Identification of enzymes responsible for extracellular alginate depolymerization and alginate metabolism in Vibrio algivorus

Alginate is a marine non-food-competing polysaccharide that has potential applications in biorefinery. Owing to its large size (molecular weight >300,000 Da), alginate cannot pass through the bacterial cell membrane. Therefore, bacteria that utilize alginate are presumed to have an enzyme that degrades extracellular alginate. Recently, Vibrio algivorus sp. SA2T was identified as a novel alginate-decomposing and alginate-utilizing species. However, little is known about the mechanism of alginate degradation and metabolism in this species. To address this issue, we screened the V. algivorus genomic DNA library for genes encoding polysaccharide-decomposing enzymes using a novel double-layer plate screening method and identified alyB as a candidate. Most identified alginate-decomposing enzymes (i.e., alginate lyases) must be concentrated and purified before extracellular alginate depolymerization. AlyB of V. algivorus heterologously expressed in Escherichia coli depolymerized extracellular alginate without requiring concentration or purification. We found seven homologues in the V. algivorus genome (alyB, alyD, oalA, oalB, oalC, dehR, and toaA) that are thought to encode enzymes responsible for alginate transport and metabolism. Introducing these genes into E. coli enabled the cells to assimilate soluble alginate depolymerized by V. algivorus AlyB as the sole carbon source. The alginate was bioconverted into l-lysine (43.3 mg/l) in E. coli strain AJIK01. These findings demonstrate a simple and novel screening method for identifying polysaccharide-degrading enzymes in bacteria and provide a simple alginate biocatalyst and fermentation system with potential applications in industrial biorefinery.


Introduction
Alginate is an abundant sugar in marine brown macroalgae (Chapman 1970) that is considered as an efficient and nonfood-competing candidate raw material for biorefinery. Sugars from cane or corn starch are currently the major raw materials of biorefinery; however, an ethical challenge associated with their production is that it competes with food production. In contrast, the cultivation of marine brown macro algae does not require arable land, fresh water, pesticide, or fertilizer (John et al. 2011) and has the advantage of rapid growth (Stephens et al. 2013).
Alginate consists of a long-chain polymer of α-L-guluronic acid and β-D-mannuronic acid that form a high molecular weight macromolecule (>300,000 Da) that is poorly soluble in water (Gacesa 1988). Raw alginate is too large to import through the cell membrane. As such, most microorganisms cannot degrade and utilize alginate. However, novel Electronic supplementary material The online version of this article (doi:10.1007/s00253-016-8021-7) contains supplementary material, which is available to authorized users. alginate-utilizing microbial species have recently been discovered (Kita et al. 2015;Doi et al. 2016), while various fermentation processes using alginate as raw material have been proposed. For example, ethanol for biofuel production has been derived from alginate by fermentation using metabolically engineered Sphingomonas sp. A1 strain (Takeda et al. 2011), Escherichia coli (Wargacki et al. 2012), and Saccharomyces cerevisiae (Enquist-Newman et al. 2014), while pyruvate has been produced by Sphingomonas sp. A1 (Kawai et al. 2014). These studies exploit specific alginate-assimilating species and/or their enzymes. However, there are certain challenges for the industrialization of alginate fermentation, including the need to pre-treat alginate for degradation and ensuring efficient bioconversion of alginate into products. These can potentially be circumvented by identifying novel alginatedegrading and alginate-utilizing enzymes, which could be used for the production of commodity chemicals such as Llysine, a food and feed additive. To date, there have been no reports of L-lysine production from alginate as a carbon source, although the bioconversion of D-glucose to L-lysine in E. coli strain AJIK01 has been described (Doi et al. 2015).
We recently isolated Vibrio algivorus sp. strain SA2 T from the gut flora of the turban shell marine snail (Doi et al. 2016), which can depolymerize and assimilate alginate as a sole carbon source, although the underlying mechanisms are unclear. To address this issue, in the present study we developed a novel system to screen for genes encoding extracellular active alginate-degrading enzymes and identified V. algivorus alyB as a candidate. The protein was expressed in E. coli and exhibited extracellular alginate-depolymerizing activity. We also identified seven putative alginate utilization pathway genes in V. algivorus (toaA, alyB, alyD, oalA, oalB, oalC, and dehR) that were expressed in wild-type E. coli and conferred the cells with the capacity to convert depolymerized alginate into Llysine. This is the first report identifying genes encoding enzymes for alginate degradation and utilization in V. algivorus and demonstrating the bioconversion of alginate into L-lysine.

