Fermentation medium optimization, molecular modelling and docking analysis of the alginate lyase of a novel Pseudomonas sp. LB56 isolated from seaweed waste

Abstract Alginate lyase has the potential for the ecofriendly depolymerization of alginate and generation of value-added products important in agriculture, bioenergy, food processing, medical diagnostic and pharmaceutical industries. Substantial research has focused on the cost-effective production of alginate lyase via screening of potent microbial isolates for use as biocatalysts and the optimization of fermentation media with low-cost feedstocks. This study evaluated the alginate lyase-producing efficacy of a novel Pseudomonas sp. LB56 (GenBank accession no. MT176182) isolated from Sargassum seaweed waste. The feedstock for solid-state fermentation (SSF) was decomposed Sargassum biomass amended with cow dung. The production of alginate lyase was maximal using 9.5 g Sargassum, 14 g cow dung and at pH 8 after optimization via a 23–factorial central composite design. Comparative modelling and molecular docking studies were conducted to elucidate the 3D structure of the alginate lyase of this novel isolate, as well as to investigate its catalytic interactions with different polymeric sizes of alginate [polymannuronate (mannuronate disaccharide and mannuronate trisaccharide), polyguluronate (guluronate disaccharide and guluronate trisaccharide) and alginate]. The overall results suggest that biowaste (Sargassum seaweed amended with cow dung) can be used as a feedstock for economical production of alginate lyase. Additionally, the identified active site residues of alginate lyase, which were shared by different polymers of alginate, can be tailored via rational and/or combinatorial approaches for efficient biotransformation of alginate-enriched wastes into biologically active alginate oligomers and third generation biofuels.


