*4.1. Preparation of Fucoidans*

Three different brown seaweeds, *Ascophyllum nodosum* (collected from the Irish Sea), *Laminaria Japonica* (collected from the South China Sea), and *Kjellmaniella crassifolia* (collected from the South China Sea), were purchased from Kunshan Yihong Seaweed Co. Ltd. (Kunshan, Jiangsu, China). The extraction and purification processes of AnF, LjF, and KcF were carried out according to the method, as previously described [30,56]. Briefly, brown algae were dried, powdered, and then pass through a 12-mesh net, followed by delipidating (60 ◦C, 4 h, 95% ethanol, 1:20) and two cycles of extraction with double-distilled water (ddH2O) (80 ◦C, 3 h, 1:20). After centrifugation at 5000 rpm for 10 min, the supernatants were combined and concentrated, and then anhydrous ethanol was added to achieve a final concentration of 80%, and the supernatants were left overnight at 4 ◦C. After centrifugation at 5000 rpm for 10 min, the crude fucoidans were obtained. Then, the precipitate was redissolved in ddH2O, and 3 M of CaCl2 solution was added to remove alginate completely until no precipitation occurred. The solution was centrifuged at 8000 rpm for 10 min to remove precipitates, and the supernatant containing fucoidan was dialyzed in a dialysis bag with a 7000 Da MW cutoff, then lyophilized to obtain the refined fucoidans. Chemicals reagents were obtained from the Sigma Chemical Co. (Sigma–Aldrich, St. Louis, MO, USA) unless otherwise stated.

### *4.2. Physicochemical Properties Analysis of Fucoidans*

The sulfate content was determined by the BaCl2-gelatin method as follows [57]. Fucoidan (3 mg/mL) was degraded in 1 M of HCl at 110 ◦C for 6 h, then the absorbance of the degraded solution was determined at 400 nm after mixing with an equal volume of BaCl2-gelatin. The sulfate content was calculated based on a standard curve, which was established with a Na2SO4 standard. The monosaccharide composition was analyzed with the acid hydrolysis method described previously [58]. In brief, the monosaccharide composition was determined using a1-phenyl-3-methyl-5-pyrazolone precolumn derivatization HPLC on an Agilent Eclipse XDB-C18 Column (Agilent, Santa Clara, CA, USA). Sample (10 μL) was eluted with 0.1 mol/L phosphate buffer (pH 6.7) and acetonitrile (83:17 volume fraction) at a flow rate of 1 mL/min at 30 ◦C. Then, a UV detector was used to detect the signal at 245 nm. MW was determined using high-performance gel permeation chromatography coupled with a multi-angle laser light scattering instrument. The MWs of fucoidans were determined using an Agilent 1260 HPLC system (Agilent Technologies, CA, USA) on the Shodex Ohpak SB-HQ 804 column in series with an SB-HQ 803 column (TosoHaas Corp., Tosoh, Japan) detected with a Wyatt Dawn Heleos II multi-angle laser scattering system (Wyatt Technology, Santa Barbara, CA, USA) and refractive index detector. One hundred microliters of sample was eluted with 0.1 M of Na2SO4 solution at a flow rate of 0.6 mL/min at 35 ◦C. MWs were calculated using Astra 5.3.4.20 software (Wyatt Technology, Santa Barbara, CA, USA). The type of glycosidic linkages of fucoidans was determined by the published papers, i.e., FvF (31), AnF (32), KcF (33), and LjF (34).

