Lactate dehydrogenase A inhibitors with a 2,8-dioxabicyclo[3.3.1]nonane scaffold: A contribution to molecular therapies for primary hyperoxalurias

Bioorganic Chemistry 129 (2022) 106127 Available online 8 September 2022 0045-2068/© 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Lactate dehydrogenase A inhibitors with a 2,8-dioxabicyclo[3.3.1]nonane scaffold: A contribution to molecular therapies for primary hyperoxalurias Alfonso Alejo-Armijo a , Cristina Cuadrado a , Joaquin Altarejos a , Miguel X. Fernandes b , Eduardo Salido c , * , Monica Diaz-Gavilan d , Sofia Salido a , * a Departamento de Química Inorg´ anica y Org´ anica, Facultad de Ciencias Experimentales, Universidad de Ja´ en, Campus de Excelencia Internacional Agroalimentario ceiA3, 23071 Ja´ en, Spain b Instituto Universitario de Bioorg´ anica, Universidad de La Laguna, 38206 La Laguna, Spain c Hospital Universitario de Canarias & Center for Rare Diseases (CIBERER), 38320 Tenerife, Spain d Departamento de Química Farmac´ eutica y Org´ anica, Facultad de Farmacia, Universidad de Granada, 18071 Granada, Spain ARTICLE INFO Keywords: Primary hyperoxaluria Selective lactate dehydrogenase A inhibitors 2,8-Dioxabicyclo[3.3.1]nonane scaffold Flavylium salts Hyperoxaluric mouse hepatocytes Oxalate ABSTRACT Human lactate dehydrogenase A (hLDHA) is one of the main enzymes involved in the pathway of oxalate synthesis in human liver and seems to contribute to the pathogenesis of disorders with endogenous oxalate overproduction, such as primary hyperoxaluria (PH), a rare life-threatening genetic disease. Recent published results on the knockdown of LDHA gene expression as a safe strategy to ameliorate oxalate build-up in PH patients are encouraging for an approach of hLDHA inhibition by small molecules as a potential pharmacological treatment. Thus, we now report on the synthesis and hLDHA inhibitory activity of a new family of compounds with 2,8-diox- abicyclo[3.3.1]nonane core (23–42), a series of twenty analogues to A-type proanthocyanidin natural products. Nine of them (25–27, 29–34) have shown IC 50 values in the range of 8.7–26.7 µM, based on a UV spectrophotometric assay, where the hLDHA inhibition is measured according to the decrease in absorbance of the cofactor β-NADH (340 nm). Compounds 25, 29, and 31 were the most active hLDHA inhibitors. In addition, the inhibitory activities of those nine compounds against the hLDHB isoform were also evaluated, finding that all of them were more selective inhibitors of hLDHA versus hLDHB. Among them, compounds 32 and 34 showed the highest selectivity. Moreover, the most active hLDHA inhibitors (25, 29, 31) were evaluated for their ability to decrease the oxalate production by hyperoxaluric mouse hepatocytes (PH1, PH2 and PH3) in vitro, and the relative oxalate output at 24 h was 16% and 19 % for compounds 25 and 31, respectively, in Hoga1 -/- mouse primary hepatocyte cells (a model for PH3). These values improve those of the reference compound used (stiripentol). Compounds 25 and 31 have in common the presence of two hydroxyl groups at rings B and D and an electron-withdrawing group (NO 2 or Br) at ring A, pointing to the structural features to be taken into account in future structural optimization. 1. Introduction Primary hyperoxaluria (PH) is a rare disease linked to the liver metabolism that results in an overproduction of oxalate anion. The main risk associated with this excess of oxalate is the formation of poorly soluble calcium oxalate (CaOx) crystals and stones in kidneys and urinary tract. The continuous CaOx deposition leads to a remarkable kidney damage and, in most cases, to the consequent end stage renal disease (ESRD). PHs are inherited errors of glyoxylate metabolism and three types of PH (PH1, PH2 and PH3) have been reported. PH1 is the most common and severe form of PH, due to mutations in the AGXT gene; PH2 is caused by mutations in the GRHPR gene; and PH3 is caused by mutations in the HOGA1 gene. Loss-of-function mutations in any of these three genes result in a deficit to detoxify glyoxylate, which is then converted into oxalate by hepatic lactate dehydrogenase A (LDHA) [1,2]. Until very recently, the applied treatments were not sufficient to avoid recurring stones or ESRD and, thus, for severe forms of PH, combined liver and kidney transplantation was the only curative treatment [3]. However, current advances in molecular therapy and clinical * Corresponding authors. E-mail addresses: [email protected] (E. Salido), [email protected] (S. Salido). Contents lists available at ScienceDirect Bioorganic Chemistry journal homepage: www.elsevier.com/locate/bioorg https://doi.org/10.1016/j.bioorg.2022.106127 Received 14 February 