Introduction
The increase of nutrient-loading in coastal waters from agricultural, storm water and municipal runoff has a negative effect on the global marine ecosystem by causing microalgal and seaweed blooms [1][2][3]. Additionally, the changes in land-use practices and climate have been reported as the contributory environmental factors for triggering the blooms of seaweed [4]. The bloom-forming ability of seaweed has been documented for the species of three phyla, Chlorophyta, Phaeophyta and Rhodophyta. In recent years, two floating species of Sargassum (Phylum: Phaeophyta), S. fluitans and S. natans, have caused large blooms annually in the coastal regions of the Caribbean, Mexico, the USA, West Africa and South America [2,5]. In 2018, these blooms covered a distance of 8850 km and the biomass of the bloom was estimated as 20 million tons via satellite imageries [6]. Sargassum bloom is a blow to the regional economy since it adversely affects the coastal biodiversity as well as the seafood and tourism industries [4]. The coastal inundation of Sargassum biomass is also a public health concern owing to the emission of toxic hydrogen sulphide and insect infestation which can cause skin irritation [7, 8,].
Recently, there has been a growing industrial interest in the use of Sargassum biomass as a feedstock in biorefinery for the production of third generation biofuels and value-added chemicals including platform chemicals and biofertilizers [9][10][11]. The advantages of utilizing the biomass of Sargassum as a feedstock include the absence of recalcitrant lignin and no requirements of land, fertilizer and freshwater for their cultivation [12]. The major structural component of Sargassum is alginate (ca. 40% of dry weight) [11,13]. Alginate has been commercially extracted from seaweed species, such as Macrocystis, Laminaria, Ascophyllum and Sargassum of phylum Phaeophyta, because of its extensive industrial applications as a gelling agent in food processing, as a hydrogel for treating burns and as a carrier for drug delivery [14].
Alginate is a linear acidic polysaccharide and consists of two types of uronate, β-D-mannuronate (M) and its C5 epimer α-l-guluronate (G). In alginate, the G and M types of uronates form three different blocks, homopolymeric mannuronate (polyM), homopolymeric guluronate (polyG) and heteropolymeric polyGM [15]. The depolymerization of alginate into alginate oligosaccharides was accomplished by physicochemical processes (hydrothermal treatment and/or acidic/alkaline hydrolysis) [16]. Generally, these processes generate secondary pollution. Therefore, there has been substantial industrial interest in developing environmentally friendly bioprocesses for sustainable depolymerization of alginate via the application of microbial enzymes. The advantages of bioprocess include operation at ambient temperature and pressure, high substrate specificity and product yield, maximal utilization of raw materials and minimal generation of undesirable by-products [17,18]. Alginate lyase is the enzyme that degrades alginate via a β-elimination reaction with cleavage of the 1,4 glycosidic linkage. This enzymatic reaction produces unsaturated oligosaccharides and monosaccharides of uronic acids. Alginate lyase has a preference for the substrate polyM (EC 4.2.2.3) or polyG (EC 4.2.2.11) [15,19]. Some microbial alginate lyases can cleave polyGM blocks [20]. Furthermore, the enzymatic cleavage is either endolytic, by internally cleaving the blocks with the production of unsaturated alginate oligosaccharides, or exolytic, by cleaving the blocks at the end with the production of unsaturated monosaccharides [19,20].
In recent years, alginate lyase has attracted considerable interest for the biodegradation of alginate into biologically active alginate oligosaccharides and other value-added chemicals which may have potential applications in agriculture, biorefinery, food processing, medical diagnostic and pharmaceutical industries [21]. The high cost of commercially available alginate lyase is sufficient motivation for exploring a more cost-effective approach for the production of this enzyme. This may be accomplished by, for example, the screening of potent microbial isolates from diverse environments to be used as the biocatalyst and optimization of the fermentation medium using economical feedstocks. The advantage of microbial biocatalysts in a bioprocess is that the crude enzyme can be used directly, thereby circumventing the expensive purification steps [22,23]. In the present study, the alginate lyase-producing efficacy of a novel Pseudomonas sp. LB56 isolate was assessed by utilizing decomposed Sargassum biomass amended with cow dung as the feedstocks in solid-state fermentation (SSF). Cow dung makes available nutrients (trace elements and vitamins) essential for the growth and proliferation of microorganisms [24]. The global annual production of livestock faecal matters has been documented as 13 billion tons [25]. Because cow dung is abundant, it has emerged as an inexpensive feedstock for the production of industrial enzymes [26,27]. The conditions of SSF were optimized for the maximal production of alginate lyase by this novel Pseudomonas sp. LB56 via response surface methodology (RSM). Furthermore, comparative modelling and molecular docking studies were also performed to assess the structure-function relationship of the alginate lyase of this novel isolate.

Alginate-cleaving bacterial isolate
The potent alginate-cleaving bacterial isolate LB56 was isolated from the Sargassum seaweed waste off Barbados' coast using serial dilutions in sterile seawater [28] and spread plate methods with a mineral medium (0.02% MgSO 4 , 0.002% CaCl 2 , 0.1% KH 2 PO 4 , 0.1% K 2 HPO 4 , 2% NaCl and 0.1% NH 4 NO 3 ) amended with 0.5% (w/v) sodium alginate (Sigma-Aldrich, Saint Louis, MO) and 1.5% (w/v) agar. The resulting bacterial colonies were purified by streaking on Luria-Bertani (LB) agar amended with 2% NaCl and screened for alginate lyase activities by spot inoculation onto the above-mentioned alginate-containing mineral agar plates [29]. After incubation of the plates at 30 °C for 24 h, the plates were flooded with Gram's iodine solution. The alginate lyase positive isolates exhibited a halo around the colony. LB56 isolate displayed a halo diameter of 17 mm.

