Characterization of alginate extracted from Sargassum latifolium and its use in Chlorella vulgaris growth promotion and riboflavin drug delivery

Alginates derived from macroalgae have been widely used in a variety of applications due to their stability, biodegradability and biocompatibility. Alginate was extracted from Egyptian Sargassum latifolium thallus yielding 17.5% w/w. The chemical composition of S. latifolium is rich in total sugars (41.08%) and uronic acids (47.4%); while, proteins, lipids and sulfates contents are 4.61, 1.13 and 0.09%, respectively. NMR, FTIR and TGA analyses were also performed. Crystallinity index (0.334) indicates alginate semicrystalline nature. Sodium alginate hydrolysate was evaluated as Chlorella vulgaris growth promoter. The highest stimulation (0.7 g/L biomass) was achieved by using 0.3 g/L alginate hydrolysate supplementation. The highest total soluble proteins and total carbohydrates were 179.22 mg/g dry wt and 620.33 mg/g dry wt, respectively. The highest total phenolics content (27.697 mg/g dry wt.), guaiacol peroxidase activity (2.899 µmol min−1 g−1) were recorded also to 0.3 g/L alginate hydrolysate supplementation. Riboflavin-entrapped barium alginate-Arabic gum polymeric matrix (beads) was formulated to achieve 89.15% optimum drug entrapment efficiency (EE%). All formulations exhibited prolonged riboflavin release over 120 min in simulated gastric fluid, followed Higuchi model (R2 = 0.962–0.887) and Korsmeyer–Peppas model with Fickian release (n ranges from 0.204 to 0.3885).

Proton nuclear magnetic resonance ( 1 H NMR) analysis. 1 H NMR spectroscopy is a significant physicochemical method for elucidating structure of polysaccharides. Structural features of alginate are elucidated by the 1 H NMR profile (Fig. 1A) which depict the chemical shifts and single monomers and blocks characteristic to the sodium alginate fraction (Fig. 1A), revealing purity 32 . Data revealed the recognition of β-anomeric protons in the alginate sample in addition to appearance of protons signals within a 2-ppm chemical shift range (3 to 4 ppm). Assignments of the signs in the anomeric zone are recognized 33 whereas, 2 signals in the anomeric region were documented.
The present results illustrated Signal II corresponding to the overlap between mannuronic acid anomeric proton (M-1) and the H-5 of alternating blocks (GM-5) besides, Signals III corresponding to proton H-5 guluronic acid from the GG-5G block (G-5) in the alginate sample indicating the heterogeneous (FGM) blocks of alginate 34 . The 1 H NMR profile revealed presence of guluronic acid H-5 (GG-5G) that was recognized at 4.282 ppm (signal III) as suggested by Usoltseva et al. 32 and Flórez-Fernández et al. 35 . Furthermore, the characteristic signals of  24 . Moreover, the overlapping signals in the 1 H NMR could be an assignment for the structure complexity of the alginate sample 34 . Aside from alginate extracted from Sargassum vulgare by Hussein et al. 31 , most Sargassum-derived alginate have M/G < 1 and significant magnitudes of homopolymeric block M (η < 1). The heterogeneity of alginate composition could be affected by the environmental factors 23 .  Table S1) revealed different chemical functional groups. FTIR spectrogram of alginate is dominated with strong absorption broad band at 3465 cm −1 ascribing hydroxyl group (-OH stretching). Whereas, spectral peaks allocated in the range 4000 cm −1 -3400 cm −1 could be assigned to alcohol and acids 35 . Strong peaks in the range of 1628 to 1428 cm −1 designated asymmetric and symmetric stretching vibrations that attributed to carboxylate anions (COO − ) 35 . These significant spectral peaks could take a part in the structure elucidation of alginates recognizing the metal-carboxylate interactions according to Flórez-Fernández et al. 35 . The spectral band around 2923-2854 cm −1 is in agreement with that obtained by Aprilliza 36 and are attributed to aliphatic -CH stretching, and symmetrical and asymmetrical (C-H)CH 2 stretching, beside aromatic and/or vinylic C-H stretching and (CH)-anomer stretching 37 . FTIR spectrogram of alginate also illustrated peak of C-H stretching vibrations recognizing alkanes, C=O indicated carbonyl group (amide I band), COO − stretching vibrations ascribed to carboxylate as well as C-O-C stretching vibrations 36 . Peaks attained the range of 1090-1030 cm −1 are assigned to C-O stretching of pyranosyl ring, C-O-C asymmetric stretching (glycosidic linkage), C-C stretching which are attributed to alginate saccharide structure 38 . Also, the stretching of C = O group was documented at 1734 cm −1 as reported by Carpenter and Saharan 39 . The present results are in accordance with those of Cardenas-Jiron et al. 40 and Bouissil et al. 24 . According to Gomaa et al. 27 , the spectral band 848 cm −1 confirms the presence of sulfate groups of fucoidan. Peaks around 600 cm −1 could be due to symmetric and asymmetric O=S=O deformation as reported by Flórez-Fernández et al. 35 .