2022; Received in revised form 22 May 2022; Accepted 2 September 2022 Bioorganic Chemistry 129 (2022) 106127 2 research have led to interesting results for the reduction of the excess systemic oxalate levels in PH. These therapeutic approaches range from targeting molecular defects in the liver to the use of probiotics to enhance oxalate intestinal elimination [4–6]. However, the decreasing of hepatic oxalate synthesis by inhibition of the enzymes involved in the production of oxalate presents a more feasible therapeutic approach. Thus, the inhibition of glycolate oxidase (GO) enzyme [7–9] or hydroxyproline dehydrogenase (HYPDH) enzyme [10,11] are substrate reduction strategies, since both enzymes are involved in the production of glyoxylate. In addition, there is plenty of evidence that supports that the hepatic isozyme lactate dehydrogenase A (LDHA) is the key enzyme responsible for converting glyoxylate to oxalate. Therefore, the inhibition of this enzyme is an attractive strategy in the prevention of oxalate formation for the three types of PH. In fact, two iRNA therapies have been designed to specifically inhibit hepatic expression of HAO1 gene, which encodes GO enzyme [8], and of LDHA gene, which encodes the major isoform of LDH enzyme in the liver (LDHA) [12,13]. As a result, the FDA and EMA agencies approved in November 2020 the first pharmacological treatment for PH1, based on siRNA inhibition of GO, for Lumarisan (Alnylam Pharmaceuticals). Thereafter, Nedorisan (Dicerna Pharmaceuticals Inc.), based on siRNA inhibition of LDHA, started its phase 3 clinical trial [14]. Results from this ongoing clinical trial using siRNA to target liver LDHA expression have not shown significant adverse effects (personal communication, article in preparation) and no safety concerns were raised in the phase I study already published [15]. Once GO and LDHA inhibition with siRNA have been clinically validated as a safe therapeutic method for the treatment of PHs [8,13], an alternative or complementary strategy is the inhibition of GO or LDHA by small molecules. These constitute a classical approach and, in contrast to biopharmaceuticals, present the advantage of possible oral administration and, in general, lower production costs. Several patent applications and recent publications have described GO inhibitors (GOi’s), most of them sharing the common chemical feature of an aryl carboxylic acid, such as 1,2,3-thiadiazole-4-carboxylic acid [7], 1,2,3- triazole-4-carboxylic acid [16] and more recently salicylic acid derivatives [9] and indazole-3-carboxylic acid [17]. Although LDHA was suggested as a more potent target than GO to reduce urinary oxalate levels [12], to date there are few reports exploring the inhibition of LDHA with small molecules as potential therapeutic agents for PH treatment. Since there are different tetrameric LDH isoforms located in different organs, such as LDHA in skeletal muscle and liver, and LDHB in heart and brain, mainly, a liver organ selective inhibition of hLDHA is considered necessary to validate hLHDA as a safe therapeutic strategy in PH patients, thus avoiding likely nondesired secondary effects [5]. In a recent study, a patient affected by severe PH1 was treated with a commercial anticonvulsant drug, stiripentol, traded as Diacomit® (Biocodex, France) (Fig. 1), for several weeks showing a decrease of urine oxalate excretion without side effects [18]. However, the effectiveness of this drug seems to depend on the state of the renal function of the patient [19,20]. Forthcoming phase 2 clinical trial (NCT03819647) could yield more information on safety and efficacy of its monotherapy for the treatment of PHs [21]. The hypothesis of double GO/LDHA inhibition leading to more effective and safe drugs has been introduced [5,9,22]. Díaz-Gavil´ an and col. bet on using the same pharmacophore with a salicylic acid core against both enzymes [22]. Likewise, Lowther and col. have synthesized and evaluated compounds whose structures result from the merger of two scaffolds, one of them is present in a reported GOi and the other one in LDHA inhibitors (LDHAi’s) [23]. Given the known role of LDHA in aerobic glycolysis in malignant cells [24], a wide number of LDHAi’s have been reported in several review articles [5,25,26] (Fig. 1). Pharmaceutical companies, such as AstraZeneca [27], GlaxoSmithKline [28] and Genentech [29], have developed some of them as potential therapeutics in oncology. However, limitations related to the lack of selectivity or poor pharmacokinetic properties have not led to additional in vivo experiments and none of them are currently in clinical trials. Within the huge structural variability of reported LDHAi’s, our