#### *4.3. E*ff*ects of Fucoidans on Glucose Absorption Using Caco-2 Monolayer Model*

The inhibitory effects of fucoidans on glucose uptake in vitro were measured using a human colon cancer cell line monolayer model (Caco-2 cells, from the Cell Bank of the Chinese Academy of Sciences, Shanghai, China) incubated with 2-NBDG (MedChem Express, Monmouth, NJ, USA) in various conditions. Caco-2 cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) with 10% FBS, 1% penicillin/streptomycin, 25 mM HEPES, and 0.35 g/<sup>L</sup> sodium bicarbonate (Gibco, Carlsbad, CA, USA) at 37 ◦C with 5% CO2. A Caco-2 monolayer model was established as follows [59]. Briefly, Caco-2 cells were adjusted to 2 × 10<sup>5</sup> cells/mL, and 100 μL of this cell suspension was inoculated in the upper layer of a transwell compartment (0.4 μm, 1.12 cm2, PET) (Corning Inc., Corning, NY, USA) and incubated at 37 ◦C for 2 min. Then, 500 μL of DMEM was added to the upper layer, while 1.5 mL DMEM was added to the lower layer for Caco-2 cells to differentiate into enterocyte-like cells at 37 ◦C. Then, the TEER was measured, which could evaluate the integrity of the Caco-2 monolayer cells. In addition, another index to evaluate the successful construction of the Caco-2 monolayer model was to determine the ALP activity ratio of the apical side to the basolateral side. In the Caco-2 monolayer cell model, the apical side located to the upper side of the transwell had higher ALP activity, while the basolateral side located to the lower side had lower ALP activity. The enzyme activities of both sides of the transwell were measured using the ALP ELISA kit (Shanghai Enzyme-linked Biotechnology Co. Ltd., Shanghai, China) according to the manufacturer's recommended protocol on the 4th, 8th, 12th, 16th, and 21st-day post-induction, respectively. Then, the ALP activity ratio was calculated. The 2-NBDG uptake in the Caco-2 monolayer model was conducted as previously described [60]. Briefly, the culture medium was removed from each well and replaced with 100 μL of HBSS buffer in the presence of 2-NBDG (100 μM) or 2-NBDG (100 μM) together with 400 μg/mL of specific fucoidan. Then the cells were incubated at 37 ◦C for 30 min. Finally, the fluorescence intensity (Ex/Em = 485/535 nm) in the lower layer was measured using a Spark 10M (Tecan Trading AG, Männedorf, Switzerland).

#### *4.4. E*ff*ects of Fucoidans on OGTT in Kunming Mice and Glucose Transport Using Everted Gut Sac Model*

An everted gu<sup>t</sup> sac model was established as previously described [40]. Briefly, six-week-old male Kunming mice were purchased from the Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China). The mice were raised in ventilated cages, maintained in a light-dark cycle of 12 h at 23 ◦C–25 ◦C, with free access to water and food. After a two-week adaptive period, the mice were randomly divided into four groups of six each. OGTT was performed as follows [61,62]. The mice of each group were fasted for 15 h, then the experimental groups were given the specific fucoidan by gavage at a dose of 200 mg/kg, while the Control group was given the same volume of saline. Then, mice were given a 20% glucose solution at a dose of 2 g/kg by oral gavage in 15 min. Next, blood glucose levels were detected using a standard glucometer (Johnson & Johnson, New Brunswick, NJ, USA) at 0, 30, 60, 90, and 120 min by cutting the tail tip. In addition, the increment of plasma glucose following glucose loading was expressed in terms of the area under curve, using the trapezoidal rule. The jejunum was separated from the Kunming mice one week after the OGTT experiment. Mice were euthanized by pentobarbital sodium injection (80 mg/kg) accompanied by isoflurane inhalation to maintain anesthesia. The jejunum was cut into 5-cm segments and quickly transferred into cold Krebs–Ringer bu ffer in the state of oxygen maintenance. Due to the overturning of the intestinal sac, SGLT1 protein originally on the serosal side moved to the outside, so the intestinal epithelial cells could absorb glucose from the outside to the inside in. Thus, Krebs–Ringer bu ffer was injected into the intestinal capsule of the everted gu<sup>t</sup> sac model, then placed in Krebs–Ringer bu ffer (containing 30 mM D-glucose) with various concentrations of specific fucoidan in the experimental groups, while the group treated with Krebs–Ringer bu ffer (containing 30 mM D-glucose) was used as the Control. After incubation for 30 min at 37 ◦C, the glucose concentrations of the inside and outside of the intestinal capsule were determined using the glucose oxidase-peroxidase method with a glucose oxidase kit (Applygen Technologies Inc., Beijing, China) [63]. Additionally, the glucose intake ratio was calculated by the glucose concentration of the inside, divided by the glucose concentration of the outside. All animal procedures were approved by the Committee of Experimental Animals of School of Medicine and Pharmacy, Ocean University of China (OUCSMP-18081201), and conformed to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health (NIH Publication No 85-23, revised 1996).