Thermogravimetric analysis (TGA). The thermal behavior of alginate ( Fig. 2A) indicated mass progressive decreasing pattern with increasing temperature, demonstrating four distinctive decomposition stages characterized by temperature range for each of them. The first stage began with weight loss of 10.67% in the temperature range 49.74-66.68 °C, followed by a 24.90% mass loss in 245.13-296.99 °C temperature range. Thereafter, 11.54% mass loss has documented in the temperature range 532.39-566.69 °C, ended with 8.97% mass loss within temperature range 649.02-659.01 °C. Under progressive elevating temperature, sodium alginate exhibited initial dehydration process during the first stage. The initial dehydration followed by two decomposition stages characterized by the production of carbonaceous residue. The decomposition stages followed by production of sodium carbonate with 40.91% carbonized matter residue at the end of the experiment which is degraded gradually according to the interpretation of Guedes Soares et al. 41 , exhibiting good thermal stability. The present thermal degradation behavior of sodium alginate is in agreement with those of Rani et al. 42 and dos Santos Araújo et al. 43 . On the other hand, Xu et al. 44 reported the thermal stability of calcium alginate capsules up to 160 °C. X-ray diffraction (XRD). The XRD pattern was used to investigate the micro structural features of Sargassum latifolium-derived sodium alginate (Fig. 2B). The XRD profile illustrates three distinctive intensive crystalline diffraction, which observed at 2θ degrees values of 20.51, 21.04 and 29.73 with inter planar spacing (d-spacing) of 4.328 Å, 4.22 Å and 3.003 Å, respectively indicating a rather amorphous structure. The obtained diffractogram demonstrate crystallinity index of 0.334 indicating the semi-crystalline nature of this sodium alginate sample 45 . Furthermore, owing to the interpretation of Kanimozhi et al. 46 the peak height intensity and the degree of crystallinity could diagnose the amorphous nature of alginate. In accordance with our results, Aprilliza 36 reported that sodium alginate extracted from brown algae has a semi-crystalline structure.