attention has been focused on polyphenolic flavone-based inhibitors, such as morin [30], epigallocatechin [31], galloflavin [32] and luteolin- 7-O-β-D-glucopiranoside [33], which present micromolar IC 50 values (Fig. 1). These natural products were found to be potential anticancer agents due to their inhibitory activity against hLDHA [30–33]. In addition, a LDHA inhibition assay on procyanidin-enriched fractions from Spatholobus suberectus extract showed moderate inhibition percentages (55 %) for those fractions predominantly composed of tetrameric to hexameric procyanidins [34]. Procyanidins and other proanthocyanidins are a large group of natural compounds with a 2,8-dioxabicyclo[3.3.1]nonane scaffold, whose biological activities have been evaluated in several occasions [35–40]. However, the ability of compounds with such scaffold to inhibit hLDHA has not been studied yet. Thus, according to our knowledge in the synthesis of this kind of compounds [38,39], we report herein the synthesis of a series of 2,8-dioxabicyclo[3.3.1]nonane derivatives, the evaluation of their hLDHA and hLDHB inhibitory activities, and the in vitro assay of a selection of them on hyperoxaluric mouse hepatocytes, in order to find some preliminary structure–activity relationships that help to design more powerful compounds in the future. Regarding the synthetic protocol chosen, most of the reported methods to synthesize these bicyclic compounds use 2-hydroxychalcones as starting materials [41–45], but we have taken advantage of our experience on the Fig. 1. Structures of some known LDHA inhibitors (LDHAi’s). A. Alejo-Armijo et al. Bioorganic Chemistry 129 (2022) 106127 3 synthesis of flavylium salts [46] and their use as starting materials [38]. 2. Materials and methods 2.1. General experimental methods All chemicals were purchased and used without further purification, except absolute MeOH and THF, which were prepared and dried, prior to use, according to standard methods [47]. All reactions were carried out under argon atmosphere and at room temperature or in oil baths with electronic temperature control, unless otherwise mentioned. The progress of the reactions was monitored by (a) analytical high-performance liquid chromatography (HPLC) and (b) analytical thin-layer chromatography (TLC) on silica gel 60 F 254 precoated aluminum sheets (0.25 mm, Merck Chemicals, Darmsdadt, Germany) and spots visualized under UV light (254 nm). Purifications of synthesized compounds were performed by semipreparative HPLC or by column chromatography (CC) using Sephadex LH-20 or Silica gel 60 (particle size 0.040–0.063 mm) (Merck Chemicals, Darmsdadt, Germany). Analytical and semipreparative HPLC were conducted on a Waters 600E instrument (Waters Chromatography Division, Milford, MA, USA) equipped with a diode array detector (DAD), scan range: 190–800 nm (Waters CapLC 2996 Photodiode Array Detector, Waters Chromatography Division, Milford, MA, USA), and operating at 30 ◦C. Analytical HPLC analyses were performed on a C18 reversed-phase Spherisorb ODS-2 column, 250 mm ×3 mm i.d., 5 μ m (Waters Chromatography Division, Milford, MA). The best separation was obtained with H 2 O:CH 3 COOH, 99.8:0.2, v/v (solvent A) and methanol: CH 3 COOH, 99.8:0.2, v/v (solvent B) at a flow rate of 0.7 mL/min: linear gradient from 30 % to 100 % B for 25 min; 100 % B for 15 min; and 5 min to return to the initial conditions. The total run time excluding equilibration was 45 min. Purity of the compounds was assessed by analytical HPLC at 280 nm on the C18 reversed-phase described above (data are included in the Supplementary Material section). Semipreparative HPLC separations were performed on a C18 reversed-phase Spherisorb ODS-2 column, 250 mm ×10 mm i.d., 5 μ m (Waters Chromatography Division, Milford, MA) at a flow rate of 5 mL/ min. Infrared (IR) spectra were recorded on a Bruker Tensor 27 Fourier transform infrared (FTIR) spectrometer (Bruker Optik GmbH, Ettlingen, Germany) using an Attenuated Total Reflectance (ATR) accessory, and only characteristic absorptions ( ν , cm −1 ) are reported. 