Synthesis and Evaluation of Hypoglycemic Activity of Structural Isomers of ((Benzyloxy)phenyl)propanoic Acid Bearing an Aminobornyl Moiety

Free fatty acid receptor-1 (FFAR1) agonists are promising candidates for therapy of type 2 diabetes because of their ability to normalize blood sugar levels during hyperglycemia without the risk of hypoglycemia. Previously, we synthesized compound QS-528, a FFA1 receptor agonist with a hypoglycemic effect in C57BL/6NCrl mice. In the present work, structural analogs of QS-528 based on (hydroxyphenyl)propanoic acid bearing a bornyl fragment in its structure were synthesized. The seven novel compounds synthesized were structural isomers of compound QS-528, varying the positions of the substituents in the aromatic fragments as well as the configuration of the asymmetric center in the bornyl moiety. The studied compounds were shown to have the ability to activate FFAR1 at a concentration of 10 μM. The cytotoxicity of the compounds as well as their effect on glucose uptake in HepG2 cells were studied. The synthesized compounds were found to increase glucose uptake by cells and have no cytotoxic effect. Two compounds, based on the meta-substituted phenylpropanoic acid, 3-(3-(4-(((1R,2R,4R)-1,7,7-trimethylbicyclo-[2.2.1]heptan-2-ylamino)methyl)benzyloxy)phenyl)propanoic acid and 3-(3-(3-(((1R,2R,4R)-1,7,7-trimethylbicyclo [2.2.1]heptan-2-ylamino)methyl)benzyloxy)phenyl)propanoic acid, were shown to have a pronounced hypoglycemic effect in the oral glucose tolerance test with CD-1 mice.


Introduction
Type 2 diabetes mellitus (T2DM) is a major public health problem worldwide. The global diabetes prevalence in 2019 was estimated to be 463 million people worldwide, with 90% of cases associated with type 2 diabetes [1]. More than one million deaths each year can be associated with T2DM, making it the ninth leading cause of death worldwide. [2]. The development of T2DM is considered to be caused by insulin resistance, which is triggered by a combination of two main factors: defective insulin secretion by pancreatic β-cells and the inability of insulin-sensitive tissues to respond to insulin [3,4]. Insulin resistance leads to a chronic condition of elevated blood glucose levels, i.e., hyperglycemia, which, if ignored, can lead to serious complications such as nephropathy [5], retinopathy [6], and cardiovascular system disorders [7].
Nowadays, T2DM pharmacotherapy is symptomatic and consists of normalizing glucose levels. There are several classes of hypoglycemic drugs that act on different biological targets and are able to exert a hypoglycemic effect. Therapy selection is individualized for each patient, which makes the search for new classes of hypoglycemic agents an important task [8]. New hypoglycemic agents are being scouted both for known targets, such as dipeptidyl peptidase-4 (DPP4) inhibitors [9,10], thiazolidinediones [11,12], α-glycosidase inhibitors [13][14][15], and sodium-glucose linked transporter-2 (SGLT-2) inhibitors [16,17], as The aim of this work was to synthesize structural analogs of compound QS-528 by varying the relative positions of structural fragments in the molecule, including the synthesis of structural analogs containing a meta-substituted fragment of phenylpropanoic acid, for further exploring the structure-activity relationship in in vitro and in vivo. In this work, we evaluated the influence of target compounds on cellular glucose uptake and their hypoglycemic effect in the oral glucose tolerance test in mice. The aim of this work was to synthesize structural analogs of compound QS-528 by varying the relative positions of structural fragments in the molecule, including the synthesis of structural analogs containing a meta-substituted fragment of phenylpropanoic acid, for further exploring the structure-activity relationship in in vitro and in vivo. In this work, we evaluated the influence of target compounds on cellular glucose uptake and their hypoglycemic effect in the oral glucose tolerance test in mice.