Rheological measurement. Rheological characteristics are significant parameters for the biotechnology issue of phycocolloids. The apparent viscosity of sodium alginate fraction in aqueous solutions (5, 10 and 15 mg/mL) was quantified as a function of shear rate in the range 20-500 s −1 that is illustrated in Fig. 3. Flow profile demonstrated that 40 s −1 shear viscosity giving maximum values of viscosity 8.03, 12.1, 22.1 centipoise (cP); respectively (Fig. 3A). At the maximum shear rate used (500 s −1 ), the viscosity exhibit marked decrease in all alginate concentrations to 2.3, 5.1, 11.2 cP; respectively designating a shear thinning behavior 47 . Non-Newtonian flow is studied via gathering the viscosity data covering the range of shear rates used to perform a rheogram representing viscosity versus shear rate. Despite the suggestion of Wildemuth and Williams 48 about the presence of conflicting issue of the distinctive shear-thinning fluids that displays an area of Newtonian flow at shear rate extremes, the present data did not illustrate any Newtonian flow behavior for the studied alginate solutions. According to the findings of Shyamali 49 , Sargassum species-derived alginates include a higher quantity of guluronic acid blocks which yielded strong gels relative to that derived from Macrocystis. The detected shear-thinning pseudoplastic performance, which is considered as an irreversible structural disturbance, and the reduction of viscosity takes place due to molecular arrangement which occurs in the structural rearrangement as indicated by Glicksman 50 . Flow curves of different concentrations of alginate (Fig. 3B) exhibited decreases in shear stresses showing a restrictive stable value at reduced values of shear rate designating limited yield stress of these solutions. Though greater yield stress levels could be reached with the high concentrations of alginate. The pseudoplastic property showed as shear thinning of alginate solutions is illustrated in Fig. 3C. The highest viscosity levels as well as the marked shear thinning behavior were documented to the highest concentration ending with the smallest mg alginate/mL. Rheogram (Fig. 3D) torque percent increases with increasing spindle speed (RPM). Figure 3E explains viscosity of alginate aqueous solutions & spindle speed (RPM) relationship. It is obvious that viscosity range is inversely proportional to the rotational speed. According to Truus et al. 51 alginate viscosity of Fucus is comparatively low depending on the gathering time and drying processes of seaweeds, whereas rheological properties of alginate are temperature dependent process. Generally, a reduction in viscosity is often noticed after increasing shear rate, whereas increasing alginate concentrations induced viscos-  52 . This viscosity performance was illustrated by polysaccharides of other algae 53 . The dynamic viscosity manners of alginate polymer are influenced by both polymer structure and mass 54 . Shear thinning pattern of the extracted alginate may be contributed to the hydrodynamic potential resulted throughout the shear decline of alginate structural units 55 . Sutherland 56 reported that polysaccharide characterized by this rheological pattern can be incorporated into many food industries to modify dynamic viscosity behavior of the present water, changing product texture according to their gel formation ability.
Growth responses of C. vulgaris to alginate hydrolysate supplementation. Different alginate hydrolysate supplementations (0, 0.05, 0.1, 0.3 and 0.5 g/L) induced various levels of promoting effects on Chlorella vulgaris growth and metabolism over 14 days incubation period as can be seen in Fig. 4 A. All the trials were performed in triplicate in order to minimize errors and to calculate standard deviation. As shown in Fig. 4A, C. vulgaris growth curve exhibits an exponential phase from the 2nd day till the 12th day after which the growth tends to be in the stationary phase. Significant increments in dry biomass (Fig. 4B), specific growth rate (Fig. 4C), protein and carbohydrate contents (Fig. 5B), total phenolics (Fig. 5C) as well as guiacol peroxidase (Fig. 5D) over the www.nature.com/scientificreports/ control experiments were documented. On the other hand, all alginate supplementations decreased chlorophyll a and b contents below the control experiment while carotenoids content increased from 0.071 to 0.232 mg/g dry biomass with increasing the alginate supplementation from 0 to 0.5 g/L (Fig. 5A). In general, under this mixotrophic nutritional mode, Chlorella vulgaris demonstrated marked increases in most measured growth parameters, whereas, the magnitude of response followed dose response manner. The maximum stimulatory effect was induced by 0.3 g/L alginate hydrolysate supplementation for dry biomass, protein, carbohydrate contents recording values of 0.7 g/L, 179.222 mg/g dry weight and 620.332 mg/g dry biomass; respectively. Higher sodium alginate hydrolysate dose (0.5 g/L) induced nonsignificant increments in the dry biomass, protein and carbohydrate content recording 0.32 g/L, 131.215 mg/g dry weight and 433.143 mg/g dry weight; respectively. On the other hand, the specific growth rate increased in a dose-response manner, taking the bell shape response, achieving the maximum result (0.104 g/L/d) with 0.3 g/L alginate supplementation as illustrated in Fig. 4C. Figure 5C demonstrated the induced changes in the total phenolic content of chlorella vulgaris under this mixotrophic mode of nutrition. Data revealed significant increases in total phenolic contents relative to the control level with increasing alginate hydrolysate concentration (18.315, 20.79, 27.697 and 26.468 mg/g); respectively.   