1 H NMR and 13 C NMR spectra were recorded on a Bruker Avance 400 spectrometer (Bruker Daltonik GmbH, Bremen, Germany) operating at 400 and 100 MHz for 1 H and 13 C, respectively. Deuterated methanol (CD 3 OD) was used to prepare solutions of purified compounds for NMR. For flavylium salts a drop of DCl was added to ensure acid conditions. The chemical shifts (in ppm) were referenced to solvent peaks as internal reference. The coupling constants (J) are quoted in hertz (Hz). The following abbreviations are used: d, doublet; t, triplet; m, multiplet; br s, broad singlet; br d, broad doublet; dd, doublet of doublets; ddd, doublet of doublet of doublets; td, triplet of doublets; dq, doublet of quartets. The complete assignment of 1 H and 13 C signals was performed by analysis of the correlated homonuclear (COSY) and heteronuclear (HMBC, HSQC) 2D NMR spectra. High-resolution mass spectra (HRMS) were recorded on an Agilent 6520B Quadrupole time-of-flight (QTOF) mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) with an electrospray ionization (ESI) interface operating in positive or negative mode. 2.2. General method A for the synthesis of flavylium salts (10–19) A mixture of the salicylic aldehyde derivative (2 mmol), the acetophenone or propiophenone derivative (2 mmol), 98 % H 2 SO 4 (0.6 mL; 10.8 mmol) and HOAc (2.6 mL) was stirred overnight at room temperature according to procedures previously used by us [38,39,46]. Then, Et 2 O (30 mL) was added and a reddish solid precipitated. The solid was filtered off and carefully washed with Et 2 O and dried. The flavylium salts 10 (0.605 g, 90 % yield), 11 (0.506 g, 79 % yield), 12 (0.586 g, 77 % yield), 13 (0.555 g, 76 % yield) and 14 (0.672 g, 85 % yield) were described previously with similar yields and their structures confirmed by comparison of their spectral data with those reported in the literature: 10 [48], 11 [49], 12 [39], 13 [50], and 14 [38]. 2.2.1. 3-Methyl-4′-hydroxy-6-nitroflavylium hydrogen sulfate (15) Method A was followed by using 2-hydroxy-6-nitrobenzaldehyde (2) (0.334 g, 2 mmol) and 4′-hydroxypropiophenone (9) (0.300 g, 2 mmol). Pure compound 15 was obtained as a red–orange solid (0.690 g, 91 % yield). Melting point: 260 ◦C (decomposes); 1 H NMR (400 MHz, DCl/ CD 3 OD, pD ≈1.0) δ 8.11–8.00 (m, 2H, H-4, H-7), 7.33–7.20 (m, 2H, H- 2′, H-6′), 6.98 (d, J =8.4, 1H, H-8), 6.79–6.74 (m, 2H, H-3′, H-5′), 6.73 (br s, 1H, H-5), 1.66 (d, J =1.3, 3H, CH 3 ); 13 C NMR (100 MHz, DCl/ CD 3 OD, pD ≈1.0) δ 159.3 (C-2, C-4′), 159.1 (C-9), 143.4 (C-6), 134.9 (C- 3), 133.4 (C-10), 128.9 (C-2′, C-6′), 125.8 (C-7), 123.9 (C-5), 123.1 (C- 4), 116.8 (C-8), 116.1 (C-3′, C-5′), 106.8 (C-1′), 19.1 (CH 3 ); FT-IR (ATR) ν max : 3069, 1589, 1535, 1489, 1408, 1308, 1277, 1165, 1119, 1055, 943, 862, 824, 750, 706, 681 cm −1 ; HRMS (ESI-TOF) m/z [M] + Calcd. for C 16 H 12 NO 4 282.0766, found 282.0762. 2.2.2. 3′,4′-Dihydroxy-6-chloroflavylium hydrogen sulfate (16) Method A was followed by using 2-hydroxy-5-chlorobenzaldehyde (3) (0.313 g, 2 mmol) and 3′,4′-dihydroxyacetophenone (6) (0.304 g, 2 mmol). Pure compound 16 was obtained as a red-brownish solid (0.608 g, 79 % yield). Melting point: 245 ◦C (decomposes); 1 H NMR (400 MHz, DCl/CD 3 OD, pD ≈1.0) δ 9.09 (d, J =9.3, 1H, H-4), 8.63 (d, J =9.3, 1H, H-3), 8.28 (d, J =2.5, 1H, H-5), 8.26 (d, J =9.1, 1H, H-8), 8.21 (dd, J =8.8, 2.4, 1H, H-6′), 8.13 (dd, J =9.1, 2.5, 1H, H-7), 7.98 (d, J =2.4, 1H, H-2′), 7.13 (d, J =8.8, 1H, H-5′); 13 C NMR (100 MHz, DCl/ CD 3 OD, pD ≈1.0) δ 176.0 (C-2), 160.3 (C-4′), 154.9 (C-9), 152.8 (C-4), 148.7 (C-3′), 138.7 (C-7), 135.8 (C-6), 130.0 (C-5), 129.8 (C-6′), 125.6 (C-10), 121.9 (C-8), 121.3 (C-1′), 119.6 (C-3), 118.7 (C-5′), 117.3 (C-2′); FT-IR (ATR) ν max 3373, 3088, 1630, 1593, 1543, 1504, 1448, 1339, 1313, 1145, 1119, 1032, 908, 885, 817, 796, 784 cm −1 ; HRMS (ESITOF) m/z [M] + Calcd. for C 15 H 10 ClO 3 273.0318, found 273.014. 2.2.3. 4′-Hydroxy-6-chloroflavylium hydrogen sulfate (17) Method A was followed by using 2-hydroxy-5-chlorobenzaldehyde (3) (0.313 g, 2 mmol) and 4′-hydroxyacetophenone (7) (0.272 g, 2 mmol). Pure compound 17 was obtained as a red–orange solid (0.447 g, 63 % yield). Melting point: 235 ◦C (decomposes); 1 H NMR (400 MHz, DCl/CD 3 OD, pD ≈1.0) δ 9.19 (d, J =9.3 Hz, 1H, H-4), 8.71 (d, J =9.3 Hz, 1H, H-3), 8.67–8.57 (m, 2H, H-2′, H-6′), 8.40–8.26 (m, 2H, H-5, H-8), 8.16 (dd, J =9.1 Hz, J =2.4 Hz, 1H, H-7), 7.19 (m, 2H, H- 3′, H-5′); 13 C NMR (100 MHz, DCl/CD 3 OD, pD ≈1.0) δ 177.0 (C-2), 170.7 (C-4′), 155.2 (C-9), 153.8 (C-4), 139.1 (C-7), 136.6 (C-2′, C-6′), 136.1 (C-6), 130.3 (C-5), 125.8 (C-10), 122.1 (C-8), 121.0 (C-1′), 119.6 (C-3), 119.5 (C-3′, C-5′); FT-IR (ATR) ν max : 3153, 3089, 1610, 1581, 1539, 1500, 1450, 1375, 1331, 1311, 1269, 1176, 1157, 1137, 1126, 1116, 1049, 941, 856, 810, 688 cm −1 ; HRMS (ESI-TOF) m/z [M] + Calcd. for C 15 H 10 ClO 2 257.0369, found 257.0366. 2.2.4. 