Chemistry
The proposed scheme for the synthesis of the target compounds was as follows (Scheme 1). The Mitsunobu reaction was used to conjugate alcohols 1a,b with phenols 2a,b [29]. The reaction of alcohols 1a,b with a small excess of phenols 2a,b, diisopropyl azodicarboxylate (DIAD), and PPh 3 was carried out in tetrahydrofuran (THF) for 48 h, and the products 3a-d were isolated by column chromatography. The yields of the obtained compounds 3a-d (scaffolds with a protected aldehyde group) were in the range of 67-76%. The dimethylacetal protection was removed just before the next step by treating substances 3a-d in THF with dilute sulfuric acid for a few hours at room temperature, which resulted in the formation of aldehydes 4a-d in quantitative yield. Reductive amination of aldehydes 4a-d and exo-and endobornylamines was carried out in methylene chloride in the presence of an equimolar amount of acetic acid and a small excess of sodium triacetoxyborhydride. Compounds 5a-h were isolated with yields of 67-81% after column chromatography.
Alkaline hydrolysis of compounds 5a-h with lithium hydroxide followed by treatment of the reaction mixture with 2M hydrochloric acid allowed us to isolate the target compounds 6a-h as hydrochlorides in yields of 48-67%.

In Vitro Study
Compounds 6a-h were tested for FFAR1 activation using the Cayman FFAR1 Assay Kit at a single concentration (10 µM). GW9508 was used as a positive control at the same concentration; its efficiency was taken as 100%. All the investigated compounds showed the ability to activate the receptor at a concentration of 10 µM. The efficacy of the compounds ranged from 76% to 120%.
The effects of compounds 6a-h on glucose consumption and lactate release as well Reductive amination of aldehydes 4a-d and exo-and endobornylamines was carried out in methylene chloride in the presence of an equimolar amount of acetic acid and a small excess of sodium triacetoxyborhydride. Compounds 5a-h were isolated with yields of 67-81% after column chromatography.
Alkaline hydrolysis of compounds 5a-h with lithium hydroxide followed by treatment of the reaction mixture with 2M hydrochloric acid allowed us to isolate the target compounds 6a-h as hydrochlorides in yields of 48-67%.

In Vitro Study
Compounds 6a-h were tested for FFAR1 activation using the Cayman FFAR1 Assay Kit at a single concentration (10 µM). GW9508 was used as a positive control at the same concentration; its efficiency was taken as 100%. All the investigated compounds showed the ability to activate the receptor at a concentration of 10 µM. The efficacy of the compounds ranged from 76% to 120%.
The effects of compounds 6a-h on glucose consumption and lactate release as well as their cytotoxicity was investigated in HepG2 (human liver cancer) cells at concentrations of 2.5, 5, 10, and 25 µM. All compounds were found to have no significant effect on cell viability over the whole range of concentrations investigated.
All compounds tested at concentrations of 2.5 and 5 µM showed no effect on glucose consumption and lactate release. Increasing the concentration of compounds to 10 and 25 µM (Table 1) led to increases in glucose consumption and lactate release to 30-40% and 18-37%, respectively (except 6e and 6f at a concentration of 10 µM). Additionally, we used known FFAR1 agonist GW9508 at concentrations of 10 and 25 µM, which induced significant increases in glucose consumption to 44-54% and lactate release to 34-46%.

In Vivo Study
To evaluate the hypoglycemic effect of synthesized compounds 6a-h, the OGTT was conducted with CD-1 mice at doses of 5 and 15 mg/kg. Vildagliptin (Galvus ® ) was used as a positive control at a dose of 10 mg/kg. Table 2 shows the differences between the AUC (area under the curve) values of the experimental and control groups, presented in percentage. The data obtained indicated that compound 6e at the studied doses induced a hypoglycemic effect in a dose-dependent manner, since increasing the dose led to a stronger effect ( Figure 2). Compound 6g showed a hypoglycemic effect only at a dose of 15 mg/kg ( Figure 3). --

In Vivo Study
To evaluate the hypoglycemic effect of synthesized compounds 6a-h, the OGTT was conducted with CD-1 mice at doses of 5 and 15 mg/kg. Vildagliptin (Galvus ® ) was used as a positive control at a dose of 10 mg/kg. Table 2 shows the differences between the AUC (area under the curve) values of the experimental and control groups, presented in percentage. The data obtained indicated that compound 6e at the studied doses induced a hypoglycemic effect in a dose-dependent manner, since increasing the dose led to a stronger effect (Figure 2). Compound 6g showed a hypoglycemic effect only at a dose of 15 mg/kg ( Figure 3).

In Vivo Study
To evaluate the hypoglycemic effect of synthesized compounds 6a-h, the OGTT was conducted with CD-1 mice at doses of 5 and 15 mg/kg. Vildagliptin (Galvus ® ) was used as a positive control at a dose of 10 mg/kg. Table 2 shows the differences between the AUC (area under the curve) values of the experimental and control groups, presented in percentage. The data obtained indicated that compound 6e at the studied doses induced a hypoglycemic effect in a dose-dependent manner, since increasing the dose led to a stronger effect (Figure 2). Compound 6g showed a hypoglycemic effect only at a dose of 15 mg/kg ( Figure 3).