Fig. 5D. Thereby, guaiacol peroxidase exhibited significant increases with increasing alginate hydrolysate concentration, displaying maximum activity with 0.3 g/L alginate hydrolysate treatment (2.899 µ Mol min −1 g −1 ). Shen et al. 57 revealed that Chlorella vulgaris could grow under different nutritional modes, especially the mixotrophic condition which could induce higher biomass than autotrophic and heterotrophic cultures. Furthermore, substrate concentration and external carbon source, induced final biomass of the mixotrophic cultivations while the lipids content remains unchanged. Many previous studies documented that alginate oligomers have growth promoting effect on higher plants, since C. vulgaris is a photosynthetic eukaryotic cell, having structural and functional similarities with higher plant cells so it responds to alginate oligomers in a comparable manner 58 . Yamasaki et al. 59 demonstrated that alginate oligosaccharides may function as growth enhancing agents for certain plant cells and some green algae. In accordance with the present results, Yokose et al. 60 indicated that oligoalginate preparations provided by enzymatic hydrolysis using bacterial alginate lyase, improved growth of Nannochloropsis oculate. Whereas, generally, 20 mg/L oligoalginate preparations induced the optimal growth. On the other hand, they found that 40 mg/L slightly decreased the observed growth enhancing activity. As mannuronic acid and guluronic acid are the main components of alginate, it may produce complex with Ca 2+ and overcome Cu 2+ induced growth suppression of N. oculate 60 . Diatom Chaetoceros gracilis also responded positively to oligoalginate preparation except for diatom genus Skeletonema sp., demonstrating a species-specific growthenhancing significance of the oligo-alginate preparation in microalgae as previously suggested by Yokose et al. 60 . In the present study, alginate hydrolysate concentrations are effective on C. vulgaris growth and metabolism in a concentration dependent manner with a bell shaped profile. It can be observed that control experiment (0 g/L alginate hydrolysate) achieved 0.24 g/L biomass whereas the optimum growth (0.7 g/L) was achieved at 0.3 g/L  Moreover; 1 mg/L alginate oligosaccharides produced via alginate lyase decomposition, promoted Chlamydomonas reinhardtii growth. Moreover; the alginate oligomers produced after acid hydrolysis had no growth-enhancing potentiality as documented by Schafer et al. 9 . Also, increments in C16:0, C18:2, and C18:3 fatty acids content of C. reinhardtii were noticed after treating with enzyme-decomposing oligo-alginates excluding C18:0 level. On the other hand, acid hydrolysis produced oligo-alginates with no influence on fatty acid content as reported by Schafer et al. 9 . Thus, the oligo-alginate mixture induces the growth of microchlorophytes and increases the content of fatty acids, which may have future prospective in the biotechnological practices for biodiesel production.
According to Naeem et al. 62 , the administration of the depolymerized form of irradiated sodium alginate (ISA) as a fertilizing agent resulted in a significant increase in photosynthetic parameters, chlorophyll and carotenoids content as well as growth promotion of Mentha arvensis. They supposed that ISA might enable plants to catch more light energy for increasing photosynthesis and/or improving chlorophyll and carotenoids contents consequently after foliar application of ISA as documented by El-Chaghaby et al. 63 previous study. Moreover, when sodium alginate oligomers functioned as plant growth enhancers, they might stimulate plant vegetative growth leading to increased plant productivity relative to control 64 . Degraded alginate (oligosaccharides) induced cell signaling resulted in enhancing different physiological processes in plants 65 . Chlorophyll is one of the cellular compounds on the basis of which microalgal biomass in the culture is estimated and it can be used to measure cell   65 reported that external organic carbon source supplementation may affect photosynthesis and respiration. Also, oligomers produced by alginate depolymerization stimulates the growth and promote the germination and shoot elongation in plants 63 . In accordance with our data, plant growth induction was achieved by the use of radiated sodium alginate oligosaccharides. Gamma irradiation of sodium alginate influences all polymer cross-linking manner, its use affects the biological activities of plant cells 64 .