3′,4′-Dihydroxy-6-bromoflavylium hydrogen sulfate(18) Method A was followed by using 2-hydroxy-5-bromobenzaldehyde (4) (0.402 g, 2 mmol) and 3′,4′-dihydroxyacetophenone (6) (0.304 g, 2 mmol). Pure compound 18 was obtained as a red-brownish solid (0.744 g, 90 % yield). Melting point: 272 ◦C (decomposes); 1 H NMR (400 MHz, DCl/CD 3 OD, pD ≈1.0) δ 9.08 (d, J =9.3,1H, H-4), 8.64 (d, J =9.3, 1H, H-3), 8.45 (d, J =2.3, 1H, H-5), 8.27–8.19 (m, 3H, H-7, H-8, H-6′), 8.00 (d, J =2.3, 1H, H-2′), 7.15 (d, J =8.7, 1H, H-5′); 13 C NMR (100 MHz, DCl/CD 3 OD, pD ≈1.0) δ 175.6 (C-2), 160.3(C-4′), 155.3 (C- 9), 152.5(C-4), 148.8 (C-3′), 141.4 (C-7), 133.3 (C-5), 123.5 (C-6), 129.9 (C-6′), 126.0 (C-10), 121.9 (C-8), 121.3 (C-1′), 119.6 (C-3), 118.7 (C-5′), A. Alejo-Armijo et al. Bioorganic Chemistry 129 (2022) 106127 4 117.3 (C-2′); FT-IR (ATR) ν max : 3153, 3089, 1610, 1581, 1539, 1500, 1450, 1331, 1311, 1269, 1176, 1157, 1137, 1126, 1116, 1049, 941, 856, 810, 688 cm −1 ; HRMS (ESI-TOF) m/z [M] + Calcd. for C 15 H 10 BrO 3 316.9813, found 316.9809. 2.2.5. 4′-Hydroxy-6-bromoflavylium hydrogen sulfate (19) Method A was followed by using 2-hydroxy-6-bromobenzaldehyde (4) (0.402 g, 2 mmol) and 4′-hydroxyacetophenone (7) (0.272 g, 2 mmol). Pure compound 19 was obtained as a red–orange solid (0.545 g, 68 % yield). Melting point: 235 ◦C (decomposes); 1 H NMR (400 MHz, DCl/CD 3 OD, pD ≈1.0) δ 9.18 (d, J =9.2, 1H, H-4), 8.68 (d, J =9.2, 1H), 8.64–8.55 (m, 2H, H-2′, H-6′), 8.48 (d, J =2.2, 1H, H-5), 8.28 (dd, J =9.0, 2.2, 1H, H-7), 8.23 (d, J =9.0, 1H, H-8), 7.21 (d, J =8.9, 2H, H-3′, H-5′); 13 C NMR (100 MHz, DCl/CD 3 OD, pD ≈1.0) δ 175.4 (C-2), 170.3 (C-4′), 155.4 (C-9), 153.8 (C-4), 142.1 (C-7), 136.8 (C-2′, C-6′), 133.6 (C-5), 126.1 (C-10), 124.8 (C-6), 122.3 (C-8), 121.1 (C-1′), 119.6 (C-3, C-3′, C-5′); FT-IR (ATR) ν max : 3153, 3089, 1610, 1581, 1539, 1500, 1450, 1375, 1331, 1311, 1269, 1176, 1157, 1137, 1126, 1116, 1049, 941, 856, 810, 688 cm −1 ; HRMS (ESI-TOF) m/z [M] + Calcd. for C 15 H 10 BrO 2 300.9864, found 300.9859. 2.3. Method B for the synthesis of 3′,4′-dihydroxy-6-carboxyflavylium chloride (20) A mixture of 2-hydroxy-5-carboxybenzaldehyde (5) (0.332 g, 2 mmol) and 3′,4′-dihydroxyacetophenone (6) (0.304 g, 2 mmol) in absolute EtOH (40 mL) was saturated with dry HCl (g) for 1 h. The reaction mixture was stirred overnight following a similar method to that described by Kraus et al [51]. Then, the solvent was removed, Et 2 O (30 mL) was added and a solid precipitated. The solid was filtered off and carefully washed with Et 2 O and dried. Pure compound 20 was obtained as a brownish solid (0.449 g, 96 % yield). Melting point: 235 ◦C (decomposes); 1 H NMR (400 MHz, DCl/CD 3 OD, pD ≈1.0) δ 9.17 (d, J =8.5, 1H, H-4), 8.79 (d, J =1.2, 1H, H-5), 8.64 (br d, J =8.5, 2H, H-3, H-7), 8.31 (d, J =8.5, 1H, H-8), 8.25 (br d, J =8.7, 1H, H-6′), 8.02 (br s, 1H, H-2′), 7.15 (d, J =8.7, 1H, H-5′); This compound is not soluble and stable enough to record a 13 C NMR spectrum; FT-IR (ATR) ν max : 3500, 1712, 1596, 1552, 1506, 1458, 1033 cm −1 ; HRMS (ESI-TOF) m/z [M] + Calcd. for C 16 H 11 O 5 283.0606, found 283.0602. 2.4. General method C for the synthesis of 2,8-dioxabicyclo[3.3.1]nonanes (23–42) A mixture of flavylium salt (10–20) (0.5 mmol) and phloroglucinol (21) or resorcinol (22) (0.5 mmol) in absolute MeOH (8 mL) (or absolute THF for compounds 34 and 42) was stirred overnight at 50 ◦C following a similar method to that described by Kraus et al. [54] and used previously by us [38,39]. Then, the solvent was removed and the crude purified by semipreparative HPLC or by CC using Sephadex LH-20 or Silica gel 60. The dioxabicyclic derivatives 23 (0.069 g, 34 % yield from the starting aldehyde 1), 25 (0.170 g, 64 % from 2) and 27 (0.170 g, 63 % from 2) were described previously with similar yields and their structures confirmed by comparison of their spectral data with those reported in the literature: 23 [54], 25 [54] and 27 [38]. Analytical HPLC (λ =280 nm): compound 23 (purity: 98 %; t R =15.9 min); compound 25 (purity: 97 %; t R =21.2 min); compound 27 (purity: 94 %; t R =17.9 min). 2.4.1. 2-(4′-Hydroxyphenyl)chromane-(4 → 4,2 → O-5)-phloroglucinol (24) Method C was followed by using the flavylium salt 11 (0.140 g) and phloroglucinol (21, 0.063 g, 0.5 mmol). Then, the solvent was removed and the crude purified by semipreparative HPLC. Purification eluting with MeOH-H 2 O (50:50) yielded pure analogue 24 as a white amorphous solid (0.032 g, 22 % from 1). Melting point: 155 ◦C (decomposes); 1 H NMR (400 MHz, CD 3 OD) δ 7.49 (m, 2H, H-2′(B), H-6′(B)), 7.38 (dd, J =7.5, 1.7, 1H, H-5(A)), 7.07 (ddd, J =8.0, 7.5, 1.7, 1H, H-7(A)), 6.89 (dd, J =8.0, 1.1, 1H, H-8(A)), 6.82 (m, 3H, H-6(A), H-3′(B), H-5′(B)), 5.93 (m, 2H, H-2′′(D), H-6′′(D)), 4.35 (t, J =3.1, 1H, H-4(C)), 2.20 (d, J =3.1, 2H, H-3(C)); 13 C NMR (100 MHz, CD 3 OD) δ 159.1 (C-4′(B)), 158.3* (C-1′′(D)), 156.4* (C-5′′(D)), 154.9* (C-3′′(D)), 153.9 (C-9(A)), 134.8 (C-1′(B)), 129.5 (C-10(A)),129.1 (C-5(A)), 128.6 (C-7(A)),128.4 (C-2′(B), C-6′(B)), 122.2 (C-6(A)), 117.1 (C-8(A)), 116.1 (C-3′(B), C-5′ (B)), 107.6 (C-4′′(D)), 100.3 (C-2(C)), 97.1 # (C-2′′(D)), 96.0 # (C-6′′(D)), 35.1 (C-3(C)), 28.4 (C-4(C)) ( *,# these signals could be interchanged); FT- IR (ATR) v max : 3300, 2960, 2931, 1606, 1514, 1475, 1437, 1336, 1300, 1230, 1171, 1117, 1062, 1009, 885, 811, 783 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 16 O 5 348.0998, found 348.1002. Analytical HPLC (λ =280 nm): purity: 93 %; t R =18.1 min. 2.4.2. 