Bacterial strains and plasmids
All strains and plasmids used in this study are listed in Table 1. Primers used for the construction of plasmids and strains are listed in Table S1. DNA fragments were PCR-amplified and purified with the Wizard SV Gel and PCR Clean-up system (Promega, Madison, WI, USA). To construct plasmids, purified insert DNA was cloned into the linearized vector with the In-Fusion HD PCR cloning system (Clontech, Mountain View, CA, USA). To express V. algivorus sp. SA2 T genes (dehR, alyB, alyD, oalA, oalB, and oalC) in E. coli, the genes were PCR amplified and inserted downstream of the chloramphenicol resistance marker gene cat and the promoter sequence attR-cat-attL-P 14 using the plasmids shown in Table 1 as a template (i.e. pM08, pM03, pM09, pM10, pM11, and pM12) and the λ-red system (Datsenko and Wanner 2000). The cat gene was excised from the genome as previously described (Katashkina et al. 2009). To express the V. algivorus sp. SA2 T toaA gene in E. coli, we carried out crossover PCR amplification of toaA (using V. algivorus sp. SA2 T genomic DNA as the template) and inserted the amplicon downstream of attR-cat-attL-P tac1000 , which was amplified using a chemically synthesized DNA template (Katashkina et al. 2005) and the λ-red system. The cat gene was then excised from the genome as described above.

Screen for extracellular active alginate lyase
We prepared sheared 40-50 kb DNA fragments of V. algivorus sp. SA2 T genomic DNA using HydroShear (Gene Machines, San Carlos, CA, USA) followed by gel purification. E. coli EPI300 cells were transformed with the sheared fragments as previously described (Wargacki et al. 2012). E. coli EPI300 colonies expressing the fragments were covered with M9 minimal medium (Miller 1992) containing 4 g/l agar and 10 g/l sodium alginate (300-400 cP, CAS no. 9005-38-3; Wako Pure Chemical Industries, Osaka, Japan) (Fig. S1A). After a 16-h incubation at 37°C, we observed an indentation over the colonies (Fig. S1B), which were transferred to a Luria-Bertani (LB) plate containing 12.5 mg/l chloramphenicol (Cm), 1 mM isopropyl β-D-1-thiogalactopyranoside, and 40 μg/l X-gal. Colonies were then cultured in liquid LB medium containing 12.5 mg/l Cm at 37°C with constant shaking at 120 rpm until the optical density at 600 nm (OD 600 ) was 0.8. We added 0.1% (v/v) of Copy Control Induction Solution (Epicentre, Madison, WI, USA) and incubated the cultures at 37°C for 6 h with constant shaking at 120 rpm. Cells were collected by centrifugation for 20 min at 3000×g and fosmid DNA was extracted using the Plasmid Midi kit (Qiagen, Hilden, Germany). The terminal sequences of extracted fosmids were determined by Sanger sequencing using EPI forward and reverse primers ( Table S1). The whole insert sequence of the extracted fosmid was determined based on the draft genome of SA2 T (Doi et al. 2016).
In vitro determination of alginate lyase activity V. splendidus ATCC33125 T and V. algivorus sp. SA2 T cells were grown on LB plates with 15 g/l NaCl for 16 h at 30°C. The cells were scraped and crude cell lysates were obtained using BugBuster Master Mix (Merck Millipore, Billerica, MA, USA). A similar procedure was used to obtain crude cell lysates of E. coli grown on LB plates with 40 mg/l Cm for 16 h at 30°C. Protein concentration of the lysates was measured as previously described (Chial and Splittgerber 1993) using Coomassie Brilliant Blue (CBB) G (Nacalai Tesque, Kyoto, Japan). Alginate lyase activity was measured using a published protocol (Iwamoto et al. 2001;Tang et al. 2009). Briefly, 180 μl of the reaction mixture (0.08 g/l protein sample, M9 minimal medium, and 2 g/l sodium alginate) was transferred to a 96-well microplate (Greiner Bio One, Frickenhausen, Germany) followed by incubation at different temperatures (34, 37, 40, and 44°C). After 18 min, the increase in absorbance at 235 nm (Abs 235 ) was measured with a Spectra Max190 microplate reader (Molecular Devices, Sunnyvale, CA, USA). One unit of alginate lyase activity was defined as an increase in Abs 235 of 0.100 per minute (Iwamoto et al. 2001).

