The synthesized plant metabolite 3,4,5-tri-O-galloyl quinic acid methyl ester inhibits calcium oxalate crystal growth in a Drosophila model, downregulates renal cell surface Annexin A1 expression, and decreases crystal adhesion to cells

Abstract

The plant metabolite 3,4,5-tri-O-galloyl quinic acid methyl ester (TGAME, compound 6) was synthesized and its potential effect on calcium oxalate monohydrate (COM) crystal binding to the surface of Madin-Darby Canine Kidney Cells type I (MDCKI) and crystal growth in a Drosophila melanogaster Malpighian tubule (MT) model was investigated. Membrane, cytosolic and total Annexin A1 (AxA1), α-enolase and heat shock protein 90 (HSP90) amounts were examined by Western blot analysis after subcellular fractionation, then confirmed by immunofluorescence staining of cultured cells. Pretreatment of MDCKI cells with TGAME for up to 6 h significantly diminished COM crystal-binding in a concentration-dependent manner. TGAME significantly inhibited AxA1 surface expression by immunofluorescence microscopy, whereas intracellular AxA1 increased. Western blot analysis confirmed AxA1 expression changes in the membrane and cytosolic fractions of compound-treated cells, whereas whole cell AxA1 remained unchanged. TGAME also significantly decreased the size, number, and growth of calcium oxalate (CaOx) crystals induced in a Drosophila melanogaster MT model and possessed a potent antioxidant activity in a DPPH assay.

Graphical Abstract

INTRODUCTION

Renal stone disease, also known as urolithiasis, is common with a recent overall estimated prevalence rate of 14.8% that appears to be rising, with a 5 year recurrence rate of up to 50%. ^{1, 2} The first stages of stone formation include nucleation ³, crystallization and retention of these crystals within the kidney. Crystal nucleation is the first step in crystallization and can occur homogeneously or heterogeneously. Homogeneous nucleation demands a high degree of supersaturation with respect to the mineral concerned. By contrast, heterogeneous nucleation is the much more likely mechanism through which crystal initiation occurs in the urine ⁴ and this process can occur in the presence of particulate matters consisting of cell debris, proteins or crystals of another mineral, and is contained within receptacles lined with chemically active cell surfaces. Once a crystal nucleus is formed inside the kidneys ⁴, exposure to the urine makes the stone able to grow by encrustation. ⁵ There are two basic pathways (free-particle and fixed-particle mechanisms) for the establishment of a stone nucleus, both of which can be active in any stone former. Crystallization in turn requires supersaturation of tubular and/or interstitial fluid with respect to ionic compounds such as calcium oxalate (CaOx) and calcium phosphate (CaP). If these processes occur in tubular fluid, the retention of crystals in the kidney may involve aggregation of smaller crystals and/or adhesion to the tubular cell surface. ^{3, 6}

Calcium oxalate monohydrate (COM) is found in up to 70–80 % of stones. ⁷ Many studies have demonstrated that COM crystal binding to renal tubular cells triggers diverse cellular responses including proliferation, mitochondrial dysfunction, oxidative stress, cytotoxicity and cell death. ^{8, 9} One possible strategy to prevent stone formation is to arrest these early intrarenal events after COM crystal nucleation including crystal growth, crystal aggregation, and crystal adherence to renal epithelial cells. ^{10, 11} Previous studies have suggested some possible COM crystal binding proteins on the apical surface of renal tubular cells (e.g. α-enolase, annexin A1 and heat shock protein 90 (HSP90)). ^11–14

Several studies have reported the role of natural products in the prevention of renal stones by inhibiting calcification and hyperoxaluria-induced oxidative stress in animal models of nephrolithiasis. We previously demonstrated that the Copaifera langsdorffii Desf. (Fabaceae) leaf extract increased urinary magnesium and decreased urinary uric acid excretion in rats. ¹⁵ Moreover, treated rats had lower urinary oxalate and higher urinary citrate excretion, reduced intratubular calcification, and decreased osteopontin (OPN) expression. ¹⁶ Galloylquinic acid compounds are the major secondary metabolites found in C. langsdorffii leaf that might significantly contribute to its antiurolithic activity. It is interesting to note that the compounds have an affinity for the kidney. A high concentration (C_max = 15.84 μg/g tissue) was measured in renal tissue after an intravenous pharmacokinetic study performed in an animal model by our group. ¹⁷ This relatively high distribution of the compounds within the kidney might make them particularly effective for preventing kidney stone formation.

Recently, we isolated 16 galloylquinic compounds from the aqueous fraction of the leaves, which displayed potential gastroprotective activity in a mouse model of gastric ulcer and cytotoxicity for gastric adenocarcinoma cells. ¹⁸ The methyl and ethyl 3,4,5-tri-O-galloylquinic acid were previously isolated from Lepidobotrys staudtii Egnl. stem bark and the leaves of Guiera senegalensis J.F. Gmel, respectively. ^{19, 20}

Previous studies also investigated chemically related compounds to galloyquinic acids, such as gallotannin, epigallocatechin-3-gallate and 1,2,3,4,6-penta-O-galloyl-beta-D-glucose in suppressing oxalate-induced oxidative stress and COM crystal binding to renal epithelial cells. ^7, 11, 21 The present study involved the total synthesis of 3,4,5-tri-O-galloyl quinic acid methyl ester (Compound 6, TGAME) (Figure 1) and its evaluation on renal cells and COM crystal adhesion, as well as investigating its effect on CaOx crystallization in a Drosophila melanogaster Malpighian tubule in vivo and ex vivo models.

Figure 1.

Chemical structure of 3,4,5-tri-O-galloylquinic acid methyl ester (TGAME, 6).

RESULTS AND DISCUSSION

Chemistry.