2-(4′-Hydroxyphenyl)-6-nitrochromane-(4 → 4,2 → O-5)- phloroglucinol (26) Method C was followed by using the flavylium salt 13 (0.195 g) and phloroglucinol (21, 0.063 g, 0.5 mmol). Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM- EtOH (97:3) yielded pure analogue 26 as a white amorphous solid (0.088 g, 45 % from 2). Melting point: 170 ◦C (decomposes); 1 H NMR (400 MHz, CD 3 OD) δ 8.29 (d, J =2.8, 1H, H-5(A)), 7.99 (dd, J =9.1, 2.8, 1H, H-7(A)), 7.50 (d, J =8.9, 2H, H-2′(B), H-6′(B)), 7.02 (d, J =9.1, 1H, H-8(A)), 6.84 (d, J =8.9, 2H, H-3′(B), H-5′(B)), 5.96(br s, 2H, H-2′′(D), H-6′′(D)), 4.47 (t, J =3.0, 1H, H-4(C)), 2.30 (m, 2H, H-3(C)); 13 C NMR (100 MHz, CD 3 OD) δ 159.6 (C-9(A)), 159.3 (C-4′(B)), 158.7 (C-3′′(D)), 156.4 (C-1′′(D)), 154.1 (C-5′′(D)), 142.8 (C-6(A)), 133.4 (C-1′(B)), 130.4 (C-10(A)), 128.2 (C-2′(B), C-6′(B)), 124.6 (C-5(A), C-7(A)), 117.7 (C-8 (A)), 116.1 (C-3′(B), C-5′(B)), 106.0 (C-4′′(D)), 101.2 (C-2(C)), 97.24* (C-2′′(D)), 96.0* (C-6′′(D)), 33.9 (C-3(C)), 28.2 (C-4(C)) (*these signals could be interchanged); FT-IR (ATR) v max : 3300, 2945, 2912, 1602, 1508, 1475, 1330, 1242, 1170, 1124, 1085, 1064, 1006, 970, 889, 817, 746 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 14 NO 7 392.0776, found 392.0780. Analytical HPLC (λ =280 nm): purity: 98 %; t R =19.4 min. 2.4.3. 2-(4′-Hydroxyphenyl)-3-methyl-6-nitrochromane-(4 → 4,2 → O-5)- phloroglucinol (28) Method C was followed by using the flavylium salt 15 (0.200 g) and phloroglucinol (21, 0.063 g, 0.5 mmol). Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM- EtOH (97:3) yielded pure analogue 28 as a white amorphous solid (0.102 g, 50 % from 2). Melting point: 280 ◦C (decomposes); 1 H NMR (400 MHz, CD 3 OD) δ 8.27 (d, J =2.7, 1H, H-5(A)), 7.98 (dd, J =8.9, 2.7, 1H, H-7(A)), 7.45 (d, J =8.9, 2H, H-2′(B), H-6′(B)), 6.96 (d, J =8.9, 1H, H-8(A)), 6.85 (d, J =8.9, 2H, H-3′(B), H-5′(B)), 6.02* (d, J =2.3, 1H, H- 6′′(D)), 5.99* (d, J =2.3, 1H, H-2′′(D)), 4.23 (d, J =2.3, 1H, H-4(C)), 2.42 (dq, J =6.4, 2.3, 1H, H-3(C)), 0.77 (d, J =6.4, 2H, CH 3 ); 13 C NMR (100 MHz, CD 3 OD) δ 159.6 (C-9(A)), 159.1 (C-3′′(D)), 158.6 (C-4′(B)), 157.1 (C-1′′(D)), 153.1 (C-5′′(D)), 142.5 (C-6(A)), 131.9 (C-10(A)), 131.7 (C-1′(B)), 128.7 (C-2′(B), C-6′(B)), 124.4 (C-7(A)), 124.0 (C-5(A)), 117.2 (C-8(A)), 115.8 (C-3′(B), C-5′(B)), 104.0 (C-2(C)), 103.1 (C-4′′ (D)), 97.5 (C-2′′(D)), 95.6 (C-6′′(D)), 35.5 (C-3(C)), 34.3 (C-4(C)), 13.8 (CH 3 ) (*these signals could be interchanged); FT-IR (ATR) v max : 3396, 3168, 1608, 1502, 1477, 1330, 1247, 1209, 1168, 1132, 1083, 1051, 1002, 916, 821, 748, 692 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 22 H 16 NO 7 406.0932, found 406.0933. Analytical HPLC (λ =280 nm): purity: 98 %; t R =19.8 min. 2.4.4. 2-(3′,4′-Dihydroxyphenyl)-6-chlorochromane-(4 → 4,2 → O-5)- phloroglucinol (29) Method C was followed by using the flavylium salt 16 (0.185 g, 0.5 mmol) and phloroglucinol (21, 0.063 g, 0.5 mmol). Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM-EtOH (97:3) yielded pure analogue 29 as a colorless foam (0.084 g, 43 % from 3). 1 H NMR (400 MHz, CD 3 OD) δ 7.34 (d, J =2.7, A. Alejo-Armijo et al. Bioorganic Chemistry 129 (2022) 106127 5 1H, H-5(A)), 7.10 (d, J =2.3, 1H, H-2′(B)), 7.03 (dd, J =8.7, 2.7, 1H, H- 7(A)), 6.98 (dd, J =8.3, 2.3, 1H, H-6′(B)), 6.86 (d, J =8.7, 1H, H-8(A)), 6.80 (d, J =8.3, 1H, H-5′(B)), 5.95* (d, J =2.1, 1H, H-6′′(D)), 5.94* (d, J =2.1, 1H, H-2′′(D)), 4.32 (t, J =3.1, 1H, H-4(C)), 2.21 # (dd, J =13.4, 3.1, 1H, H-3b (C)), 2.18 # (dd, J =13.4, 3.1, 1H, H-3a (C)); 13 C NMR (100 MHz, CD 3 OD) δ 158.6 ǂ (C-1′′(D)), 156.4 ǂ (C-5′′(D)), 154.7 (C-3′′ (D)), 152.8 (C-9(A)), 147.1 (C-4′(B)), 146.3 (C-3′(B)), 135.0 (C-1′(B)), 131.4 (C-10 (A)), 128.6 (C-5 (A)), 128.3 (C-7(A)), 126.7 (C-6(A)), 118.6 (C-8(A), C-6′(B)), 116.2 (C-5′(B)), 114.5 (C-2′(B)), 106.8 (C-4′′(D)), 100.4 (C-2(C)), 97.2 (C-6′′(D)), 96.1 (C-2′′(D)), 34.7 (C-3(C)), 28.4 (C-4 (C)) ( *,#,ǂ these signals could be interchanged); FT-IR (ATR) v max : 3549, 3474, 3414, 2925, 2856, 1636, 1617, 1520, 1480, 1338, 1301, 1234, 1119, 1058, 817, 619 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 15 ClO 6 397.0484, found 397.0487. Analytical HPLC (λ =280 nm): purity: 99 %; t R =18.8 min. 2.4.5. 