#### *4.5. E*ff*ects of AnF on Alleviating Hyperglycemia in db*/*db Mice*

Briefly, eight-week-old male db/db mice were provided by the Model Animal Research Center of Nanjing University (Nanjing, China). In addition, eight-week-old male C57BL/6J mice were purchased from the Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China) as a Control group. The mice were raised as described in the methods in Section 4.4. After a two-week acclimation period, the db/db mice were randomly divided into three groups as follows, with six mice in each group: the Metf and AnF groups received either metformin (200 mg/kg/d dissolved in saline) or AnF (200 mg/kg/d dissolved in saline) by gavage for four weeks, while the Model group was given an equal amount of saline. Body weights were measured every week. OGTT was conducted at the end of the trial as follows: mice were fasted for 15 h, and the fasting blood glucose levels were detected. Then, the mice were given a 20% glucose solution by gavage at a dose of 2 g/kg body weight. The changes in blood glucose levels were detected, as described in Section 4.4. All animal procedures were approved by the Committee of Experimental Animals of School of Medicine and Pharmacy, Ocean University of China (OUCSMP-18081201), and conformed to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health (NIH Publication No 85-23, revised 1996).

#### *4.6. E*ff*ects of AnF on Biochemical Indexes in db*/*db Mice*

Db/db mice with various treatments were finally euthanized by pentobarbital sodium injection (80 mg/kg) accompanied by isoflurane inhalation to maintain anesthesia after being fasted for 15 h. Blood samples were collected via retro-orbital bleeding, then centrifuged at 2000× *g* for 15 min to obtain serum for serological assays. The levels of fasting insulin and HbA1c in serum were determined using a mouse insulin ELISA kit and a mouse HbA1c ELISA kit from Omnimabs (Alhambra, CA, USA) according to the manufacturer's instructions, respectively. HOMA-IR was calculated as fasting insulin (mU/L) × fasting glucose (mM)/22.5. For tGLP-1 and aGLP-1 contents detection, a DPP-4 inhibitor was quickly added into the blood samples and mixed evenly. Then, the blood was centrifuged for 10 min at 2000× *g* to obtain the supernatant for assays. The contents of tGLP-1 and aGLP-1 were determined using mouse ELISA kits (Linco, St. Charles, MO, USA) according to the manufacturer's instructions, respectively.

#### *4.7. Binding Kinetics Analysis of Interaction between SGLT1 and Fucoidans*

The binding kinetics between various fucoidans and SGLT1 protein was determined by an SPR biomacromolecule interaction analyzer BIAcore T200 (General Electric Company, Boston, MA, USA), as previously described [64,65]. After washing the surface of the CM5 chip (General Electric Company, Boston, MA, USA) with PBS-P running bu ffer (General Electric Company, Boston, MA, USA), the surface

of the chip was activated with 0.4 M EDC/0.1 M NHS for 420 s at a flow rate of 10 μL/min. Immediately after activation, an SGLT1 solution (20 μg/mL) (ab152683, Abcam, Cambridge, UK) in sodium acetate buffer (pH 4.5) was added onto the chip surface for 30 s at a flow rate of 10 μL/min. After that, the chip surface was sealed by incubating with 1 M of ethanolamine (pH = 8.5) for 30 min. PBS-P buffer was running for at least 2 h to stabilize the baseline. To assess the real-time binding of fucoidans to SGLT1, varying concentrations of specific fucoidan were injected over the sensor chip surface at a flow-rate of 30 μL/min for 120 s, followed by another 900 s dissociation period. The sensor surface was regenerated by 0.1 mM NaOH for 10 s. The response was monitored as a function of time (sensor gram) at 25 ◦C and subtracted from the response of the reference surface. The binding constant (*K*a), dissociation constant (*K*d), and average dissociation constant (*K*D = *K*d/*K*a) of the interaction between various fucoidans and SGLT1 protein could be calculated from curve fitting. Kinetic parameters were evaluated using the BIAcore T200 evaluation software 3.1 (General Electric Company, Boston, MA, USA).