The promising diverse bioactivities of plant extracts^{18, 20, 22} rich in galloylquinic acids prompted our interest to synthesize the tri-substituted 3,4,5-tri-O-galloylquinic acid methyl ester, with the goal of developing lead compounds for kidney stone prevention. The total synthesis included six steps starting from commercially available quinic and Gallic acids. The key step in the synthetic pathway was the affordable Steglich esterification ²³ of methyl quinate with 3,4,5-tribenzyloxybenzoic acid using dicyclohexylcarbodiimide (DCC) and N,N-(dimethylamino) pyridine (DMAP) as the coupling reagents. The chemical structures of the final compound and its synthetic intermediates were elucidated by all possible spectroscopic, spectrometric and spectrophotometric methods of analyses.

The designed synthetic route of the target TGAME is summarized in Scheme 1. The type of esterification reaction used is a mild one, which allowed conversion of the sterically demanding and acid labile methyl quinate, a compound with complex stereochemistry, into the triester form. A protection protocol using benzyl chloride was applied for the hydroxyl groups of Gallic acid to mask their reactivity and formation of undesired side products.

Scheme 1.

Synthesis of 3,4,5-tri-O-galloylquinic acid methyl ester (TGAME, 6). Reagents and conditions: (a) & (b) H₂SO₄, methanol, reflux/18 h for (a) and 24 h for (b); (c) KI, K₂CO₃, benzyl chloride, acetone, reflux/21 h; (d) NaOH, ethanol, reflux/2 h then HCl; (e) DCC, DMAP, CH₂Cl₂, reflux/72 h followed by purification using Silica gel column chromatography, CHCl₃/methanol (100:1); (f) H₂, pressure 80 p.s.i equivalent to 6 kgf/cm², ethyl acetate, 10% Pd/C for 6 h.

The proposed mechanism of the reaction is illustrated in Scheme 2. The chemical structure of TGAME and its intermediates were elucidated and confirmed by IR, ¹H NMR, ¹³C NMR, and mass spectrometric data, which were in full agreement with their structures. ¹⁸ TGAME was obtained as a light yellow amorphous solid. Its specific rotation ${\left[\alpha \right]}_D^{25}$ was - 42.6, which is levorotatory similar to the naturally occurring one in plants. ²⁴ HPLC-UV analysis of the final synthesized product at 280 nm and 254 nm demonstrated one single peak with t_R at 33.57 min. The UV spectrum of this peak showed two characteristic bands at 225 and 275 nm, indicating that the compound belongs to the galloylquinic acids class. The IR spectrum of TGAME showed a broad band at 3387 cm⁻¹ indicating the presence of hydroxyl groups which are H-bonded. A characteristic carbonyl group of the ester appeared at 1707 cm⁻¹. The bands at 1216 and 1034 indicated the presence of a C−O functional group. Two bending bands of C−H appeared at 1475 and 1338 cm⁻¹ reflecting an aliphatic methyl group. Other characteristic IR bands of the compound are presented in the Supporting Information. The ESI-MS/MS analysis of the important synthetic intermediate 5 (C₉₂H₈₀O₁₈), a precursor of the final compound, revealed the following molecular ions: m/z [M+H]⁺ 1473.5239, MS [M+Na]⁺ 1496.5005, MS [M+K]⁺ 1511.4664.

Scheme 2.

The proposed mechanism of the coupling reaction between compound 1 and 4 for the synthesis of 3,4,5-tri-O-galloylquinic acid methyl ester (TGAME, 6).

The molecular weight of TGAME was determined by ESI-MS/MS in both negative and positive ion modes. The experimental and calculated m/z of the protonated molecular ion [M+H]⁺ were 663.1190 and 663.1200. The MS analysis in the negative ion mode revealed a molecular ion of m/z [M-H]⁻ at 661.0878. The base peak formed at m/z 169.0089 through McLafferty rearrangement pathway. Other fragments were formed through the loss of water and CO₂ as shown in the proposed fragmentation pathways in the Supporting Information. The sodiated molecular [M+Na]⁺ ion was also observed in the positive ion mode at m/z 685.1012. The calculated m/z of this ion was 685.1000 with an error of precision of +1.75 ppm. The base peak was at m/z 153.0179. The proposed fragmentation patterns of the protonated and sodiated molecular ion are also illustrated in Supporting Information. The NMR and 2D NMR (HMBC and HSQC) spectral data of compound 5 and 6 agreed with our previously published data. ¹⁸

MDCKI cell cytotoxicity of TGAME and its hydrolytic products.

To determine the safe in vitro concentration, cytotoxicity was assessed by the MTT assay after exposing Madin Darby Canine Kidney type I (MDCKI) cells to increasing concentrations of TGAME (0, 5, 10, 25, 50 μM) for 24 h. Gallic acid and methyl quinate, which are products of galloylquinic acid hydrolysis, were also tested. TGAME did not exert any apparent MDCKI cell cytotoxicity in concentrations up to 155+/−4 μM (Figure 2). The inhibitory concentration 50% (IC₅₀) of the hydrolytic products of the compound, methyl quinate and Gallic acid, were 120+/−7 μM and 34+/−6 μM, respectively. Therefore, the nontoxic concentrations below this IC₅₀ were used in subsequent experiments for the compounds.

Figure 2.

Cytotoxic effects of TGAME (6) and its hydrolytic products in MDCKI cells. The assay of cytotoxicity as evaluated by MTT assay. Cells were plated onto 96-well microplates (1×10 cells/well) and treated with various concentrations of each tested compound (0, 5, 10, 25, 50 μM in DMSO/PBS) for 24 h. Data were expressed as means ± S.D. of three independent experiments.

COM crystal-adherence assay.

MDCKI cells (originally derived from the distal nephron of a female dog), were used as an in vitro model since the initial site of nephrolithiasis has been proposed to be in the distal nephron as supported by histopathological analysis. ^{25, 26} MDCKI cells were incubated with COM crystals (100 μg/mL of culture medium) after pretreatment with 50 μM TGAME for up to 6 h. The concentrations of TGAME used in this study were based upon concentration-response and cytotoxicity studies together with results from the crystal binding assay (concentration-dependence study) that demonstrated TGAME significantly decreased the number of crystals bound to the cell surface at both 25 μM and 50 μM versus the control (Figure 3), while the concentration of 50 μM was more inhibitory compared to 25 μM (p <0.05). Results from the crystal-binding assay (time-course study) demonstrated that pretreatment with TGAME (50 μM) significantly (p <0.05) decreased the number of cell surface associated-COM crystals by approximately 50% at 3 h and 6 h compared to control (Figure 4). Assays were done at early time points (not exceeding 6 h) to focus on the crystal-binding stage and largely avoiding subsequent COM crystal phagocytosis. ^{11, 27}

Figure 3.