Phylogeny of LB56 isolate
The phylogeny of potent alginate-cleaving bacterial isolate LB56 was determined via amplification and sequencing of the 16S rRNA gene of its genomic DNA. The genomic DNA of LB56 was isolated using InstaGene TM DNA extraction kit (Bio-Rad Laboratories, Canada) and the amplification of the 16S rRNA gene was carried out in a polymerase chain reaction (PCR) reaction mixture containing Thermo Scientific TM 2X PCR Master Mix, 27 F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492 R (5′-GGTTACCTTGTTACGACTT-3′) primers set, genomic DNA of LB56 and nuclease-free water. An Edvotec TM thermal cycler was used for the 16S rRNA gene amplification with the PCR conditions, initial denaturation at 95 °C for 5 min; 30 cycles of denaturation at 94 °C for 30 s, annealing at 56 °C for 1 min, extension at 72 °C for 90 s; final extension at 72 °C for 10 min [30]. The size of the resulting 16S rRNA gene amplicon (ca. 1500 bp) was checked with 1% (w/v) agarose gel electrophoresis [30]. Then the PCR product was purified with IBI PCR/DNA fragment gel extraction kit (IBI Scientific, Peosta, IA) and sequenced using an ABI 3730XL DNA analyser (Applied Biosystems, Foster City, CA). The raw 16S rRNA gene sequence was analyzed with the BioEdit software [31]. The consensus sequence of LB56 was subjected to a BLASTn search using the 16S rRNA sequences (Bacteria and Archaea) database of GenBank to determine its identity. The consensus 16S rRNA gene sequence of LB56 was submitted to GenBank and obtained an accession number MT176182. A bootstrap neighbour-joining dendrogram with Jukes-Cantor algorithm and 1000 iterations was drawn using MEGA 7.0 [32] to depict the phylogenetic relatedness.

SSF of feedstocks
Cow dung samples were collected from a dairy farm in St. James, Barbados. Decomposed Sargassum biomass was collected from Long Beach, Barbados and rinsed with distilled water. The cow dung and Sargassum samples were processed via autoclaving at 121 °C for 20 min and then dried at 60 °C for 24 h. Afterward, both samples were ground individually and sifted through a 5 mm nominal pore size sieve [33]. The optimization of the SSF conditions (Sargassum biomass, cow dung and pH) was performed for the maximal production of alginate lyase in sterile 250 mL Erlenmeyer flasks by applying a 2 3 -factorial central composite design (CCD) with six centre points [34]. In CCD, alginate lyase specific activity (U AL /mg-protein) was denoted as the dependent variable, whereas Sargassum (X 1 ), cow dung (X 2 ) and pH (X 3 ) were designated as independent variables (Table 1). In order to estimate the optimal values of X 1 , X 2 and X 3 , 20 SSF experiments were carried out with different amounts and combinations of these variables (Table 2). After addition of different amounts of X 1 and X 2 to each flask, the substrates were mixed with 20 mL of sterile buffers with different pH values (50 mmol/L phosphate-citrate for pH 6.32; 50 mmol/L Trizma TM for pH 7 and 8; and 50 mmol/L borate for pH 9 and 9.68) and inoculation of 1 mL cell resuspension of LB56 (6 × 10 6 cells/mL). The addition of buffers to the feedstock also maintained the humidity at 95-100%. LB56 cell resuspension was prepared by culturing the cells in LB broth supplemented with 2% NaCl at 30 °C for 72 h. The cell density was adjusted to 6 × 10 6 cells/ mL with the mineral medium specified above (see Alginate-cleaving bacterial isolate). The flasks were incubated at 30 °C for 5 days. After addition of 50 mL of 50 mmol/L Trizma TM buffer (pH 7.5) to each flask, the contents were mixed by shaking at 200 rpm for 60 min at 25 °C. The resulting resuspensions were filtered through Whatman no. 1 filter papers, followed by centrifugation (12 000 × g, 4 °C) for 15 min [33]. The supernatants were used for assaying the alginate lyase activities.

Assay of alginate lyase activity
The standard reaction mixture consisted of 0.1 mL of enzyme preparation, 0.1 mL of 50 mmol/L Trizma TM buffer (pH 7.5) and 2.2 mL of 0.5% (w/v) of sodium   alginate resuspended in the same buffer. The assay mixture was incubated at 45 °C for 30 min and then placed in a boiling water bath for 5 min to terminate the reaction. The unsaturated uronic acids produced during the enzymatic reaction were measured spectrophotometrically at 235 nm against a blank in which the enzyme solution was replaced by a heat-treated enzyme preparation (boiled at 100 °C for 30 min). One unit (U AL ) of alginate lyase activity was defined as an increase of 0.1 absorbance unit per minute under the standard assay conditions [35]. Protein concentration was determined by the Bradford method using bovine serum albumin as a standard [36].