2-(4′-Hydroxyphenyl)-6-chlorochromane-(4 → 4,2 → O-5)- phloroglucinol (30) Method C was followed by using the flavylium salt 17 (0.171 g) and phloroglucinol (21, 0.063 g, 0.5 mmol). Then, the solvent was removed and the crude purified by semipreparative HPLC. Purification eluting with MeOH-H 2 O (60:40) yielded pure analogue 30 as a colorless foam (0.077 g, 42 % from 3). 1 H NMR (400 MHz, CD 3 OD) δ 7.52–7.43 (m, 2H, H-2′(B), H-6′(B)), 7.35 (d, J =2.6, 1H, H-5(A)), 7.04 (dd, J =8.6, 2.6, 1H, H-7(A)), 6.87 (d, J =8.6, 1H, H-8(A)), 6.85–6.78 (m, 2H, H-3′(B), H- 5′(B)), 5.95* (d, J =2.2, 1H, H-2′′(D)), 5.93* (d, J =2.2, 1H, H-6′′(D)), 4.33 (t, J =3.1, 1H, H-4(C)), 2.21 (dd, J =13.4, 3.1, 2H, H-3(C)); 13 C NMR (100 MHz, CD 3 OD) δ 159.3 (C-4′(B)), 158.7 # (C-1′′(D)), 156.5 # (C- 5′′(D)), 154.8 # (C-3′′(D)), 152.8 (C-9(A)), 134.4 (C-1′(B)), 131.4 (C-10 (A)), 128.6 (C-5 (A)), 128.3 (C-7(A), C-2′(B), C-6′(B)), 126.8 (C-6(A)), 118.7 (C-8(A)), 116.2 (C-3′(B), C-5′(B)), 106.8 (C-4′′(D)), 100.5 (C-2 (C)), 97.2 (C-6′′(D)), 96.1 (C-2′′(D)), 34.6 (C-3(C)), 28.4 (C-4(C)) ( *, # these signals could be interchanged); FT-IR (ATR) v max : 3300, 2960, 2931, 1606, 1514, 1475, 1437, 1336, 1300, 1230, 1171, 1117, 1062, 1009, 885, 811, 783 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 15 ClO 5 382.0608, found 382.0609. Analytical HPLC (λ =280 nm): purity: 96 %; t R =20.7 min. 2.4.6. 2-(3′,4′-Dihydroxyphenyl)-6-bromochromane-(4 → 4,2 → O-5)- phloroglucinol (31) Method C was followed by using the flavylium salt 18 (0.208 g, 0.5 mmol) and phloroglucinol (21, 0.063 g, 0.5 mmol). Then, the solvent was removed and the crude purified by semipreparative HPLC. Purification eluting with MeOH-H 2 O (60:40) yielded pure analogue 31 as a brown syrup (0.096 g, 41 % from 4). 1 H NMR (400 MHz, CD 3 OD) δ 7.48 (d, J =2.5, 1H, H-5(A)), 7.17 (dd, J =8.7, 2.5, 1H, H-7(A)), 7.11 (d, J =2.1, 1H, H-2′(B)), 6.98 (dd, J =8.3, 2.1, 1H, H-6′(B)), 6.82 (d, J =8.7, 1H, H-8(A)),6.80 (d, J =8.3, 1H, H-5′(B)), 5.95* (d, J =2.3, 1H, H-6′′(D)), 5.94* (d, J =2.3, 1H, H-2′′(D)), 4.31 (t, J =3.0, 1H, H-4(C)), 2.21 # (dd, J =13.4, 3.0, 1H, H-3β (C)), 2.17 # (dd, J =13.4, 3.0, 1H, H-3 α (C)). 13 C NMR (100 MHz, CD 3 OD) δ 158.6 (C-1′′(D)), 156.4 (C-5′′(D)), 154.7 (C-3′′(D)), 153.3 (C-9(A)), 147.1 (C-4′(B)), 146.3 (C-3′(B)), 135.0 (C-1′(B)), 131.9 (C-10 (A)), 131.5 (C-5 (A)), 131.3 (C-7(A)), 119.1 (C-8 (A)), 118.6 (C-6′(B)), 116.2 (C-5′(B)), 114.5 (C-2′(B)), 114.0 (C-6(A)), 106.8 (C-4′′(D)), 100.4 (C-2(C)), 97.2 (C-6′′(D)), 96.1 (C-2′′(D)), 34.7 (C- 3(C)), 28.4 (C-4(C)). ( *,# these signals could be interchanged); FT-IR (ATR) v max : 3300, 1606, 1514, 1475, 1436, 1336, 1299, 1230, 1116, 1062, 1008, 885, 811 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 15 BrO 6 440.9979, found 440.9980. Analytical HPLC (λ =280 nm): purity: 99 %; t R =19.3 min. 2.4.7. 2-(4′-Hydroxyphenyl)-6-bromochromane-(4 → 4,2 → O-5)- phloroglucinol (32) Method C was followed by using the flavylium salt 19 (0.191 g) and phloroglucinol (21, 0.063 mg, 0.5 mmol). Then, the solvent was removed and the crude purified by semipreparative HPLC. Purification eluting with MeOH-H 2 O (60:40) yielded pure analogue 32 as a white foam (0.100 g, 49 % from 4). 1 H NMR (400 MHz, CD 3 OD) δ 7.67–7.32 (m, 3H, H-5(A), H-2′(B), H-6′(B)), 7.14 (dd, J =8.7, 2.5, 1H, H-7(A)), 6.92–6.65 (m, 3H, H-8(A), H-3′(B), H-5′(B)), 5.92* (d, J =2.3, 1H, H-6′′ (D)), 5.91* (d, J =2.3, 1H, H-2′′(D)), 4.29 (t, J =3.1, 1H, H-4(C)), 2.17 (dd, J =13.4, 3.1, 2H, H-3(C)); 13 C NMR (100 MHz, CD 3 OD) δ 159.2 (C- 4′(B)), 158.6 # (C-1′′(D)), 156.4 # (C-5′′(D)), 154.7 # (C-3′′(D)), 153.3 (C-9 (A)), 134.4 (C-1′(B)), 131.9 (C-10 (A)), 131.5 (C-5 (A)), 131.3 (C-7(A)), 128.4 (C-2′(B), C-6′(B)), 119.1 (C-8(A)), 116.1 (C-3′(B), C-5′(B)), 114.0 (C-6(A)), 106.8 (C-4′′(D)), 100.5 (C-2(C)), 97.2 (C-6′′(D)), 96.1 (C-2′′ (D)), 34.6 (C-3(C)), 28.3 (C-4(C)) ( *,# these signals could be interchanged); FT-IR (ATR) v max : 3300, 2960, 2931, 1606, 1514, 1475, 1437, 1336, 1300, 1230, 1171, 1117, 1062, 1009, 885, 811, 783 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 15 BrO 5 426.0103, found 426.0104. Analytical HPLC (λ =280 nm): purity: 96 %; t R =21.1 min. 2.4.8. 2-(3′,4′-Dihydroxyphenyl)-6-carboxymethylchromane- (4 → 4,2 → O-5)-phloroglucinol (33) and 2-(3′,4′-dihydroxyphenyl)-6- carboxychromane-(4 → 4, 2 → O-5)-phloroglucinol (34) Method C was followed by using the flavylium salt 20 (0.300 g, 0.94 mmol) and phloroglucinol (21, 0.200 g, 1.58 mmol). Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM-MeOH (97:3) yielded pure analogue 33 as white foam (0.241 g, 58 % from 5) and 34 as a brownish syrup (0.036 mg, 9 % from 5). When THF was used as solvent instead of MeOH, the pure analogue 34 was formed as the only product (0.204 g, 51 % from 5).Compound 33: 1 H NMR (400 MHz, CD 3 OD) δ 8.11 (d, J =2.2, 1H, H-5(A)), 7.79 (dd, J =8.5, 2.2, 1H, H-7(A)), 7.15 (d, J =2.2, 1H, H-2′(B)), 7.03 (dd, J =8.3, 2.2, 1H, H-6′(B)), 6.98 (d, J =8.5, 1H, H-8(A)), 6.84 (d, J =8.3, 1H, H-5′ (B)), 5.97 (br s, 2H, H-2′′(D), H-6′′(D)), 4.44 (t, J =3.1, 1H, H-4(C)), 3.22 (s, 3H, COOCH 3 ), 2.32–2.22 (m, 2H, H-3(C)). 