system (Agilent Technologies, Santa Clara, CA, USA) was equipped with a refractive index (RI) detector (Agilent Technologies) using an Aminex HPX-87H column (Bio-Rad). The column and RI detector temperatures were set to 65 and 55 °C, respectively, and the column was eluted with 0.005 M H 2 SO 4 at a flow rate of 0.5 mL/min.

Identification of NAOSs using thin-layer chromatography
To identify the hydrolyzed products of agarose during fermentation, thin-layer chromatography (TLC) analysis was performed. During fermentation, 1 mL of cell culture containing the fermentation products was obtained for each time point (0, 12, 24, 36, 48, and 72 h). For accurate measurements, the obtained cell culture was boiled to terminate the possible enzymatic reaction. After centrifugation at 16,609×g for 15 min at 4 °C, a 1-µL aliquot from each supernatant was loaded onto silica gel 60 plates (Merck, Damstadt, Germany). After drying the TLC plates, they were visualized with 10% (v/v) H 2 SO 4 in ethanol and 0.2% (w/v) naphthoresorcinol in ethanol, as previously described [38].

NeoDP4 production by engineered yeast
To produce NeoDP4 by engineered yeast, β-agarase secretion from yeast was necessary for agarose degradation. Therefore, the expression and secretion of an endotype β-agarase, BpGH16A, by yeast S. boulardii, was tested first. To confirm that BpGH16A is functionally expressed and secreted by S. boulardii, SB-HTU_16A_C containing CL, which has been previously proven to work in S. boulardii, was used [17].
In the HPLC analysis of SB-HTU_16A_C fermentation products, a significant peak at an approximate retention time of 7.6 min, corresponding to NeoDP4, was detected in the sample (Fig. 2). In contrast, no peak was detected in the sample from the control strain, SB-HTU_16A_E, harboring the empty vector. Thus, these results showed that BpGH16A was functionally expressed, secreted from S. boulardii, and produced NeoDP4 by hydrolyzing agarose.

Screening of the optimal signal peptide for NeoDP4 production by S. boulardii
As the enzyme BpGH16A was confirmed to be functionally expressed and secreted by S. boulardii, the next step was to identify the optimal signal peptide to further increase NeoDP4 production. Four different signal peptides, CL, α-MF, STA1, and SED1, were tested. Each signal peptide was individually pre-fixed to the BpGH16A sequences and introduced into SB-HTU to identify the signal peptide that produces more NeoDP4. Production of NeoDP4 by the engineered yeasts was verified using TLC analysis at 72 h (Fig. 3A); NeoDP4 was strongly produced by the strains containing CL and SED1. The NeoDP4 spot was weakly detected by the strain with α-MF. To compare the amount of NeoDP4 produced by each engineered yeast more accurately, HPLC analysis was performed (Fig. 3B). NeoDP4 was found to have gradually increased during the 72-h fermentation by each strain, and the highest amount of NeoDP4 was produced by the strain containing SED1. The amount of NeoDP4 produced after 72 h of fermentation was 1.73, 0.95, 0.99, and 1.86 g/L when signal peptides, namely CL, α-MF, STA1, and SED1, respectively, were used. The production of NeoDP4 by the strain containing SED1 was 1.08, 1.96, and 1.88 times higher than that by the strains containing CL, α-MF, and STA1, respectively. Thus, SED1 from S. cerevisiae, which showed the highest production of NeoDP4, was selected for further NeoDP4 production.
Meanwhile, NeoDP4 was not produced in any control group, whereas it was produced in all groups in which the signal peptide was present (Fig. 3). In particular, SB-HTU_16A_W, which was used as a negative control, was constructed to confirm that the extracellular activity of BpGH16A was not derived from cell lysis but from secretion due to the heterologously expressed signal peptides. As NeoDP4 was not detected in the culture broth of SB-HTU_16A_W, the degradation of agarose into NeoDP4 was confirmed to be caused by the secreted BpGH16A.

Fermentation for the production of NeoDP4 by the engineered yeast
Based on the signal peptide screening results, the strain SB-HTU_16A_D containing the SED1 signal peptide was fermented in YSC medium without uracil containing 2.5 g/L agarose for 72 h. The fermentation products were analyzed using TLC and HPLC. NeoDP4 production was confirmed using TLC analysis (Fig. 4A). The initially added glucose was confirmed to be depleted after 36 h. Based on HPLC analysis of the fermentation products at each time point, 1.86 g/L NeoDP4 was obtained after 72 h of fermentation (Fig. 4B). Cell growth entered the stationary phase from 24 h onwards and reached an OD 600 = 6.7. Both ethanol and acetic acid accumulated up to a concentration of 4.8 g/L. In conclusion, NeoDP4 was produced as the target major product by the engineered yeast SB-HTU_16A_D.