Preparation of supernatant and washed cell samples
E. coli strains were grown overnight at 37°C on LB plates. Cells were then inoculated into 40 ml of fermentation medium composed of 20 g/l glucose, 5 g/l tryptone, 2.5 g/l yeast extract, 5 g/l NaCl, 40 mg/l Cm, and 0.3 M 3-(N-morpholino) propanesulphonic acid (MOPS; adjusted to pH 7.0 with NaOH) in a Sakaguchi flask at an initial OD 600 of 0.05 at 37°C for 20 h with shaking at 120 rpm. The cultures were centrifuged at 5000 rpm (7830×g) and 4°C for 10 min (CR20GIII; Hitachi, Tokyo, Japan). The pellet was washed three times with 5 ml of 0.85% NaCl followed by centrifugation at 5000 rpm and 4°C for 10 min.

Determination of protein concentration in the supernatant
The supernatant from the above-described cultures was concentrated using the Amicon Ultra-15 centrifugal filter unit with an Ultracel-50 membrane (Merck Millipore). Protein concentration was measured with the CBB assay using bovine serum albumin (Bio-Rad, Hercules, CA, USA) as a standard.

Preparation of alginate depolymerized with commercial alginate lyase
Sodium alginate (2.5 g) and 50 ml distilled water were mixed in a Sakaguchi flask (500 ml) at 120 rpm and 37°C for 16 h to obtain a uniformly dispersed, clear alginate gel containing 50 g/l sodium alginate. We added 40 ml of 0.85% NaCl and 1 mg/l commercial alginate lyase (A1603; Sigma-Aldrich, St. Louis, MO, USA) with shaking at 120 rpm and 37°C for 25 h.
Preparation of alginate depolymerized with alginate lyase purified from E. coli E. coli strain JM109/pM03 was grown at 37°C on LB plates containing 40 mg/l Cm for 16 h. Colonies were inoculated into 40 ml of fermentation medium composed of 20 g/l glucose, 5 g/l tryptone, 2.5 g/l yeast extract, 5 g/l NaCl, 40 mg/l Cm, and 0.3 M MOPS (adjusted to pH 7.0 with NaOH) in a Sakaguchi flask at an initial OD 600 of 0.05 and incubated at 37°C for 20 h with shaking at 120 rpm. Cells were collected by centrifugation at 5000 rpm and 4°C for 10 min then inoculated in 0.85% NaCl such that the total volume was 40 ml.
The mixture was combined with the uniformly dispersed clear alginate gel containing 50 g/l sodium alginate and incubated at 37°C for 25 h with shaking at 120 rpm. After centrifugation at 5000 rpm and 4°C for 10 min, the supernatant containing depolymerized alginate was recovered and sterilized using Nalgene Rapid-Flow filters (pore size 0.2 μm; Thermo Fisher Scientific) before addition of minimal medium.

Test-tube cultivation with minimal medium
For test-tube cultivation on minimal medium, E. coli MG1655 and its derivative strains were grown overnight at 37°C on LB plates. Colonies were inoculated into 5 ml M9 minimal medium supplemented with 1 mM MgSO 4 , 0.001% thiamine, and different carbon sources in L-shaped test tubes at an initial OD 600 of 0.05 and incubated at 37°C for 96 h with constant shaking at 70 rpm on a TVS062 CA rocking incubator (Advantec, Tokyo, Japan).

Test-tube cultivation for L-lysine bioconversion
Test-tube cultivation for L-lysine bioconversion was carried out by growing E. coli AJIK01 and its derivative strains overnight at 37°C on LB plates. Colonies were inoculated into 5 ml of medium composed of 0.25 g/l yeast extract, 4 g/l (NH 4 ) 2 SO 4 , 0.01 g/l FeSO 4 ·7H 2 O, 0.01 g/l MnSO 4 ·7H 2 O, 0.246 g/l MgSO 4 ·7H 2 O, 0.001% thiamine, 10 g/l piperazine-N, N′-bis(2-ethanesulphonic acid) (adjusted to pH 7.0 with NaOH), and different carbon sources in test tubes at an initial dry cell weight of 0.05 g/l. Cells were cultured at 34°C for 90 h with constant shaking at 120 rpm.