Effect of TGAME (6) on COM crystal-binding capability of the cells was dose dependent. MDCKI cells were pretreated with various doses of compound 6 (0, 1, 10, 25, and 50 μM) for 3 h followed by incubation with COM crystals (100 μg/mL of culture medium) for 30 min. A | After washing with plain DMEM culture media, images of the remaining crystals were captured under a phase-contrast microscope for 15 high-power fields (HPF) in each well. The original magnification power was ×400. B | The remaining crystals that adhered on the cell surface were then counted. *p < 0.05 vs. control and **p <0.05 for the effect of the concentration 50 μM vs. 25 μM of compound 6.

Figure 4.

Pretreatment of TGAME (6) reduced COM crystal binding onto the cell surface. MDCKI cells were treated with 0, 25 or 50 μM compound 6 for 1, 3 or 6 h prior to incubation with COM crystals (100 μg/mL of culture medium) for 30 min. A | After washing with plain DMEM culture media, images of the remaining crystals were captured under a phase-contrast microscope for at least 15 high-power fields (HPF) in each well. B | Crystals remaining on the cell surface were then counted. *p < 0.05 vs. control.

TGAME decreased crystal-binding in a concentration-dependent manner across this range with maximal effects at 50 μM. Thus, further experiments used exposure to 50 μM of TGAME for 3 h. Of note, one of the hydrolytic products (Gallic acid) of the compound significantly (p <0.05) decreased the crystal binding at a lower concentration (10 μM) after 3 h versus the control (Figure 5). This indicates that TGAME is able to generate active metabolites, if it undergoes in vivo metabolism by body esterases.

Figure 5.

Effects of TGAME (6) and its hydrolytic products on COM crystal binding to MDCKI cells. The cells were treated with 0, 1 or 10 μM of each compound for 3 h prior to incubation with COM crystals (100 μg/mL of culture medium) for 30 min. After washing with plain DMEM culture media, images of the remaining crystals were captured under a phase-contrast microscope for 15 high-power fields (HPF) in each well. Crystals remaining on the cell surface were then counted. *p < 0.05 vs. control.

Analysis of COM-binding proteins (Annexin A1, α-enolase and HSP90) by Western blot.

We reported previously that when COM crystals were pre-coated with gallotannin (in green tea), their subsequent adhesion to renal tubular cells was reduced. ⁷ A similar study demonstrated that aluminum citrate decreased COM crystal binding to the same cells, when the crystals were pre-coated with it. ²⁸ Moreover, we recently showed that an aqueous extract of Costus arabicus, a plant used in Brazilian folk medicine to treat urolithiasis, inhibited the growth of COM crystals and their adhesion to renal epithelial cells. ²⁹ The inhibitory effect on the binding of crystals to the cell was only significant when pre-coated COM crystals were investigated. Pretreatment of cells with this extract had no effect on crystal adhesion. Unlike the previous studies, we found that TGAME exerted effects when added directly to renal cells followed by treatment with uncoated COM crystals. We propose that the observed biological activities might be related to cellular effects of TGAME and reduction of COM crystal-binding protein expression on the surface of renal cells. One of these proteins is AxA1, ¹⁴ in addition to α-enolase and HSP90. ^{6, 30}

A previous publication documented that epigallocatechin gallate strongly interacts with phospholipid bilayers due to hydrogen bond formation, and it inhibits COM crystal adhesion to renal tubular cells by decreasing α-enolase surface expression. ¹¹ It is noteworthy that AxA1, α-enolase, and HSP90 are all in the database of the lipid raft proteome (http://lipid-raft-database.di.uq.edu.au/). AxA1 is present in the basolateral cell membrane, cytoplasm, cell projection and nucleus. Thus TGAME could decrease cell surface expression of these proteins through transcriptional suppression, altered molecular shuttling, or trafficking from the cytoplasm to the cell surface, since the compound is polyphenolic and could interact with lipid bilayers through hydrogen bond formation. We studied surface expression of these three putative COM crystal receptors on MDCKI cells pretreated with TGAME by immunofluorescence microscopy and confirmed results by Western blot analysis of whole cell lysate and subcellular fractions. The analysis revealed that TGAME significantly reduced the presence of AxA1 and the other two proteins in the membrane fractions (P <0.05), whereas their concentrations in cytosolic fractions was greater (Figure 6). Overall, the total amount of protein in the whole cell lysate was the same. Interestingly, the inhibitory effect of TGAME on AxA1 in the membrane fraction was more significant in comparison with its effect on α-enolase and HSP90 proteins (p<0.05).

Figure 6.

A | Western blot analysis of COM crystal-binding proteins. Proteins obtained from whole cell lysate, cytosolic fraction and membrane fraction of MDCKI with or without 50 μM TGAME (6) treatment, were subjected to Western blotting for annexin A1, α-enolase, HSP90. β-actin served as the loading control. B | Bands intensity was quantitated using AlphaEase FC software (Mayo Clinic). Each bar represents mean ± SEM from 3 independent experiments. *p < 0.05 vs. control. **p < 0.05 for the effect of TGAME on Annexin A1 vs. α-enolase and HSP90.