Scanning electron microscopy (SEM). Alginate as a smart matrix for riboflavin delivery systems was scanned using SEM. The surface and cross-sectional scanning electron microscopy images of the prepared barium alginate beads are illustrated in Fig. 6A. The dry prepared beads were virtually spherical with a mean diameter of 2.37 mm with rough surface. The cross-sectional SEM images demonstrated many closed pores with varying diameters (mean value 19.24 to 166.4 µm) (Fig. 6B). In accordance with the current results, Malakar et al. 66 reported that SEM photographs of liquid paraffin entrapped calcium alginate bead surface showed a rough surface with small pores or channels and no drug crystals were found on the bead's surface, revealing the dispersion of drug crystals in the alginate matrix. Moreover, Rashidzadeh et al. 67 reported that SEM images of alginate/Ag/Fe 3 O 4 hydrogel beads are illustrated a clear rough and flat surface. According to Hosseini et al. 71 , the characteristic chemical groups of riboflavin are demonstrated in lower intensities and may be obscured by other signals confirming the existence of electrostatic interactions between the blend components. FTIR charts of Arabic gum-entrapped barium alginate beads loaded with riboflavin exhibited comparable features indicating absence of interference between R and Alg-AG ingredients besides non-participation of the ketonic carbonyl in coordination with alginate binding centers 72 .
For Alg-AG-R physical interaction characterization, the R carbonyl stretching region (1730-1620 cm −1 ) was analyzed. The present data illustrated that characteristic acid carbonyl stretching band of the pure drug appeared unchanged in the polymer/drug physical mixtures, and the spectra seemed to be the sum of the spectra of the pure components.
Riboflavin encapsulation efficiency of the prepared riboflavin alginate-Arabic gum beads. Sodium alginate has the ability to form rigid gels with divalent cations. Although it is relatively easy to describe alginates in terms of M and G units, the detailed molecular compositions of alginates in terms of block lengths and block distributions are more difficult to determine. The formation of riboflavin-entrapped barium alginate beads is a simple and significant process for drug delivery. Riboflavin encapsulation efficiency of the prepared riboflavin alginate-Arabic gum beads ranged from 82.81 to 89.15% (Table 1, Fig. 7A) according to the formulation composition of the beads. The highest encapsulation efficiency was positively correlated to the Arabic gum concentration which was observed in Alg-AG-R 0.2 formulation. This pattern of response could be attributed to the partitioning of certain amounts of riboflavin in the Arabic gum phase and/or constitution of an Arabic gum barrier that prevents external passage of riboflavin molecules throughout preparation. Whereas, the physical interaction and/or enlargement of the complicated cross-linked barium alginate network may facilitate the entrapment of the drug (riboflavin) according to the interpretation of Malakar et al. 66 . Results of Azad et al. 73 are in line with the current results whereas, they documented that the EE% of black seed oil in alginate beads ranged from 67.20 to 104.50% and increased with increasing voltage and flow rate. The encapsulation efficiency was found to be dependent on the encapsulating matrix's strength 74 . On the other hand, the cross-linking ability of alginate may be able to increase the encapsulation efficiency of oil in alginate 75 . In vitro release studies. The in vitro prolonged, sustained release profile (Fig. 7B) of Alg-AG-R with different alginate-Arabic Gum composite-riboflavin formulations could be designated as two-step biphasic process. Concerning Alg-AG-R 0.1, the first phase continued for 70 min which designated by burst release, followed by a stationary phase, whereas in Alg-AG-R 0.15 and Alg-AG-R 0.2 formulations the exponential R release con-  Fig. 7C showed different patterns of cumulative drug release (%) profiles of riboflavin with the following hierarchy Alg-AG-R 0.1 (37.598%) > Alg-AG-R 0.15 (28.375%) > Agl-AG-R 0.2 (23.802%). Increments in R release from Alg-AG-R was parallel to increasing alginate content and to decreasing Arabic gum content in the beads on the other side. The current results indicated that riboflavin release rate significantly increased with decreasing Arabic gum incorporated concentration i.e., by the relative increasing alginate concentration. Owing to interpretation of Peppas and Narasimhan 76 , drug molecules dissolution from polymeric blends depends generally on two key phenomena. Firstly, decomposing of the encapsulating substance and dispersion of the drug molecules through the polymeric matrix; however, the second phenomenon concerned with swelling, chemical decomposition, osmotic consequences. Based on Bera et al. 77 interpretations, slow prolonged