Results
Screen for extracellular alginate-decomposing enzyme in V. algivorus sp. SA2 T genome and phenotypic analysis of mutant AlyB We used a plate assay method to screen colonies with extracellular alginate-decomposing activity (Fig. S1A, B). A colony of the SA2 T strain (Doi et al. 2016) was first prepared and covered with a layer of alginate-containing gel. A visible indentation in the gel formed over the colony (data not shown), which was presumed to result from the decomposition of alginate by the underlying colony. Over 3000 E. coli EPI300 colonies harbouring a fosmid containing a 40-to 50-kb fragment of SA2 T genomic DNA were inoculated on LB plates and covered with a layer of gel. After 16 h, three indentations appeared; the corresponding colonies were harvested and the fosmid DNA was extracted and sequenced. All three colonies harboured the same polysaccharide lyase family gene that showed high similarity to alyB of V. splendidus (Badur et al. 2015); we therefore named the gene alyB of V. algivorus. The gene was inserted into an expression plasmid (pM03) that was introduced into E. coli JM109 (Table 1). The in vitro alginate lyase activity of V. algivorus SA2 T , V. splendidus, and E. coli JM109/pM03 whole cell lysates was analysed. V. algivorus SA2 T and V. splendidus lysates showed alginate lyase activities; the activities in the JM109/pM03 lysate were significantly higher than those in the V. algivorus SA2 T and V. splendidus lysates (Fig. 1). We also constructed a vector control strain (E. coli JM109/pM01) and a strain expressing alyB of V. splendidus (E. coli JM109/pM02) (Table 1). JM109/pM03 but not JM109/pM01 and JM109/pM02 grown on LB medium lowered the viscosity of alginate in the alginate viscosity test (Figs. 2a and S1). Alginate was added to the supernatants of JM109/pM01 and JM109/pM03 cultures, followed by incubation; the supernatants were then analysed by GPC (Fig. 3b). Alginate exhibited a single peak within the retention time of 7.7-10 min; the centre of the peak was at 8.7 min (Fig. 3b). The alginate peak was lower for the supernatant of JM109/pM03, although the peak was retained in the supernatant of JM109/pM01 (Fig. 3b).
The domain structure of V. algivorus AlyB was modelled with SignalP 4.0 (http://www.cbs.dtu.dk/services/SignalP/; Petersen et al. 2011) and Pfam 25.0 (http://pfam.xfam.org; Finn et al. 2014) software. AlyB had a 17-amino acid (a.a.) signal peptide (SP) for secretion, a 123-a.a. CBM32 domain, and a 265-a.a. PL7 domain (Fig. 4a, b). We constructed plasmids harbouring AlyB sequences lacking each of these domains (Fig. 4c); the plasmids were then introduced into E. coli JM109 and the alginate viscosity test was performed (Fig. 2). The viscosity-reducing activity was lost in the supernatant of cells expressing alyB lacking the N-terminal SP (Table 2), while the cell lysate of the SP deletion mutant retained this activity. PL7 deletion also caused the loss of viscosity-lowering activity in the supernatant, washed cells, cell lysate, and whole broth ( Table 2). Deletion of the CBM32 domain had no effect on viscosity, which was similar to that observed with wild-type alyB expression.
We assessed the alginate depolymerization potential of the filtered and concentrated supernatants. The protein concentration of the supernatant was measured with the CBB protein assay, and the same amount of protein from each sample was Fig. 1 Results of alginate lyase enzyme activity assay. 1, crude lysate of V. splendidus ATCC33125 T ; 2, crude lysate of V. algivorus SA2 T ; 3, crude lysate of E. coli JM109/pM01 (vector control); 4, crude lysate of E. coli JM109/pM03 (harbouring alyB of V. algivorus). Blue, red, green, and purple bars represent activity at 34, 37, 40, and 44°C, respectively incubated with alginate. The decrease in the amount of alginate and average molecular weights were evaluated by GPC (Fig. 5). We found that deletion of SP or PL7 reduced alginatedecomposing activity (Fig. 5a, b). A similar observation was made upon deletion of the CBM domain, although in this case half of the activity remained (Fig. 5a, b). A western blot analysis of the supernatants revealed a band of the same size as AlyB (57 kDa), which disappeared upon deletion of the SP domain (Fig. 5c).