Comparative modelling
The 16S rRNA gene sequence of the LB56 isolate displayed 99.8% similarity with Pseudomonas stutzeri. However, the 3D structure of P. stutzeri alginate lyase is not available yet. Therefore, the target sequence selection was made via a homology search in UniprotKB (https://www.uniprot.org/). Afterward, the target sequence in a FASTA format was incorporated into SWISS-MODEL [37] to predict the template sequence. The selected template sequence was used to predict the homology 3D structure of Pseudomonas sp. via structural alignments with Global Model quality Estimation (GMqE) and quality Model Energy Analysis (qMEAN) scores. The statistical robustness of the modelled alginate lyase of Pseudomonas sp. was also analyzed using PROCHECK [38], Verify3D [39] and ERRAT [40].
The primary structural properties of modelled alginate lyase of Pseudomonas sp. were analyzed by ExPaSy's ProtParam program [41]. Furthermore, the characteristics of the secondary structure were predicted via self-optimized prediction method with alignment (SOPMA) [42].

Molecular docking of modelled alginate lyase
The modelled alginate lyase of Pseudomonas sp.

Results
The BLASTn search of 16S rRNA gene sequence of LB56 identified the isolate as Pseudomonas sp. LB56 since it showed 99.8% similarity with P. stutzeri. The 16S rRNA gene sequence of LB56 was deposited into NCBI GenBank with an accession number MT176182. Figure 1 depicts the phylogenetic relatedness of LB56 isolate with other Pseudomonas sp.
A 2 3 -factorial CCD was applied for the optimization of SSF components, such as the amounts of feedstock (Sargassum and cow dung) and pH for maximal production of alginate lyase by the new isolate Pseudomonas sp. LB56 (Table 2). A second-order quadratic model (Eq. 1) was generated via multiple regression analysis from the CCD matrix using the experimental values of alginate lyase (Table 2) where the coded values of independent variables were specified. This quadratic model denotes alginate lyase specific activity (Y = U AL /mg-protein) as a function of the interactive effects of Sargassum (X 1 ), cow dung (X 2 ) and pH (X 3 ). . .
The statistical significance of the quadratic model was evaluated via Fisher's test for the analysis of variance (ANOVA) and the R 2 value was estimated as 0.93. This R 2 value suggests that 93% of the response of the independent variables for alginate lyase production can be detected via Eq. 1. The model is significant at 5% level since it shows a higher F value (15.5) than the tabular F 10,9 -value (3.14). Furthermore, the model is also useful in predicting the experimental results by displaying higher values of the adjusted R 2 (0.87) and lack-of-fit F (57.4) ( Table 3).
The 3D response surface plots were drawn to assess the interactions of the independent variables, X 1 , X 2 and X 3 on alginate lyase production. In the 3D response surface plots, the levels of two independent variables changed and the third variable was fixed at its zero coded-value (Figure 2(A-C)). These plots helped in estimation of the optimal values of the independent variables and visualization of their interactions. The optimal values of Sargassum, cow dung and pH were estimated as ca. 9.5 g, 14 g and 8, respectively, for maximal alginate lyase production by the new LB56 isolate via the interactive 3D response surface plots. For these optimal values of the independent variables, Figure 2. the 3D response surface plots depicting the interactive effects of SSF conditions. (a) Sargassum (X 1 )-cow dung (X 2 ), (B) cow dung (X 2 )-ph (X 3 ) and (c) Sargassum (X 1 )-ph (X 3 ). the quadratic model (Eq. 1) predicted the alginate lyase specific activity as 4.01 U AL /mg-protein. The result predicted via the quadratic model was validated by conducting experiments in triplicate with the optimized values of independent variables. These experiments yielded the alginate lyase specific activity as 5.26 ± 0.24 U AL /mg-protein, which is significantly higher (P < 0.05; one-tailed t-test) than the value predicted via the quadratic model.
The homology search via UniProtKB for the alginate lyase of P. stutzeri, which showed 99.8% of 16S rRNA sequence similarity with the new isolate LB56, selected P. putida W619 (UniProtKB ID. B1J479) as the target sequence for the comparative modelling because the amino acids sequence of this bacterium is reviewed at the protein level. Subsequently, the target sequence, when used in homology modelling via SWISS-MODEL, identified P. aeruginosa PAO1 (PDB ID 4ozw.1.A; 1.64 Å 3D structural resolution) as the template and predicted the 3D structure of the alginate lyase of Pseudomonas sp. (Figure 3(A)). The alignment of sequences between model and selected template displayed 64.7% sequence identity, 0.50 sequence similarity and 0.91 coverage in the range of 25-360 amino acids (Figure 3(B)). The predicted 3D structure of the alginate lyase exhibited the GMqE and qMEAN scores as 0.80 and −1.30, respectively. The structural robustness of the predicted alginate lyase of Pseudomonas sp. was also assessed via PROCHECK (Figure 4), Verify3D (Supplemental Figure S1(A)) and ERRAT (Supplemental Figure S1(B)).