Discussion
The structures of FFAR1 agonists, as can be seen in Figure 1, are linear molecules in which the acid head (phenylpropanoic acid or its analog) is linked through an oxygen atom to the tail part of the molecule. Generally, synthesis of this type of structure involves parallel synthesis of the ester (hydroxyphenyl)propanoic acid and the tail part, and their conjugation reaction is followed by hydrolysis of the ester group, which leads to the formation of the target compounds. Previously, we synthesized compound QS-528 containing a fragment of (4-hydroxyphenyl)propanoic acid and a bornyl fragment [24]. QS-528 was shown to be an effective FFAR1 receptor agonist and induced a hypoglycemic effect in mice at doses of 1-25 mg/kg. The key difference between QS-528 and the other synthetic agonists shown in Figure 1 is the absence of substituents in the aliphatic part of the phenylpropanoic acid. The purpose of the substituent in the aliphatic part of phenylpropanoic acid is to reduce β-oxidation and increase selectivity toward the target receptor [30].
Based on the above, we considered that such phenylpropanoic acid derivatives as QS-528 may be promising antidiabetic agents with potential multitarget action. This assumption was supported by our consequent studies.
Compound QS-528 was shown to induce a hypoglycemic effect in C57BL/6 mice but not in C57BL/6 Ay mice with developed diabetes, although some benefits were exhibited when administered for a prolonged time. QS-528 exhibited hepatoprotective properties by restoring liver functions. In comparison, compound QS-619 ( Figure 2), which is an Oendobornyl analog of compound QS-528, induced a hypoglycemic effect in both C57BL/6 and C57BL/6 Ay mice but had no hepatoprotective effect [25].
In order to further evaluate the effect of the molecular shape on its anti-diabetic properties, we decided to synthesize all eight possible combinations for this structure by varying the bond from the para to the meta position in both aromatic rings as well as changing the configuration of the asymmetric center in the bornyl moiety.
Ortho-substituted compounds were discarded due to the lack of data about known successful examples as well as the available information about the lack of effectiveness of such compounds [30], which may be related to the strong conformation deformation as compared to the endogenous ligands.
Based on the literature analysis, we assumed that the meta-substituted derivatives were likely to exhibit better activity by being full agonists [40] or positive allosteric modulators [27].
In previous work [24], we used 1,4-bis(bromomethyl)benzene 7 as a building block for the synthesis of the target compound QS-528, but in scaling up we faced problems in needing to use a considerable amount of dibromide as well as difficulties in its separation from the reaction product (Scheme 2). Based on the literature analysis, we assumed that the meta-substituted derivatives were likely to exhibit better activity by being full agonists [40] or positive allosteric modulators [27].
In previous work [24], we used 1,4-bis(bromomethyl)benzene 7 as a building block for the synthesis of the target compound QS-528, but in scaling up we faced problems in needing to use a considerable amount of dibromide as well as difficulties in its separation from the reaction product (Scheme 2). Scheme 2. The previously described synthesis of the compound.
To avoid these problems, we suggested using another building block, ((dimethoxymethyl)phenyl)methanol 1, which could be obtained from commercially available bromoaldehydes (Scheme 3). Scheme 3. Synthetic approach to the key building block.
The above-mentioned meta-or para-substituted building blocks 1a-b were reacted with methyl esters of (hydroxyphenyl)propanoic acids 2a-b under standard Mitsunobu conditions, and the products 3a-d were isolated by column chromatography.
It is worth noting that in order to prevent the deprotection of the aldehyde group when working with all compounds containing the dimethyl acetal fragment discussed in this paper, their isolation from reaction mixtures and following purification via chromatography were performed with solvents containing 1% v/v of triethylamine.
The removal of dimethyl acetal protection from compounds 3a-d occurred in quantitative yield when the substance solution in THF was treated with diluted aqueous sul-Scheme 2. The previously described synthesis of the compound.
To avoid these problems, we suggested using another building block, ((dimethoxymethyl)phenyl)methanol 1, which could be obtained from commercially available bromoaldehydes (Scheme 3). Based on the literature analysis, we assumed that the meta-substituted derivatives were likely to exhibit better activity by being full agonists [40] or positive allosteric modulators [27].
In previous work [24], we used 1,4-bis(bromomethyl)benzene 7 as a building block for the synthesis of the target compound QS-528, but in scaling up we faced problems in needing to use a considerable amount of dibromide as well as difficulties in its separation from the reaction product (Scheme 2). Scheme 2. The previously described synthesis of the compound.
To avoid these problems, we suggested using another building block, ((dimethoxymethyl)phenyl)methanol 1, which could be obtained from commercially available bromoaldehydes (Scheme 3). Scheme 3. Synthetic approach to the key building block.
The above-mentioned meta-or para-substituted building blocks 1a-b were reacted with methyl esters of (hydroxyphenyl)propanoic acids 2a-b under standard Mitsunobu conditions, and the products 3a-d were isolated by column chromatography.
It is worth noting that in order to prevent the deprotection of the aldehyde group when working with all compounds containing the dimethyl acetal fragment discussed in this paper, their isolation from reaction mixtures and following purification via chromatography were performed with solvents containing 1% v/v of triethylamine.