release could be ultimately due to the residual drug that is dispersed in Arabic gum pockets of the beads formulating a drug-Arabic gum dispersed matrix. Consequently, R delivery from Alg-AG-R beads to the dissociation medium may exhibit two steps, including the diffusion of the drug out of Arabic gum pockets into the barium alginate matrix at first, then it could be distributed outwards barium alginate matrix into the dissociation medium. Another point of view was suggested by Bera et al. 77 who described the behavior of R release to additional Arabic gum barrier formation. www.nature.com/scientificreports/ Beirão-da-Costa et al. 78 documented the importance of alginate volume expansion in the initial exponential phase of R release from Alg-AG matrix. Helrich 79 suggested that drug release from microparticles might be occurred by different modes of action comprising surface erosion, disintegration, diffusion and desorption. According to Anitha et al. 80 , the initial fast release rate could be attributed to riboflavin molecules adsorption onto and near the particle surface where the alginate dissolution rate is high. Hosseini et al. 81 and Siepmann et al. 82 documented an alternative interpretation, that slow R delivery during second stage might be attributed to the continuous diffusion of R into Alg-AG microbeads with time, in addition to the maintenance of almostlinear R concentration gradients over extended periods within the Alg-AG microbeads. Alginate microparticles were considered safe, Abdelaziz et al. 83 reported alginate nanoparticles as a safe delivery system for miltefosine in the treatment of candidiasis and cryptococcosis. Alginate nanoparticles toxicity were assessed on red blood cells and Galleria mellonella larvae. Miltefosine in alginate nanoparticles existed neither hemolytic effect, nor, toxicity in larvae. Results showed the nontoxic use of alginate-based drug-delivery systems as carriers to control the fungal infection in the in vivo model of G. mellonella 19 . Shanmuganathan et al. 84 found that various drugs and nanoparticles have been encapsulated or adhered over chitosan nanoparticles and applied for cancer treatment.
Kinetic studies. The mathematical kinetic models: zero order ( Supplementary Fig. S3A), first order (Supplementary Fig. S3B), Higuchi (Supplementary Fig. S3C) and Korsmeyer-Peppas ( Supplementary Fig. S3D) models were followed for evaluating the in vitro riboflavin release. The zero-order kinetic model is a relation between time (min) and log cumulative percent of drug release, achieving R 2 values ranging from 0.8823 to 0.7914. The first order kinetic model is a relation between time (min) and Ln(1-F), giving R 2 values ranging from 0.777 to 0.7915. Higuchi model is a relation between t 1/2 and log cumulative percent of drug release achieving R 2 results range of 0.9616 to 0.8945. Korsmeyer Peppas model is a relation between Ln t and log cumulative percent of drug release giving R 2 results range of 0.9783 to 0.9484. The resulted kinetic parameters of curve fitting in the previously mentioned mathematical models are listed in Table 1. Investigating data of corresponding correlation coefficients (R 2 ) of Alg-AG-R microbeads in the dissociation medium indicated that riboflavin release follows Higuchi model (R 2 = 0.962-0.887) and Korsmeyer Peppas model (R 2 = 0.948-0.978) over a period of 2 h. Different riboflavin preparations release behavior was assessed using Korsmeyer-Peppas model (Supplementary Fig. S3D) which differentiate between different release mechanisms: Fickian release (controlled release), non-Fickian release (irregular transport), and case-II transport (relaxation-controlled release). Value of n ≤ 0.43 designates Fickian release, n value range between 0.43 and 0.85 indicates non-Fickian release, Whereas, n ≥ 0.85, it is case-II transport which includes polymer disbanding and enlargement of polymeric chain 85 . The present data indicate Fickian release (diffusion-controlled release) (n ranges from 0.204 to 0.389) as reported by Siepmann et al. 82 . In accordance with our results, Azad et al. 73 reports that the black seed oil release from alginate fitted with Korsemeyer-Peppas kinetic model (R 2 = 0.900 to 0.997) over two h period, except for the release exponent (n) values which were in the range of 0.49 to 0.61, designating non-Fickian diffusion in acidic media (pH 1.2). According to the interpretation of Danarto et al. 86 , the ability to encapsulate riboflavin in the alginate-gum Arabic hydrogel could be contributed to the development of an "eggbox" structure during the crosslinking process with Ba 2+ ions. They suggested that higher sodium-alginate concentration will result in more loading of riboflavin. This might be attributed to the number of "egg-box" structures that are formed within the alginate molecule whereas 3% Na-alginate concentration was the optimum for the loading process.