Discussion
In the present study, the optimization of SSF conditions with respect to the feedstock (decomposed Sargassum and cow dung) quantity and pH for maximal production of alginate lyase by a novel Pseudomonas sp. LB56 isolate was performed via the RSM. After optimization via a 2 3 -factorial CCD, the maximal production of alginate lyase (5.26 ± 0.24 U AL / mg-protein) by this novel isolate was recorded with 9.5 g Sargassum, 14 g cow dung and at pH 8. A previous optimization study using a 2 2 -factorial CCD reported 15 g cow dung and 4.5 g decomposed Sargassum biomass for maximal co-production of alginate lyase and mannanase by a novel Streptomyces sp. Alg-S23 isolated from Sargassum seaweed waste [33]. The specific activities of the crude alginate lyase (U AL / mg-protein) extracted from other bacterial species were reported as 2.81, 8.3, 3.12, 6.55, 3.64 and 2.22 for Arthrobacter sp. AD-10 [21], Alteromonas sp. 272 [35], Exiguobacterium sp. Alg-S5 [29], Sphingomonas sp. A1 [44], Streptomyces sp. Alg-S23 [33] and Vibrio sp. yKW-34 [45], respectively.
The 3D structure of the alginate lyase of this novel Pseudomonas sp. LB56 isolate was predicted via homology modelling in SWISS-MODEL [37]. An advantage of homology (aka comparative) modelling is the elucidation of the relationship between structure and function via an in silico approach rather than the time-consuming and labour-and cost-intensive experimental approaches. Moreover, it is difficult to crystallize the 3D structure of the membrane-bound and larger proteins by experimental approaches [46].
The GMqE and qMEAN values indicate that the modelled alginate lyase of Pseudomonas sp. is reliable because GMqE and qMEAN scores should be between 0 and 1 [37] and <-4 [47], respectively, to give a statistically reliable model. The GMqE score of the modelled protein was calculated via the alignment of target-template sequences and the structure of the template. The degree of nativeness of the structural features of the model protein was estimated via the qMEAN score.
Additional quality checking of the predicted 3D structure of Pseudomonas sp. was also performed using three different programs; PROCHECK, Verify3D and ERRAT. PROCHECK uses a Ramachandran plot to validate the stereochemistry of the predicted 3D structure. Verify3D assesses the compatibility between the modelled 3D structure and its amino acids sequence (1D). A model protein is considered valid in Verify 3D if it returns a score of ≥0.2 for 80% of the amino acids in the 3D-1D profile [39]. ERRAT evaluates the interaction between non-bonded atoms and by linking these atoms to the residues. In ERRAT, a structurally accurate model displays an average overall quality factor score of 91% [40].
The Ramachandran plot of the torsional angles, Phi (Φ, X-axis) and Psi (Ψ, Y-axis), for the modelled alginate  lyase was generated via the PROCHECK program. This plot indicates stereochemical validity of the model structure by allocating higher numbers of residues in most favoured regions (93.3%), followed by additional allowed regions (6%), generously allowed regions (0.3%) and disallowed regions (0.3%). Additionally, Verify3D and ERRAT programs also validated the accuracy and reliability of the modelled alginate lyase by producing an average 3D-1D score of ≥ 0.2 for 100% of the residues and an average overall quality factor score of 96.6%, respectively. The numbers of negatively charged acidic (Asp and Glu) and positively charged basic (Arg and Lys) amino acids were recorded as 41 and 52, respectively, for the modelled alginate lyase during the