These data suggested that TGAME decreased AxA1 cell surface expression by inhibiting or re-localizing its shuttling from cytoplasm to the cell membrane. Annexins are Ca²⁺-regulated membrane binding proteins found throughout the animal and plant kingdoms. The Annexin core (Ca²⁺/membrane binding unit) is thought to be one essential element that facilitates membrane interactions. Thus Annexins can organize and assemble phospholipids into specific membrane domains. ³¹ AxA1 protein also plays important roles by controlling cellular Ca²⁺ storage and thus Ca²⁺-dependent signal transduction. Overall, AxA1 localizes to many cellular structures including the plasma, cytoplasm and organelle membranes. ³¹

Alpha-enolase is another COM crystal binding protein on the surface of MDCKI cells, and it was reported that expression of this protein on the MDCKI cell surface increased after exposure to oxalate. ¹¹ Another study confirmed a role of α-enolase in COM crystal binding by using a specific antibody to block α-enolase and demonstrating decreased adhesion to the cell surface. ³⁰

Surface heat shock protein 90 (HSP90) is also found on the surface of renal cells as a putative crystal binding molecule. Downregulating HSP90 expression using specific anti-HSP90 antibodies or small interfering RNA (siRNA) decreased COM crystal adhesion by 50% and 75%, respectively. ¹³

To further support our hypothesis that the COM crystal adhesion inhibition by TGAME is mediated in part by surface AxA1, we used a specific anti-AxA1 antibody. As expected, COM crystal adhesion to the surface of renal cells was significantly reduced (p <0.05) in comparison with the control and control isotypic-IgG treated cells (Figure 7).

Figure 7.

Cell-COM crystal adhesion assay and neutralization by a specific anti-Annexin A1. The confluent polarized MDCKI cell monolayer was incubated with 0.2 μg/mL anti-Annexin A1 antibody or 0.2 μg/mL rabbit isotype-controlled IgG prior to cell-crystals adhesion assay, whereas the cells without antibody treatment served as the blank control. A | After removal of unbound crystals, crystals adherent on the cell surface were imaged by phase-contrast microscopy. B | The adherent crystals were counted from 15 random high power fields (HPF). Data were expressed as means ± SEM of three independent experiments. *p < 0.05 vs. controls.

Examining the expression of surface and intracellular Annexin A1 by immunofluorescence assay and laser-scanning confocal microscopy.

Laser-scanning confocal microscopy demonstrated that AxA1 cell surface expression was dramatically inhibited by pretreatment with 50 μM of TGAME to approximately 50% of control (Figure 8). However, Annexin A1 expression inside cells increased after TGAME treatment.

Figure 8.

Confocal and laser scanning microscopy. TGAME (6) reduced the surface expression of annexin A1 but increased its intracellular level. (A) Intracellular expression of annexin A1 (with cell permeabilization). (B) Surface expression of annexin A1 (without cell permeabilization). MDCKI cells were pretreated with 50 μM compound 6 for 3 h and processed for immunofluorescence study using anti-annexin A1 as a primary antibody and secondary antibody conjugated with Alexfluor 488 donkey antirabbit (shown in green), whereas nuclei were counterstained with TOT-3 dye (shown in blue). Original magnification power was ×400. (C) The graph peaks represent the fluorescence intensities of the staining at the given distances (in μm) along the red bar.

In vivo feeding and ex vivo oxalate ± TGAME birefringence experiment in Drosophila melanogaster Malpighian tubule model.

Our group previously studied the effect of thiosulfate and sulfate on Malpighian tubule (MT) calcium oxalate crystallization in a Drosophila melanogaster (fruit fly) model. ³² Chen et al. fed fruit flies lithogenic agents such as sodium oxalate (NaOx). Their analysis used polarized light microscopy of dissected tubules to show that NaOx feeding resulted in the formation of crystals within the Malpighian tubule lumen in a dose-dependent manner. These crystals were identified as CaOx using scanning electron microscopy and energy-dispersive X-ray spectroscopy. The study also demonstrated that crystal formation decreased lifespan of the flies and that the use of potassium citrate (KCit) as an antiurolithic agent, reduced crystallization inside MT and decreased mortality. Thus, Drosophila appears to be a useful model to study CaOx renal stones, before embarking on the use of higher order experimental animals. ³³ Therefore, we investigated the effect of TGAME on the formation of CaOx crystals within MT in feeding experiments and after ex vivo exposure of the tubules to oxalate (Figure 9). Flies fed oxalate (10 μM) + TGAME manifested a decrease in CaOx crystal number, size and total crystal area within Malpighian tubules after 24 h exposure in comparison to those fed oxalate alone, but these findings were not significant (p >0.05) (Figure 10). However, in the factors such as pharmacokinetic parameters affecting TGAME absorption to reach the blood stream or the need for doses greater than 50 μM in vivo to deliver sufficient amounts to MT to alter CaOx crystallization.

Figure 9.

Malpighian tubule (MT) images from wild type female fruit flies incubated ex vivo with A | oxalate (10 mM) or B | oxalate (10 mM) and 3,4,5-tri-O-galloyl quinic acid methyl ester (TGAME) (100 μM) for 1 h. The arrows indicate CaOx crystals within the lumen of the MT. C | CaOx crystal number, D | total crystal area, and E | average area/crystal per 600 μm of tubule were determined using ImageJ. TGAME significantly decreased crystal number, total crystal area and average area/crystal. Values are means ± SE (n = 12 for both oxalate control group and compound 6-treated group). *p < 0.05 vs. control.

Figure 10.

Malpighian tubule (MT) images from wild type female fruit flies fed A | oxalate (10 mM) or B | oxalate (10 mM) and 3,4,5-tri-O-galloyl quinic acid methyl ester (TGAME) (50 μM) for 24 h. The arrows indicate CaOx crystals within the lumen of the MT. C | CaOx crystal number, D | total crystal area, and E | average area/crystal per 600 μm of tubule were determined using ImageJ. Values are means ± SE (n = 11 for both oxalate control group and compound 6-treated group). p >0.05.

Annexin b11 (Axb11) is also found in Drosophila ³⁴ as a Ca²⁺-dependent phospholipids binding protein (http://flybase.org/reports/FBgn0030749.html) that is known to play important role in cell adhesion, involved in several functions, among them differentiation, cell signaling and membrane fusions. In this case, we speculate that the effect of our compound TGAME in the fly model might be of its action on Axb11, where the decrease in cell surface Axb11 might decrease nucleation on the cell surface.

Antioxidant and DPPH assay.