Artificial alginate assimilation by E. coli expressing alginate metabolism pathway genes of V. algivorus
We searched the draft whole-genome sequence of SA2 T (Doi et al. 2016) for genes with sequences homologous to those encoding alginate utilization-related enzymes of V. splendidus 12B01 (Wargacki et al. 2012) using GENETYX v.10 software (GENETYX, Osaka, Japan). Seven such genes were identified (Table 3): alyB, alyD, oalA, oalB, oalC, dehR, and toaA were introduced into wildtype E. coli MG1655 cells, which were then cultured in M9 minimal medium with depolymerized alginate as the sole carbon source. Expression of the seven genes along with alyB enabled E. coli MG1655 cells to utilize the alginate that was depolymerized by AlyB activity (Fig. 6b, d).
Bioconversion of alginate to L-lysine E. coli strain AJIK01 can utilize D-glucose and accumulate Llysine (Doi et al. 2015). The introduction of the seven genes homologous to V. algivorus alginate metabolism genes (alyB, alyD, oalA, oalB, oalC, dehR, and toaA) into E. coli strain AJIK01 enabled the cells to utilize the alginate depolymerized by AlyB activity and accumulate L-lysine (Fig. 7a, b). There was no L-lysine accumulation in the absence of depolymerized alginate (Fig. 7b2) and without expression of all seven genes (Fig. 7b3).