assessment of primary structural traits. The modelled protein gave an extinction coefficient of 81610 M −1 cm −1 , assuming that all pairs of the Cys residues form cysteine at 280 nm. The recording of a larger extinction coefficient value suggests a higher Cys content [48]. The in vitro and thermal stabilities of the modelled alginate lyase were assessed from the values of the instability and aliphatic indices, respectively. The values were recorded as 34 for instability index and 70.1 for aliphatic index indicating in vitro [49] and thermal stabilities [50] of the modelled alginate lyase. The modelled alginate lyase of Pseudomonas sp. revealed the GRAVy index of −0.56, suggesting it was a hydrophilic globular protein rather than a membrane protein [51]. Additionally, the recording of high percentage of random coil regions during the assessment of secondary structural traits via SOPMA indicates the stability of modelled alginate lyase [52].
In docking analysis, the catalytic interaction of an enzyme and its substrate is assessed by identifying the key amino acid residues with the structural pose of lowest in silico binding energy score. This in silico approach has recently received attention for tailoring the biocatalytic efficacy of an enzyme in diverse industrial applications [53,54] and in identifying novel therapeutic compounds [55]. The recordings of negative docking scores of modelled alginate lyase of Pseudomonas sp. indicate the strong affinities of modelled alginate lyase towards different polymeric sizes of alginate. The residues involved in the binding processes via hydrogen bonding were Gly318, Lys319 and Ser348 for M2; Glu80, Asp313, Ser348 and Arg350 for M3; Tyr 257, Lys319, Ser348 and Arg350 for G2; Lys64, Tyr257, Asp313, Lys319, Ser348 and Arg350 for G3; and Tyr257, Asp313, Lys319, Asp347 and Ser348 for alginate. There were 11 (M2), 12 (M3), 10 (G2), 12 (G3) and 11 (alginate) residues also participating in the catalytic binding processes via hydrophobic interactions with modelled alginate lyase. The active site residues, which were shared by all the five substrates during the catalytic processes via hydrogen and hydrophobic interactions, were Lys64, Arg84, Trp203, Tyr257, Lys319, Ser348, Phe349 and Arg350. Past studies have documented the catalytic interactions of alginate lyase active site residues, Arg78, Gln 125, His127 and Tyr244, of Psychromonas sp. C-3 with M2 [56]; Asn138, Arg143, His200, Asn217, Lys308 and Tyr312, of Vibrio sp. with G3 [57]; and Tyr238, Arg241, Arg418 and Glu644, of Stenotrophomonas maltophilia KJ-2 with alginate as the substrate [58].

Conclusions
The sustainable depolymerization of alginate by alginate lyase has attracted industrial attention for cost-effective production of this enzyme. This study has demonstrated the potential of the novel Pseudomonas sp. LB56 for economical production of alginate lyase by utilizing biowaste (decomposed Sargassum biomass amended with cow dung) as the feedstock. Furthermore, the molecular modelling and docking analysis results revealed the participation of crucial active site residues of the catalytic binding processes. These active site residues of alginate lyase can be tailored via rational and/ or combinatorial approaches for sustainable and efficient bioconversion of seaweed waste into alginate oligomers having novel biotechnological applications.

Data availability statement
The author confirms that the data supporting the findings of this study are available within the article.

Disclosure statement
No potential conflict of interest was reported by the author.