Studies suggest that renal epithelial cell exposure to CaOx crystals, CaP crystals, and oxalate ions can lead to the generation of reactive oxygen species (ROS), potential cause of injury and/or promote inflammation. Further, the urine of stone formers and hyperoxaluric rats is enriched with markers of inflammation and oxidative stress (OS). ³⁵ Stone formers also have been reported to have reduced peripheral concentrations of antioxidants including α- and β-carotenes. There are many reported mechanisms whereby ROS can increase stone risk including decreased function of macromolecular inhibitors and/or cellular injury that promotes increased crystal adhesion. ³⁶ Antioxidant treatments have also been shown to reduce pathological injury of crystal deposition in hyperoxaluria animal models.

In this study, the DPPH assays used to gauge OS, revealed that TGAME and its metabolite Gallic acid can scavenge free radicals to a similar extent as the standard antioxidant vitamin C with an EC₅₀ 11+/−0.3, 3 and 45+/−2 (Figure 11), respectively. Therefore, the compound is considered to be a potent antioxidant. Methyl quinate did not show any radical scavenging activity as expected, since it is a sugar with no phenolic properties.

Figure 11.

Antiradical efficiency (AE) of TGAME (6) and its hydrolytic products (gallic acid and methyl quinate) in DPPH assay using vitamin C as a standard. Different volumes (2−20 μL) of the compounds (2−20 μM in DMSO) were made up to 40 μL with DMSO, and 1.96 mL of DPPH (0.1 mM) solution was added. The reaction mixture was incubated in dark condition at room temperature for 20 min. Then, the absorbance of the mixture was read at 517 nm. DPPH solution was taken as control, and the % radical scavenging activity of each compound was calculated.

In total, our study supports the hypothesis of Verkoelen and Verhulst that crystal binding is preceded by pathologic alterations in cell surface binding molecules. Nonetheless, TGAME-treatment could not entirely prevent COM adhesion to cells, suggesting that these interactions involve multiple processes and many molecules, perhaps including annexin II, osteopontin and hyaluronan. ^{6, 37} Therefore, other unknown cell surface molecules might mediate the inhibitory effect of TGAME on COM crystal-binding to MDCKI cells.

CONCLUSIONS

We report that TGAME inhibited COM crystal adhesion to renal cells and this effect is mediated by decreased expression of AxA1 on cell surface. Thus, cell surface expression of AxA1 might be a key pathogenic factor in crystal retention and urinary stone formation in vivo. Moreover, the compound significantly decreased CaOx crystal number, size and total crystal area within MT of Drosophila models. Our findings may also be relevant for the observed decrease of crystal deposition when hyperoxaluric animals treated with C. langsdorffii leaf extract. Finally, our findings support a promising role for TGAME in the prevention and modulation of new or recurrent renal stone formation. Further pre-clinical and clinical studies should be performed for the use of this compound in urolithiasis.

EXPERIMENTAL SECTION

Material and methods

The purities of the tested compound were determined by HPLC analysis and MS error of precision data as being ≥ 98%. See the Supporting Information for full details. Biological assays were conducted at Department of Internal Medicine, Division of Nephrology and Hypertension, Mayo Clinic Medical School, USA.

All methods were performed using anhydrous solvents purchased from commercial vendors and used without further purification. Quinic and Gallic acids were purchased from Sigma Adrich, Germany in addition to all other reagents. Hydrogenolysis was performed using a hydrogen reactor system (FAMABRAS, IND. BRAS. 1/8, 255046, Brazil) supported with pressure controller (CLASSE B ABNT). All melting points (°C, uncorrected) were determined in capillary tubes using Fisatom™ melting point device (Mod. 431, Brazil, Serie 1237344). Optical rotation ${\left[\alpha \right]}_D^{25}$ was done using polarimeter (Jasco P-2000), Serial No. A104161232, Japan) at 25°C and a wavelength of 589 nm. Methanol and chloroform were used as solvents. Three readings of ${\left[\alpha \right]}_D^{25}$ were recorded and the average was taken. Chromatographic purification was performed based on flash chromatography using silica gel (pore size 60A°, 40–63 μm, Sigma, batch #MKBC6227) and the solvent systems indicated. Solvent systems are expressed as v/v percent ratios. All reactions were monitored by TLC using fluorescent precoated silica gel plates (Merck, Germany) at short wavelength. Molybdate/H₂SO₄ spray solution was used as the raveling reagent. The IR spectra were recorded using FT-IR spectrophotometer (Perkin Elmer Spectrum Two) in the range of 400–4000 cm⁻¹ by KBr pellet technique. The analytical HPLC analysis was performed on a Shimadzu LC-10ADvp (Japan) operated with multisolvent delivery system, equipped with a Shimadzu SPD-MICAvp photodiode array detector (PDA). Analyses were performed using analytical reversed phase column (Polar-RP C 80 A) with dimensions of 150 mm × 4.6 mm (Shimadzu) and a particle diameter of 4 μm; the mobile phase consisted of acidified water (A) (0.1% formic acid in water) and methanol (B) in gradient conditions as follows: 15–50% of B (45 min), 50–90% of B (45–65 min) and 90–15% of B up to 75 min; flow rate, 1 mL/min, with injection volume of 20 μL. All samples were dissolved in HPLC grade methanol (2 mg/mL) 0.45 μm filtered prior to automatic injection. Mass spectra were acquired on a high resolution MicroTOF II-Q mass Spectrometer (Bruker Daltonics, Billerica, MA, US) fitted with an electrospray ionization (ESI) operating system in the positive and negative ion modes. Accurate masses were obtained using TFA-Na⁺ (sodiated trifluoroacetic acid, 10 mg/mL) as the internal standard; end plate: - 500 volts; capillary: 3500 volts; dry gas temperature: 180°C with flow rate 4L/min; nebulizer gas pressure: 0.4 bar of nitrogen gas; infusion bomb model: Cole Parmer with a flow rate 300μL/ h. ¹H NMR, ¹³C NMR, and two-dimensional (2D) spectroscopic techniques were recorded on a Bruker-Avance DRX500 spectrometer operating at a frequency of 500 MHz. Samples were dissolved in Aldrich deuterated dimethyl sulfoxide (DMSO-d6) for compounds 1–4, 6 and chloroform (CHCl₃-d) for compound 5. Data for ¹H NMR are reported as follows: chemical shift (δ, ppm), multiplicity (s = singlet, d = doublet, t = triplet, m = multiplet, coupling constant (s) in Hz, integration). Data for ¹³C NMR are recorded in terms of chemical shift (δ, ppm).