Discussion
Macroalgal utilization requires viscous polysaccharide decomposition. Brown macroalgae contain alginate, whereas red macroalgae contain carrageenan, xylan, and agarose (Chapman 1970); these are all high molecular weight polysaccharides that cannot pass through the cell membrane. It is therefore important to identify enzymes that can decompose these molecules. In this study, we used a simple double-layer screening procedure to identify AlyB, an extracellular polysaccharide-decomposing enzyme in V. algivorus, an alginate-assimilating strain. Like alginate, most polysaccharides are viscous; therefore, this double-layer screening procedure can be used to identify other polysaccharidedecomposing enzymes by adding a polysaccharide other than alginate to the upper layer gel.
Alginate degradation is important for its industrial applications (Wang da et al. 2014) and is necessary for microbial alginate assimilation. Many Vibrio species are presumed to have enzymes for extracellular alginate degradation; indeed, several Vibrio alginate lyases such as AlyVI from Vibrio sp. QY101 (Han et al. 2004) and OalA, OalB, OalC, AlyA, AlyB, AlyD, and AlyE from V. splendidus 12B01 (Jagtap et al. 2014;Badur et al. 2015) have been purified and characterized. V. algivorus AlyB exhibited alginate-depolymerizing activity when expressed in E. coli EPI300 and JM109 cells without requiring any purification or concentration.
Deletion of PL7 resulted in the loss of alginatedepolymerizing activity, suggesting that the PL7 domain is the active centre of this enzyme, similar to the alginate lyase of V. splendidus (Badur et al. 2015).
Deleting the SP domain caused the alginate-decomposing activity to be lost in the supernatant and washed cells but not in the cell lysate. The GPC analysis confirmed that the lysate from cells expressing the SP deletion mutant showed a reduced peak corresponding to alginate (data not shown). However, a western blot analysis showed that the AlyB signal  Liquefaction of alginate-containing medium. 1, LB medium with 50 g/l sodium alginate; 2, JM109/pM01 broth with 50 g/l sodium alginate (vector control); 3, JM109/pM02 broth with 50 g/l sodium alginate (whole broth of cells expressing alyB of V. splendidus); 4, JM109/pM03 broth with 50 g/l sodium alginate (whole broth of cells expressing alyB of V. algivorus). b Results of GPC analysis. Black, 2 g/l sodium alginate standard; pink, LB medium; blue, 2 g/l sodium alginate after processing with the supernatant of JM109/pM01; brown, 2 g/l sodium alginate after processing with the supernatant of JM109/pM03 in the supernatant was undetectable upon deletion of SP. The most recent study of V. splendidus alginate lyase used mutant AlyB protein in which a His tag replaced the N-terminal SP for all the experiments and did not discuss the function of the SP (Badur et al. 2015). Our results show that wild-type AlyB of V. splendidus is not released into the supernatant irrespective of the presence of the SP; in contrast, in V. algivorus, the release of AlyB into the extracellular medium is dependent on this domain.
The CBM32 domain of V. splendidus AlyB is essential for its alginate lyase activity (Badur et al. 2015). In this study, deletion of the CBM32 domain of V. algivorus AlyB reduced the extracellular alginate depolymerization potential by 50%, whereas extra-and intracellular alginate-decomposing activity was not lost completely. These results suggest that the function of the CBM32 domain differs in the two species.
Alginate depolymerized by Aly enzymes has many uses, including as an ingredient of fish jelly food products (Sato et al. 2005), an agent that promotes plant growth (Yonemoto et al. 1995), and as an antioxidant (Falkeborg et al. 2014). However, these require alginate concentration and purification prior to depolymerization, which is unsuitable for large-scale applications due to the high cost. The fact that V. algivorus AlyB does not require purification makes it a good candidate for industrial production.
ToaA expression in E. coli increased alginate utilization and cell growth, although the import of depolymerized soluble alginate into the cytosol by V. algivorus ToaA was a ratelimiting step. E. coli cells expressing ToaA solely were unable to grow using enzymatically decomposed extracellular alginate as the sole carbon source (data not shown). We therefore speculate that other enzymes involved in alginate metabolism (alyB, alyD, oalA, oalB, oalC, and dehR) also function in alginate utilization. Our results suggest that V. algivorus  Fig. 7 Results of test-tube cultivation for L-lysine bioconversion. a Dry cell weight (DCW) accumulation. b L-lysine accumulation. 1, AJIK01 strain without depolymerized soluble alginate; 2, D3000 strain without depolymerized soluble alginate; 3, AJIK01 strain with alginate depolymerized by AlyB of V. algivorus expressed in E. coli; 4, D3000 strain with alginate depolymerized by AlyB of V. algivorus expressed in E. coli assimilates depolymerized alginate via a single uronic acid molecule (beta-D-mannuronate and 4-deoxy-l-erythro-5hexoseulose urinate), similar to what has been predicted for V. splendidus (Wargacki et al. 2012) and Sphingomonas sp. strain A1 (Takeda et al. 2011). We propose a model of alginate degradation and metabolism in V. algivorus SA2 T based on our findings (Fig. 8).
In this study, we evaluated the potential for producing alginate derivatives using engineered E. coli. One possible derivative is L-lysine, a commodity chemical that is commercially produced by fermentation in the order of over 1,500,000 metric tons per year (Doi et al. 2014). There are no reports to date describing the bioconversion of alginate to L-lysine. We observed here that the alginate depolymerized by AlyB turned into cell biomass and L-lysine.
We have not introduced the previously reported technologies for improving L-lysine productivity of E. coli in this study. Concrete examples of such technologies are the introduction of feedback resistant L-lysine biosynthesis genes (Kojima et al. 1994), the introduction of heterologous ddh gene of Corynebacterium glutamicum, and the attenuation of the meso-α,ε-diaminopimelic acid synthesis pathway (Doi and Ueda 2009). In the future, L-lysine production from alginic acid can be further developed with introduction of these technologies.
Metabolically engineered E. coli fermentation can be used to produce other commodity chemicals such as L-glutamate (Nishio et al. 2013), L-tryptophan ), L-phenylalanine (Báez-Viveros et al. 2007), lactic acid (Niu et al. 2014), and succinic acid (Zhu et al. 2013). Our results indicate that expressing V. algivorus genes in E. coli is a suitable alginate degradation and utilization system that can be used in combination with fermentation to produce commercially valuable chemicals.
A higher bioconversion efficiency is necessary for the industrialization of alginate fermentation. This requires optimizing the expression levels of alginate metabolism enzymes. A recent study reported that a high-throughput method for constructing recombinant variants known as recombinase-assisted genome engineering enhanced the efficiency of bioconversion of alginate into ethanol by recombinant E. coli (Santos et al. 2013). Other efficient high-throughput engineering technologies have also been reported for E. coli (Esvelt and Wang 2013;Ronda et al. 2016). These approaches can broaden the industrial applications of V. algivorus AlyB and other enzymes involved in alginate depolymerization and metabolism.