The removal of dimethyl acetal protection from compounds 3a-d occurred in quantitative yield when the substance solution in THF was treated with diluted aqueous sul-Scheme 3. Synthetic approach to the key building block.
The above-mentioned meta-or para-substituted building blocks 1a-b were reacted with methyl esters of (hydroxyphenyl)propanoic acids 2a-b under standard Mitsunobu conditions, and the products 3a-d were isolated by column chromatography.
It is worth noting that in order to prevent the deprotection of the aldehyde group when working with all compounds containing the dimethyl acetal fragment discussed in this paper, their isolation from reaction mixtures and following purification via chromatography were performed with solvents containing 1% v/v of triethylamine.
The removal of dimethyl acetal protection from compounds 3a-d occurred in quantitative yield when the substance solution in THF was treated with diluted aqueous sulfuric acid solution. When trying to purify aldehydes 4a-d obtained by column chromatography, we noticed a partial decrease in the yield, so the optimal solution was to use the aldehydes in the next stage without additional purification.
The reductive amination using sodium triacetoxyborhydride proceeded smoothly for all combinations of aldehydes and amines. The compounds were purified by column chromatography on silica gel to afford compounds 5a-h.
The use of lithium hydroxide allowed us to perform hydrolysis of the ester groups in compounds 5a-h under mild conditions. Follow-up treatment of the reaction mixture with dilute hydrochloric acid led to precipitation of the target compounds as hydrochloride salts 6a-h.
It can be noted that the proposed approach to the synthesis of the target substances was versatile; the yields of compounds did not strongly depend on the structure of the initial substrates. Despite the increased number of stages compared to our previous approach [24], we found the new approach to be more suitable for the production of larger amounts of compounds.
Thus, seven new structural analogs of QS-528 were synthesized and studied using in vitro tests and the OGTT.
Glucose consumption and lactate release tests in HepG2 cells allowed us to estimate the effect of the compounds on glucose metabolism in liver cells, which can be used to further examine the mechanism of action [41,42]. The effects of compounds 6a-h at concentrations of 2.5, 5, 10, and 25 µM on glucose uptake and lactate release in human liver cancer HepG2 cells were studied (Table 1). The cytotoxicity of these compounds on this cell line was also estimated.
At low concentrations (2.5 and 5 µM), the studied compounds had no effect on glucose uptake and lactate release, but as the concentration increased to 10 and 25 µM, there was a dose-dependent increase in both glucose uptake and lactate release. It can be noted that the exobornyl derivatives (6a,c,e,g) exhibited slightly higher activity compared to their endo (6b,d,f,h) counterparts (except for the pair of compounds 6g and 6h at 10 µM). The highest difference compared to the control group was found for compound 6a (QS-528).
An interesting effect was observed for compounds 6c,h at a concentration of 10 µM, with a noticeable increase in glucose uptake but little difference in lactate release compared to the control group, which may be explained by the accumulation of glucose in cells.
There was no significant difference between para and meta substitution in any of the two aromatic fragments. With the exception of compound 6a (QS-528), compounds 6b-h appeared to have similar effects, with no significant difference in their effects on cellular glucose uptake and lactate release.
Analyzing the FFAR1 activation data (Figure 2), we observed that all the compounds studied were agonists of the FFA1 receptor at a concentration of 10 µM. The efficacy of compounds 6a-h ranged from 76% to 120%. Compounds 6a,b,d,g showed greater efficacy (104-121%) whereas compounds 6c,e,f,h showed slightly less efficacy (76-84%). It seemed that the derivatives containing a para-substituted phenylpropanoic acid fragment tended to exhibit greater efficacy than their meta-substituted counterparts (except for compounds 6c and 6g).
Then we examined the hypoglycemic effect of the compounds in CD-1 mice ( Table 2). The doses were chosen based on previous work [24]. In mice, derivative 6e containing a meta-substituted fragment of phenylpropanoic acid, a para-substituted linker, and an exobornyl fragment showed the clearest effect in the OGTT, showing a dose-dependent hypoglycemic effect. Its analogue 6g containing a meta-substituted fragment of phenylpropanoic acid, a meta-substituted linker, and an exobornyl fragment showed an effect only at the highest dose of 15 mg/kg studied. The other compounds did not demonstrate a hypoglycemic effect in the OGTT. The missing hypoglycemic effect induced by compound 6a (QS-528) in the OGTT may have been due to the use of a different mouse line compared with previously published results [24]. Despite the absence of a clear correlation between the data on glucose uptake, lactate release, FFAR1 activation, and OGTT, we can make some assumptions about the relationship between the structure and activity of the compounds studied. Based on our results, we can suggest that the exobornyl fragment is a more privileged structure than the endobornyl fragment. The lack of hypoglycemic effect of compounds 6f,h, which are endobornyl analogs of compounds 6e,g, can be explained by the lower metabolic stability of these compounds due to the more reactive amino group, thus causing their higher steric accessibility.
Apparently, the discrepancy in the results of the OGTT, FFAR1 activation, and glucose consumption can be attributed to the differing pharmacokinetics of the studied compounds. Thus, the synthesized compounds demonstrated interesting properties in terms of the search for new antidiabetic agents, which indicates the need for further in-depth research.