Synthetic steps of TGAME

Synthesis of quinic acid methyl ester. Methyl-(1S,3R,4S,5R)-1,3,4,5 tetrahydroxycyclohexanecarboxylate (1)

To a stirred solution of methanol (160 mL) and sulfuric acid (2 mL), quinic acid (4g) was added. The mixture was refluxed at 79–80°C for 18 h. TLC analysis demonstrated a whole conversion of quinic acid into a single product spot (R_f = 0.56; ethyl acetate/methanol/water/formic acid = 6:3:3:0.2). The reaction mixture then was poured over solid NaHCO₃ until no effervescence (neutralization was detected by litmus paper), and then it was filtered. The filtrate was evaporated to give a solid crude hygroscopic residue (8.67 g) that was recrystallized in methanol to give rosette crystals (4.60 g, 91%) with undefined melting point; molecular formula C₈H₁₄O₆.

Synthesis of methyl 3,4,5-trihydroxybenzoate (2)

Sulfuric acid (1 mL) was added drop-wise to a stirred solution of Gallic acid in methanol (10 g dissolved in 100 mL). The resulting reaction mixture was refluxed for 24 h. The solvent was evaporated and an off-white precipitate was obtained which was dissolved in ethyl acetate (150 mL). The organic phase was washed with distilled water (3 X 20 mL) and sodium bicarbonate solution (3 X 20 mL), successively. The organic phase was dried over anhydrous sodium sulphate, then evaporated to furnish 8 g of a yellowish-white precipitate (R_f = 6.3; ethyl acetate/formic acid = 8:2). The product was purified by recrystallization in methanol and clusters of plate-like crystals were formed (6 g, 60%); mp 200–202°C; molecular formula C₈H₈O₅.

Synthesis of methyl 3,4,5-tribenzyloxybenzoic acid (3)

Compound 2 (5 g (27.15 mmol)), 2 g (12 mmol) of KI and 22 g (159 mmol) of dry K₂CO₃ were mixed in 250 mL of acetone and stirred for 20 min at room temperature (RT). Next, 11 g (87 mmol) of benzyl chloride in 50 mL of acetone was added portion-wise. The suspension was refluxed for 21 h. TLC screening revealed a single spot of the formed product (R_f = 0.30; hexane/ethyl acetate = 8:2). The reaction mixture was filtered, then the filtrate was concentrated under a vacuum. The resulting residue was re-suspended in 200 mL of CH₂Cl₂, then filtered through Celite and was concentrated under a vacuum. The product was obtained as off-white solid (12 g, 97%). Molecular formula C₂₉H₂₆O₅.

Synthesis of 3,4,5-tribenzyloxybenzoic acid (4)

Compound 3 (10 g) was mixed with a solution of 1.77 g (44.25 mmol) sodium hydroxide in 250 mL of 95% ethanol. The resulting mixture was heated up for 2 h with reflux, followed by immediate pouring into 260 mL of 0.6 M HCl and stirred for 10 min, and finally filtered under vacuum. The resulting product was washed with a 1:1 mixture of 95% ethanol and water (50 mL), followed by pure water (50 mL), 95% ethanol (50 mL), methanol (2 X 25 mL), and di-ethyl ether (25 mL). The yield was dried under vacuum overnight to give a white solid (8.9 g, 94 %); mp 192–195°C; C₂₈H₂₄O₅.

Synthesis of (1S,3R,4S,5R)-5-hydroxy-5-(methoxycarbonyl)cyclohexane-1,2,3-triyl tris (3,4,5-tris (benzyloxy) benzoate) (5)

A suspension of 7 g (16 mmol) of 3,4,5-tribenzyloxybenzoic acid, 0.82 g (4 mmol) of methyl quinate, 4 g (20 mmol) of dicyclohexylcarbodiimide (DCC), and 12.24 g (18.4 mmol) of N,N-(dimethylamino) pyridine (DMAP) in 500 mL dry dichloromethane, was refluxed for 72 h. After cooling to RT, the mixture was filtered, then concentrated using a rotary evaporator. The resulting crude residue (14.5 g) was purified by flash chromatography (two times, 97 fractions of 20 mL each were collected/per one time) using a mixture of chloroform/methanol (100:1). Both the reaction and the column fractions were monitored by TLC (dichloromethane/toluene/ethyl acetate = 7.5:2.5:0.1) and the desired product fractions (f 71 to 81 and f 82 to 91) were combined and concentrated to give (1.74 g, 29%) of a faint yellowish-white glassy material (R_f = 0.2); mp 129–130°C; molecular formula C₉₂H₈₀O₁₈; ${\left[\alpha \right]}_D^{25}$ : − 49.2 (c 0.15; CHCl₃); m/z [M+H]⁺ 1473.5239, MS [M+Na]⁺ 1496.5005, MS [M+K]⁺ 1511.4664.