Strain construction using CRISPR-Cas9 and fermentation
For stable expression, BpGH16A and SED1 were introduced into the genome of S. boulardii, in this study. BpGH16A gene knock-in was performed via homologous recombination using CRISPR-Cas9 (Additional file 1: Fig. S1A). After yeast transformation, the gene encoding BpGH16A was confirmed to have entered the genome successfully, using yeast colony PCR. Primer pairs Conf-CS5-F and Conf-CS5-R were designed and used so that the size was 2.4-kb when the gene entered and 0.5-kb when the gene was not entered ( Table 2). The successful integration of BpGH16A and SED1 into S. boulardii was confirmed with yeast colony PCR, based on the formation of a 2.4-kb single band in 3 lanes, namely, 4, 7, and 8 out of total 9 lanes (Additional file 1: Fig. S1B).
Finally, strain SB_16A_D containing BpGH16A and SED1 in the S. boulardii genome was constructed using CRISPR-Cas9, and flask fermentation proceeded in YSC medium containing 2.5 g/L agarose, for 72 h. NeoDP4 production was confirmed using TLC analysis (Fig. 5A), suggesting that BpGH16A was secreted from strain SB_16A_D. For more accurate fermentation products analysis, HPLC analysis and growth measurement were performed at each time point as well (Fig. 5B). Glucose was depleted before 12 h of fermentation had passed, and the strain grew to an OD 600 = 15.51 at 72 h. After fermentation, 0.80 g/L NeoDP4 was produced, as well as 3.03 g/L ethanol and 3.65 g/L acetic acid.

Discussion
S. boulardii has been widely used as a probiotic because it can compete with diarrhea-causing pathogens in the human gut [39]. Additionally, S. boulardii may be used as a host for a microbial cell factory that produces useful proteins in the gut [11]. S. boulardii has been metabolically engineered; however, it has not yet been engineered to produce prebiotics [17]. In this study, we introduced an endo-type β-agarase, BpGH16A, into S. boulardii to produce potential prebiotics, NAOSs from red macroalgal agarose (Fig. 1). The production of NeoDP4, a NAOS, was verified by engineered S. boulardii.
NAOSs, which can be produced by hydrolyzing agarose extracted from red macroalgae by endo-type β-agarase, have been reported to possess various health benefits [22][23][24][25][26][27]. NeoDP4, which is a representative NAOS, has been reported to have various biological properties [40,41]. For example, the anti-fatigue effects of NeoDP4 via short-chain fatty acid production and regulation of the microbial composition have been demonstrated in mice [29]. Moreover, a significant increase in Lactobacillus and Bifidobacterium was observed by supplementing mice with NeoDP4, implying that NeoDP4 is a potential prebiotic [29]. NeoDP4 is also known to alleviate the inflammatory response by inhibiting the MAPK and NF-κB signaling pathways [26]. Therefore, the requirement for production of NAOSs, especially NeoDP4, is increasing rapidly [42].
One of the goals of this study was to effectively secrete BpGH16A to hydrolyze agarose into NeoDP4 by expressing a signal peptide in engineered yeast. Previous studies have used CL or α-MF signal peptides to secrete proteins from S. boulardii [11,16,17]. Based on additional screening using STA1 and SED1 used in other yeast strains [36,37], all of them (CL, α-MF, STA1, and SED1) were confirmed to secrete BpGH16A in S. boulardii (Fig. 3). Through signal peptide screening, SED1 was shown to have the highest efficiency in producing NeoDP4 via secretion of BpGH16A. Because of the lack of other reports on signal peptide screening in S. boulardii so far, this study could contribute to studies on the production and secretion of other proteins by S. boulardii.
Genomic integration can avoid problems that may arise in a complex intestinal environment when using plasmids. These problems include plasmid instability in the absence of selective pressure, potential diffusion to other microorganisms, and an increased metabolic burden associated with the maintenance of multicopy plasmids [43]. Therefore, we attempted to integrate BACPLE_01670 coding for BpGH16A into the genome of S. boulardii using the CRISPR-Cas9 system, which is a sophisticated and advanced genomic engineering tool (Additional file 1: Fig. S1A) [44], and succeeded in constructing an S. boulardii strain that secretes BpGH16A, namely, SB_16A_D. SB_16A_D strain produced 0.80 g/L NeoDP4 (Fig. 5) at 72 h. Compared to that when using a plasmid vector system with an auxotrophic marker, the final OD 600 after 72 h of fermentation of the strain constructed using the CRISPR-Cas9 system was 2.3 times higher, but the NeoDP4 production was lower. This difference was presumed to be due to the relatively strong constitutive promoter and the high copy number of the pRS426GPD plasmid [45]. Nevertheless, the successful protein secretion from S. boulardii using genomic integration showed that S. boulardii could be used as a microbial cell factory for producing useful proteins and their products in the human gut.

Conclusions
We have, for the first time, demonstrated that NAOSs can be produced by the probiotic yeast S. boulardii. Our signal peptide screening results provide more options available in S. boulardii. We also succeeded in producing health beneficial substances using probiotic yeast harboring BpGH16A, an endo-type β-agarase, originating from the human gut bacteria B. plebeius. Our results suggest that synbiotics can be achieved by engineered probiotic yeast that produce prebiotics in the human gut.