General Procedure for Dimethylacetal Deprotection
To a solution of dimethyl acetal 3a-d (0.98 g, 2.9 mmol) in tetrahydrofuran (10 mL), a solution of 3% sulfuric acid in water (4 mL) was added and the resulting solution was stirred at room temperature for 3 h. Ethyl acetate (20 mL) and water (20 mL) were added and the organic layer was separated. The aqueous layer was extracted with ethyl acetate (2 × 20 mL). The combined organic extracts were washed with brine (10 mL) and dried over anhydrous sodium sulfate overnight. The drying agent was filtered off and washed with a small amount of ethyl acetate. The combined solutions were evaporated under reduced pressure and the residue was used in the next step without further purification.

General Procedure for Reductive Amination
To an ice-cooled solution of aldehyde 4a-d (1 equiv), exo-or endobornylamine (1.15 equiv), and acetic acid (1 equiv) in dichloromethane (5 mL per 1 mmol of aldehyde), sodium triacetoxyborohydride (1.5 equiv) was added portionwise in 5 min. Then, after 30 min, the cooling was removed and the resulting solution was stirred overnight at room temperature. Dichloromethane (20 mL) and 5% potassium carbonate (10 mL) solution were added. The organic layer was separated and the aqueous layer was extracted with dichloromethane (2 × 10 mL). The combined organic layers were washed with 5% potassium carbonate (10 mL) solution, water (10 mL), and brine (10 mL) and dried over anhydrous sodium sulfate. The drying agent was filtered off and washed with a small amount of ethyl acetate. The combined solutions were evaporated under reduced pressure and the residue was purified over silica gel column chromatography using chloroform-ethanol (100:1).

General Procedure for Hydrolysis
To an ice-cooled stirred solution of methyl ester (0.195 mmol) in THF (3 mL) and methanol (1 mL), water (2 mL) was added followed by the addition of lithium hydroxide monohydrate (0.390 mmol). After completion of the reaction (monitored by TLC), the THF was evaporated under reduced pressure. Water (10 mL) was added and then 2M hydrochloric acid was added dropwise until pH 2-3 was reached. The resulting solution was left overnight in a refrigerator. The precipitate was filtered, washed with water, and dried in a vacuum desiccator over phosphorus pentoxide.