Synthesis of TGAME, (1S,3R,4S,5R)-1-hydroxy-3,4,5-tris ((3,4,5-trihydroxybenzoyl) oxy) cyclohexanecarboxylic acid (6)

A suspension of 44 mg of palladium (10 wt % on activated carbon) and 200 mg of compound 5 in 15 mL of dry ethyl acetate was stirred at RT for 6 h inside a cell of a hydrogen reactor system at RT under a hydrogen gas atmosphere with a pressure of 80 p.s.i, equivalent to 6 kgf/cm². Next, the reaction mixture was filtered through Celite filter-aid using a mixture of ethyl acetate and methanol. The filtrate was concentrated under vacuun and the obtained residue was lyophilized, recrystallized in ethyl acetate to give (93 mg, 99%). The desired synthetic compound was obtained as a light yellow amorphous solid; mp 137–140°C. UV λ_max (methanol): 280 and 254 nm. Molecular formula C₂₉H₂₆O₁₈; MS: m/z 663.1190 [M+H]⁺, 661.0878 [M-H]⁻, 685.1012 [M+Na]⁺. ${\left[\alpha \right]}_D^{25}$: − 42.6 (c 0.50; CH₃OH). IR υ_max (KBr): (3387 cm⁻¹, O−H “H-bonded”), (2953 cm⁻¹, C−H aliphatic “stretch”), (1707 cm⁻¹, C═O), (1616 cm⁻¹, C−C aromatics), (1475 and 1338 cm⁻¹, C−H (─ CH₃) “bend”), (1216 and 1034 cm⁻¹, C−O), (764 cm⁻¹, C−H aromatics “out-of-plane bend”). ¹ HNMR (DMSO/TMS): δ (ppm) 2.30 (m, 2H), 2.40 (m, 2H), 3.43 (s, 1H), 3.49 (s, 3H), 3.75 (s, 9H), 4.90 (m, 1H), 5.40 (dd, 1H), 5.60 (m, 1H), 6.89 (s, 2H), 6.92 (s, 2H), 6.94 (s, 2H). ¹³C NMR (DMSO/TMS): δ (ppm) 29.2, 35.5, 52.4 (-CH₃), 63.3, 68.4, 72.8, 109.2, 119.2, 119.8, 128.1, 128.2, 128.4, 128.7, 139.0, 139.3, 145.8, 145.9, 165.0, 165.2, 165.6, 173.9.

For more details on ¹H, ¹³C NMR, 2D NMR, DEPT and FT-IR spectral data of all synthetic compounds, see Supporting information.

Cytotoxicity by MTT assay

The cytotoxicity of TGAME, compound 6 and its synthetic intermediates was evaluated by MTT colorimetric assay (MTT Kit ThermoFisher Scientific, USA). The cells of MDCKI (ATCC® CRL-2935™, USA, provided by Dr John Lieske, Mayo Clinic Medical School, MN, USA) were seeded onto 96-well microplates at a density of 1×10⁴ cells per well and treated with different concentrations (0, 5, 10, 25, 50 μM in DMSO/ phosphate buffered saline (PBS)) of the compounds for 24 h. The MTT (15 μL) working solution (5 mg/mL in PBS) was added to each well and incubated at 37°C for 3 h. After the addition of the stop solution (100 μL) the optical density (OD) was measured at 570 nm using a microplate reader (SpectraMax 190 Microplate Reader, USA). Cell viability was calculated as a percentage of viable cells in compound-treated group versus untreated control by the following equation:

Cell viability (%) = [OD (compound) − OD (blank)] / [OD (control) − OD (blank)]×100

Cell-COM crystal binding assay and neutralization by a specific anti-Annexin A1 antibody Preparation of COM crystals:

Sodium oxalate (1mM) in Tris buffer saline (50 mL) containing 90 mM NaCl (pH 7.4) was added to a gently stirred solution of 10 mM calcium chloride (50 mL). The mixture was incubated at RT overnight and then centrifuged at 3,000 rpm for 5 min. Crystals were re-suspended in methanol (1 mL) and centrifuged for 5 min at 3,000 rpm. The methanol supernatant was aspirated and the crystals were air-dried at RT then sterilized using UV radiation for 30 min before use in cell culture work. The dried crystals were examined under Zeiss Axiophot microscope using 10 X differential interference contrast (DIC) lens and appeared as six-sided prisms.

MDCKI cells (5 × 10⁵ cells) were seeded into 6-well cell culture plate and incubated in a complete culture medium for 24 h. Compound 6 was added (0, 1 μM, 10 μM, 25 μM, 50 μM) to the cells for 3 h (concentration-dependent study), whereas the cells kept in culture medium without compound represented as the control. The cells were pretreated with 0, 25 or 50 μM (the optimal two concentrations obtained in the previous study) of compound 6 for 1, 3 or 6 h (time-course study). The culture medium was aspirated and the cells were washed gently with PBS. Thereafter, cells were incubated for 30 min with COM crystals (100 μg crystals/mL culture medium), followed by washing 5 times with plain medium to remove non-adherent crystals. Bound crystals in each well were imaged and counted under a phase-contrast microscope (Zeiss) in 15 random high power fields (HPF).

The surface expression of Annexin A1 was blocked by pretreating the cells with rabbit polyclonal anti-Annexin A1 antibody (Cell Signaling, USA). Briefly, the confluent polarized cells were incubated for 15 min with bovine serum albumin (1%) in membrane preserving buffer (0.1mM CaCl₂ and 1mM MgCl₂ in PBS) to block non-specific bindings. Cells were then washed with the same buffer (3X) followed by incubation anti-Annexin A1 antibody (0.2 μg/mL) or rabbit isotype-controlled IgG (0.2 μg/mL) at 37 °C for 30 min. COM crystals (100 μg crystal/mL medium) were incubated for 1 h at 37 °C, after washing the cells with the buffer. Unbound crystals were removed by washing with PBS (5X). Bound COM crystals were counted under a phase-contrast microscope by assessing 15 random HPF.

Analysis of Annexin A1, α-enolase and HSP90 by Western blot

Subcellular fractionations of MDCKI cell protein and subsequent Western blot were used to measure the concentrations of cytosolic, membrane and whole Annexin A1, α-enolase and HSP90. MDCKI cells (5 × 10⁵ cells) were seeded into each well of 6-well plate and cultured in a complete medium overnight. After treatment with 50 μM TGAME for 3 h, the cells were washed with cold PBS and incubated for 10 min at 4°C with cytosolic extracting buffer containing mainly 10 mM piperazine-N, N′-bis (PIPES), 0.5 mM EDTA and 0.02 % digitonin, with gentle shaking. The cells were scrapped with the buffer, collected and centrifuged for 10 min at 10,000 r.p.m. The supernatant containing cytosolic proteins was next collected. The remaining cell pellet was mixed with membrane extracting buffer (60 mM Tris-HCl of pH = 6.8, 2% sodium dodecyl sulfate (SDS), 10 % glycerol and 5 % 2-mercaptoethanol) and incubated for 30 min in ice bath with gentle shaking every 10 min, and named as “membrane fraction”. Whole cell lysate was extracted from using whole lysate buffer. The protein concentrations were determined by Bradford’s method ³⁸ using Pierce BCA Protein Assay Kit (Thermo Scientific).