Materials and methods
Collection and preparation of macroalgal sample. Sargassum latifolium (Turner), C. Agardah was collected from the shores of Safaga-Quaser, Red Sea Governorate, Egypt, on June 2017, following the institutional, national and international guidelines and legislation. The macroalga was kindly identified by prof. Dr. Mohamed S. Abdel-Kareem, Botany and Microbiology Department, Faculty of Science, Alexandria University, Egypt. The voucher specimen (Sargassum latifolium-Herb. Nasr-1ph-2021) has been deposited at the herbarium of late professor Abdel-Halim Nasr at Botany and Microbiology Department, Faculty of Science, Alexandria University, Egypt. Sargassum latifolium thalli were surface cleaned using distilled water, then dried at 60 °C until reaching constant weight. Thereafter, thalli were crushed into small pieces (0.1-0.5 cm) 87 .
Extraction of sodium alginate. Alginate extraction was conducted following Bouissil et al. 24 protocol with some modifications. 20 g crushed algal biomass was washed with 300 mL boiling distilled water for 30 min. Then the algal biomass was boiled with 300 mL of 0.5% CaCl 2 solution for 30 min. Afterward, algal residue was obtained and boiled with 300 mL of 0.5% NaCl for 1 h. After filtration the algal sample was boiled with 100 mL of Na 2 CO 3 (5%) for 30 min with intensive stirring. After filtration, sodium alginate was precipitated by dehydration using 80% ethyl alcohol and then dried at 50 °C and crushed before storage. Alginate yield was presented as a weight percentage of algal dry biomass according to Belattmania et al. 22 .
Chemical composition. Total carbohydrates content was estimated according to the method of Farmer et al. 65 whereas soluble protein was determined following Idrees et al. 88 protocol. Lipid content was assessed gravimetrically using chloroform-methanol system following the protocol of Currie and Turvey 89 . Sulfate content was assessed after hydrolysis of alginate with 2 M HCl for 2 h at 100 °C according to Association et al. 90 . Uronic acid content was evaluated following the protocol of Blumenkrantz et al. 91 . An alginate specimen was added to a solution of sodium tetraborate in concentrated H 2 SO 4 , followed by addition of m-hydroxydiphenyl reagent with shaking. The indicative color was developed within 5 min and measured spectrophotometrically at 520 nm.  Thermogravimetric analysis (TGA). Thermogravimetry (TG), is an analysis by which the change in a specimen mass exposed to progressive heating with a constant rate is documented and illustrated vs. temperature, is an efficient protocol for investigating the thermal stability of a compound. TGA of sodium alginate was performed using a TGA-50 SHIMADZU-Japan Thermogravimetric analyzer at Egyptian Atomic Authority. The process was conducted in the temperature range 20-800ºC at a heating rate of 10ºC min −1 in nitrogen atmosphere. Data were expressed as percent of mass loss (Δ mass %) versus temperature.
Growth responses of C. vulgaris to alginate hydrolysate supplementation. Alginate hydrolysate was prepared according to the protocol of Hotchkiss et al. 94 , with some modifications. Sodium alginate solution (1%) was acidified with 2 mL of 0.2 N HCl and autoclaved at 121ºC for 20 min. Then cooled and neutralized using sodium carbonate. Finally, a homologous extract of oligo-guluronic acid was obtained.