The Design of the Experiment on HepG2 Cells
Cells were seeded into a 96-well plates (TPP, Trasadingen, Switzerland) at a density of 3 × 10 5 cells/mL in DMEM High Glucose containing 10% FBS. After culturing for 24 h, the cells were washed twice with HBSS (Gibco, Paisley, UK) and the medium was replaced with serum-free low-glucose (5.5 mM glucose) DMEM (Servicebio, Wuhan, China) with compounds 6a-h at various concentrations (2.5, 5, 10, 25 µM) and GW9508 (10, 25 µM) (CAS 1115-70-4 Cayman Chemical Company, Michigan, USA). The compounds were predissolved in DMSO. The final concentration of DMSO in the cell medium was 0.1% (v/v). There were three to four replicate wells for each treatment. After 24 h, the supernatant was collected for the glucose consumption and lactate release assays, while cell viability was measured using the MTT assay.

Glucose Consumption and Lactate Release Assays
Glucose levels were assayed using a commercial kit (Vector-Best, Novosibirsk, Russia) based on the glucose oxidase method and the Multiscan Ascent photometer (Thermo Labsystems, Helsinki, Finland). Glucose consumption was calculated by subtracting the glucose concentration in the supernatant of the blank well (DMEM without cells) from the glucose concentration in the supernatant of the well containing cells. Data were presented as the percentage of maximum glucose consumption (5.5 mM) and normalized to the control (treated with vehicle) for each plate. Meanwhile, the concentration of lactate in the supernatant was also determined using a commercial kit (Vector-Best, Novosibirsk, Russia). Lactate release data were presented as percentage of the maximum value (11 mM) and normalized to the control.

MTT Assay for Cell Viability
A solution of MTT reagent (Thiazolyl Blue Tetrazolium Bromide; Panreac AppliChem, Darmstadt, Germany) (5 mg/mL) was added to the medium in a 1:10 volume ratio, and the mixture was incubated for 4 h at 37 • C in a CO 2 incubator. Then, the supernatant was removed and the precipitate was dissolved with DMSO (Reagent Component, Moscow, Russia) for determination of the optical density at 570 nm (630 nm was used as the reference). Cell viability was calculated as percentage of the control group.

In Vitro FFA1 Activation Assay
For determination of the FFAR1 activation of compounds 6a-h, the FFAR1 (GPR40) Reporter Assay Kit (Cayman chemical) was used. A known FFAR1 agonist, GW9508, was used as a positive control as part of the kit. The assay kit was based on a transfection complex containing DNA constructs for FFAR1, an engineered G protein that directs Gαq activation signals to the Gαs pathway, and a cAMP response element-regulated secreted alkaline phosphatase (SEAP) reporter. HEK293T cells were seeded at a density of 5 × 10 4 cells/well into 96-well plate coated with transfection complex in DMEM (Servicebio) containing 10% FBS (Sigma-Aldrich). After culturing for 24 h, the medium was replaced by serum-free Opti-MEM (Gibco) with the test compounds at a concentration 10 µM. After 24 h, part of the supernatant was collected to assay for SEAP activity, which was measured following the addition of a luminescence-based substrate. Luminescence was measured using a SuPerMax 3100 Multi-Mode microplate reader (Shanghai Flash Spectrum Biotechnology, Shanghai, China). The luminescence value of GW9508 was taken as 100%.

Oral Glucose Tolerance Test (OGTT)
Animals were randomized by weight and divided into groups of 6 mice after quarantine. Mice were fasted for 12 h before the test. The test compounds were introduced by oral gavage in a Tween-80-water suspension. Glucose was given orally at a dose of 2.5 g/kg 30 min after introduction of the compounds. Vildagliptin (Galvus, Novartis Farmaceutica SA, Barcelona, Spain) was used as a positive control at a dose of 10 mg/kg. Blood samples were obtained via tail incision before dosing (time 0) and at 30, 60, 90, and 120 min after the glucose load. The blood glucose concentration was evaluated with a ONE TOUCH Select blood glucose meter (LIFESCAN Inc., Milpitas, CA, USA). The area under the glycemic curve (AUC) was calculated using Tai's model [43].

Conclusions
In this study, we proposed a new synthetic approach for structural analogs of the FFAR1 agonist QS-528. Seven novel compounds were synthesized, varying the positions of the substituents in the aromatic fragments as well as the configuration of the asymmetric center in the bornyl moiety. The obtained compounds did not exhibit cytotoxic properties against HepG2 cells but caused cells to increase their glucose uptake in a dose-dependent manner. The studied compounds were shown to have the ability to activate the receptor at a concentration of 10 µM. Only two compounds were active in the OGTT with CD-1 mice, which allowed us to identify privileged motifs in the molecules, that is, meta-substituted (hydroxyphenyl)propanoic acid and the exobornyl fragment.