Each protein fraction (30 μg/lane) was mixed with laemmli sample buffer (1:1), 2x (Bio Rad) and separated by 7.5 % ready gel (BioRad), where all samples were boiled at 99°C for 5 min prior to loading onto gel. The gel was run for 90 min at 100 V then transferred onto a nitrocellulose membrane using 1x transfer buffer. The membrane was probed by incubation at 4°C overnight with rabbit polyclonal anti-Annexin A1 (1:1000), rabbit polyclonal anti-HSP90 (1:1000), rabbit polyclonal anti-α-enolase (1:1000) and mouse monoclonal anti-Actin (1:1000) antibodies (Cell Signaling ®, USA) in blocking buffer, separately. After washing with TBS/0.1 % tween three times for 5 min each, the membrane was incubated with the corresponding secondary antibody conjugated with horseradish peroxidase (Rabbit IgG for all primary antibodies or mouse IgG Horseradish Peroxidase for β-actin) (1:5000) (Cell Signaling ®) in 1x Tris-buffered saline /0.1% tween 20 and 5% milk for 1 h at RT. The immunoreactive bands were detected by developing in SuperSignal West Femto maximum sensitive chemiluminescence substrate solutions (Thermo Scientific, USA), and then visualized by autoradiogram. The intensity corresponding to each band was measured by AlphaEase FC software (Mayo Clinic).

COM crystals bound-Annexin A1 protein immunofluorescence staining and confocal microscopy

Collagen buffer (200 μL) was added to a chamber slide and incubated for 15−20 min at 37ºC followed by washing with 1X PBS. MDCKI cells (12,000/well) were seeded in each chamber, and incubated at 37°C overnight. The media was removed and cells were treated with/without compound 6 (50 μM) for 3 h. Cells were then gently washed with PBS followed by fixation with 4% P-formaldehyde in saline (200 μL) for 15 min at RT, and finally washed twice with cold 1X PBS. Triton (0.2 %) in PBS for 5 min was used for cell permeabilization, then the cells were washed three times with PBS. A blocking buffer (200 μL) was added for one hour at RT, then replaced by the primary antibody with incubation overnight at 4°C. The next day cells were washed with PBS (3X) for 5 min. The 1° antibody was replaced with the 2° antibody and incubated in the dark for 60 min, followed by washing with PBS (3X). The counterstain (TOTO-3) was added and incubated for 20 min at RT, then washed one time with 1XPBS. The chamber frames were removed. The slide was mounted with Vectashield Mounting medium then the cells were examined using Zeiss confocal microscope (Axiovert 100 M) with 63 X oil immersion lens.

DPPH assay for antioxidant activity

To 1.96 mL of DPPH (0.1 mM) solution, different volumes (2 – 20 μL) of compound 6, Gallic acid or methyl quinate (1 mM in DMSO), made up to 40 μL with DMSO were added. The reacted mixtures were incubated for 20 min at RT in the dark and the absorbance was measured at 517 nm. Two mL of DPPH was served as control. The % of radical scavenging activity of tested compounds was calculated using the following formula:

$\%RSA=\frac{\text{Absorbance of control}-\text{Absorbance of test}}{\text{Absorbance of control}}\mathit{\times }100$

Drosophila crystallization assay: Ex vivo oxalate ± TGAME

Oregon R (red) Drosophila melanogaster flies were used as wild type for these experiments, and calcium oxalate crystallization monitored as birefringence using DIC optics. ²⁴ Flies were raised on standard Drosophila media (fly food) at 25°C. Twenty four female flies (7 days age) were used in the experiment. Subsequently, twenty four MT were obtained by dissecting the flies in Schneider’s solution and transferred to poly-L-lysine coated slides with iPBS (insect phosphate buffer saline). The tubules were divided into two main groups (control and treated group; n=12 each). The treated group was pretreated with 100 μM of TGAME in iPBS for 20 min, followed by incubation with 100 μM of compound 6 and 10 μM Na-oxalate. Control group was incubated only with 10 μM Na-oxalate in iPBS.

MTs were incubated for 1 h with fresh solutions replaced every 15 min. Crystallization of CaOx was monitored by imaging the initial 600 μm of the anterior MTs using a Zeiss Observer® using DIC microscopy. Crystallization was quantified using ImageJ as previously. ³² Drosophila feeding experiment: In vivo oxalate ± TGAME

Twenty two female flies were used in the experiment (11 control and 11 treatment) and then kept for 24 h on feeding with fly food mixed either with sodium oxalate (10 μM) or a mixture of TGMAE (50 μM) and sodium oxalate (10 μM). After 24 h, the flies were dissected in Schneider’s medium and MTs were transferred immediately to poly-L-lysine coated slides with iPBS. Crystallization was monitored by imaging the anterior MTs as above.

ABBREVIATIONS USED

COM

calcium oxalate monohydrate

CaOx

Calcium oxalate

CaP

Calcium Phosphate

HSP90

heat shock protein 90

MDCKI

Madin Darby Canine Kidney type I

OPN

osteopontin

TGAME

3,4,5-tri-O-galloyl quinic acid methyl ester

DCC

N,N’-dicyclohexylcarbodiimide

DMAP

4-(Dimethylamino)pyridine

AxA1

Annexin A1

OS

oxidative stress

ROS

reactive oxygen species

DIC

differential interference contrast

DPPH

2,2-diphenyl-1-picrylhydrazyl

HPF

high power field

MT

Malpighian tubules

iPBS

insect phosphate buffer solution