Chlorella vulgaris, the test microalga, was provided from the microalgal collection of Phycology Laboratory-Faculty of Science-Mansoura University. C. vulgaris was grown up in axenic cultures and incubated for 14 days at 25 ± 2ºC under constant illumination (59.4 µmol s −1 m −2 ) using BG11 media 93 . C. vulgaris was grown under a mixotrophic mode of nutrition using alginate hydrolysate as an external carbon source in the following concentrations (0.05, 0.1, 0.3 and 0.5 g/L) to investigate C. vulgaris growth and metabolism. Dry biomass was estimated at two days intervals. Moreover, at the end of the experiment, light harvesting pigments, protein as well as carbohydrate contents were estimated as previously mentioned. Photosynthetic pigments (chlorophylls and carotenoids) were assessed following the method of Metzner et al. 95 . Soluble phenols content was assessed according to the method of Gillespie et al. 96 , guaiacol peroxidase (GPX) activity was estimated following the protocol of Curtis et al. 97 .
Sodium alginate as a carrier matrix for riboflavin drug delivery in simulated gastric juice. For preparing Arabic gum-entrapped barium alginate beads loaded with riboflavin (Alg-AG-R), the emulsion-gelation method of Lin et al. 98 was followed. Two grams of sodium alginate were dissolved in 100 mL deionized water with stirring. Liquid Arabic gum (0.1, 0.15 and 0.2 g) and riboflavin (5 mg) were mixed with the previously prepared sodium alginate solutions, then aqueous preparations of sodium alginate and liquid Arabic gum with riboflavin were stirred for 30 min at 150 rpm for emulsion stability. For producing spherical beads, the formed polymeric emulsion was dropped through glass syringe with a size-22 disposable needle into BaCl 2 solution (1 g BaCl 2 in 100 mL of 10% acetic acid) with continuous stirring. For strengthening the beads, they were permitted to stand for 15 min in the BaCl 2 solution, then filtered and washed twice with deionized water. The resulted beads formulations are Alg-AG-R 0.1 (F1), Alg-AG-R 0.15 (F2), Alg-AG-R 0.2 (F3), Alg-AG 0.1 (F4), Alg-AG 0.15 (F5) and Alg-AG 0.2 (F6), illustrated in Table 2.
The simulated gastric fluid was prepared by dissolving 3.5 g glucose, 2.05 g NaCl, 0.60 g KH 2 PO 4 , 0.11 g CaCl 2 , 0.37 KCl in 200 mL deionized water. The pH of the solution was adjusted to 2 using 1 M HCl and to final volume of 1 L.
Scanning electron microscopy (SEM). The morphology and surface of the alginate beads were assessed using scanning electron microscope (JEOL JSM 6510/V, Japan) at the Electron Microscope Unit, Mansoura University-Egypt.

Riboflavin release kinetics.
For interpreting the performance of riboflavin release from the barium alginate beads in simulated gastric fluid (pH 2), suitable mathematical models must be followed. In vitro riboflavin release from different alginate-Arabic gum beads loaded with riboflavin (Alg-AG-R) data were assessed kinetically through different mathematical models: (Zero-order, First-order, Higuchi and Korsmeyer-Peppas model) 100 as shown in Table 3; where F represents the drug fraction released at time t, K 0 represents zero-order release constant, K 1 represents the first-order release constant, K H represents Higuchi dissolution constant and K p is Korsmeyer-Peppas constant and n represents the release exponent.
Statistical analysis. Data were subjected to statistical analysis following Zobel et al. 101 using one-way analysis of variance followed by least significant difference (LSD) analysis, P-values more than 0.05 were considered statistically, non-significant, whereas P-value less than 0.05 represent statistically significant data. Results were expressed as mean ± standard deviation.

Conclusion
In this study, Sargassum latifolium-derived alginate hydrolysate can be used as C. vulgaris growth bio-stimulant.
A new sustained release system of riboflavin-entrapped barium alginate-Arabic gum drug delivery matrix was elucidated. Moreover, drug entrapment technique using alginate is a valuable developmental tool of the multi particulate system for drug delivery even for a highly water-soluble drug such as riboflavin. (1) EE(%) = total amount of loaded R/initial amount of R × 100 Table 2. Various Alginate-Arabic gum formulations with and without riboflavin.