13 C NMR (100 MHz, CD 3 OD) δ 168.6 (COOCH 3 ), 158.3* (C-1′′D), 158.2 (C-9(A)), 156.3* (C- 5′′(D)), 154.3 (C-3′′(D)), 146.9 (C-4′(B)), 146.1 (C-3′(B)), 134.9 (C-1′ (B)), 130.7 (C-5(A)), 130.3 (C-7(A)), 129.6 (C-10 (A)), 123.8 (C-6(A)), 118.3 (C-6′(B)), 117.1 (C-8(A)), 116.0 (C-5′(B)), 114.2 (C-2′(B)), 106.6 (C-4′′(D)), 100.5 (C-2(C)), 97.0 # (C-6′′(D)), 95.8 # (C-2′′(D)), 52.4 (COOCH 3 ), 34.5 (C-3(C)), 28.1 (C-4(C)).( *,# these signals could be interchanged); FT-IR (ATR) v max : 3295, 2923, 2850, 1687, 1608, 1515, 1438, 1274, 1195, 1170, 1114, 1064, 1010, 899, 819, 767, 719 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 23 H 18 O 8 421.0929, found 421.0934. Analytical HPLC (λ =280 nm): purity: 95 %; t R =25.1 min. Compound 34: 1 H NMR (400 MHz, CD 3 OD) δ 8.10 (d, J =2.1, 1H, H-5 (A)), 7.78 (dd, J =8.5, 2.1, 1H, H-7(A)), 7.12 (d, J =2.2, 1H, H-2′(B)), 7.00 (dd, J =8.3, 2.2, 1H, H-6′(B)), 6.96 (d,J =8.5,1H, H-8(A)), 6.81 (d, J =8.3, 1H, H-5′(B)), 5.95 (br s, 1H, H-2′′(D)), 5.95 (br s, 1H, H-6′′(D)), 4.42 (t, J =2.9, 1H, H-4(C)), 2.27 (dd, J =13.3, 2.9, 1H, H-3β (C)), 2.23 (dd, J =13.3, 2.9, 1H, H-3 α (C)). 13 C NMR (100 MHz, CD 3 OD) δ 170.2 (COOH), 158.6* (C-1′′(D)), 158.3 (C-9(A)), 156.5* (C-5′′(D)), 154.5 (C- 3′′(D)), 147.1 (C-4′(B)), 146.3 (C-3′(B)), 134.9 (C-1′(B)), 131.3 (C-5 (A)), 130.8 (C-7(A)), 129.7 (C-10 (A)), 124.6 (C-6(A)), 118.6 (C-6′(B)),117.2 (C-8(A)), 116.2 (C-5′(B)), 114.5 (C-2′(B)), 106.9 (C-4′′(D)), 100.8 (C-2 (C)), 97.3 # (C-6′′(D)), 96.0 # (C-2′′(D)), 34.8 (C-3(C)), 28.4 (C-4(C)) ( *, # these signals could be interchanged); FT-IR (ATR) v max : 3309, 1685, 1608, 1515, 1438, 1184, 1109, 1066 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 22 H 16 O 8 407.0772, found 407.0774. Analytical HPLC (λ =280 nm): purity: 99 %; t R =13.9 min. 2.4.9. 2-(3′,4′-Dihydroxyphenyl)-6-nitrochromane-(4 → 4,2 → O-3)- resorcinol (35) Method C was followed by using the flavylium salt 12 (0.191 g) and resorcinol (22, 0.059 g, 0.5 mmol). Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM-EtOH (97:3) yielded pure analogue 35 as a white amorphous solid (0.098 g, 50 % from 2). Melting point: 145 ◦C (decomposes); 1 H NMR (400 MHz, CD 3 OD) δ 8.22 (d, J =2.7, 1H, H-5(A)), 8.00 (dd, J =9.0, 2.7, 1H, H-7 (A)), 7.15 (d, J =9.1, 1H, H-5′′(D)), 7.13 (d, J =2.3, 1H, H-2′(B)), 7.03 A. Alejo-Armijo et al. Bioorganic Chemistry 129 (2022) 106127 6 (d, J =9.0, 1H, H-8(A)), 7.00 (dd, J =8.3, 2.3, 1H, H-6′(B)), 6.82 (d, J =8.3, 1H, H-5′(B)), 6.40 (m, 2H, H-2′′(D), H-6′′(D)), 4.17 (t, J =3.1, 1H, H-4(C)), 2.35 (dd, J =13.4, 3.1, 2H, H-3(C)); 13 C NMR (100 MHz, CD 3 OD) δ 159.2 (C-1′′(D)), 159.1 (C-9(A)), 153.3 (C-3′′(D)), 147.2 (C-4′ (B)), 146.2 (C-3′(B)), 143.2 (C-6(A)), 133.8 (C-1′(B)), 130.3 (C-10(A)), 129.4 (C-5′′(D)), 128.1 (C-8(A)), 124.9 (C-7(A)), 124.3 (C-5(A)), 118.4 (C-6′(B)), 118.2 (C-4′′(D)), 116.1 (C-5′(B)), 114.3 (C-2′(B)), 110.7* (C-6′′ (D)), 104.4* (C-2′′(D)), 100.9 (C-2(C)), 34.4 (C-4(C)), 34.2 (C-3(C)) (*these signals could be interchanged); FT-IR (ATR) v max : 3296, 2912, 1595, 1477, 1433, 1330, 1292, 1249, 1147, 1107, 1083, 1016, 966, 810, 784, 771, 746, 688 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 14 NO 7 392.0776, found 392.0777. Analytical HPLC (λ =280 nm): purity: 94 %; t R =17.8 min. 2.4.10. 2-(4′-Hydroxyphenyl)-6-nitrochromane-(4 → 4,2 → O-3)- resorcinol (36) Method C was followed by using the flavylium salt 13 (0.195 g) and resorcinol (22, 0.059 g, 0.5 mmol). Then, the solvent was removed and the crude purified by crystallization in a mixture of Et 2 O:hexane yielding pure analogue 36 as a white amorphous solid (0.121 g, 64 % from 2). Melting point: 85 ◦C (decomposes); 1 H NMR (400 MHz, CD 3 OD) δ 8.21 (d, J =2.7, 1H, H-5(A)), 7.97 (dd, J =8.9, 2.7, 1H, H-7(A)), 7.50 (d, J =8.9, 2H, H-2′(B), H-6′(B)), 7.13 (d, J =9.1, 1H, H-5′′(D)), 7.01 (d, J =8.9, 1H, H-8(A)), 6.84 (d, J =8.9, 2H, H-3′(B), H-5′(B)), 6.41(br s,1H, H-2′′(D)), 6.40 (dd overlapped, J =9.1, 2.5, 1H, H-6′′(D)), 4.14 (t, J =3.1, 1H, H-4(C)), 2.33 (dd, J =13.4, 3.1, 2H, H-3(C)); 13 C NMR (100 MHz, CD 3 OD) δ 159.3* (C-9(A)), 159.1* (C-4′(B)), 159.0 (C-1′′(D)), 153.6 (C-3′′(D)), 143.1 (C-6(A)), 133.1 (C-1′(B)), 130.2 (C-10(A)), 129.4 (C-5′′(D)), 128.2 (C-2′(B), C-6′(B)), 124.9 (C-7(A)), 124.2 (C-5(A)), 118.1 (C-8(A), C-4′′(D)), 116.1 (C-3′(B), C-5′(B)), 110.7 (C-6′′(D)), 104.4 (C-2′′(D)), 101.0 (C-2(C)), 34.4 (C-4(C)), 34.0 (C-3(C)) (*these signals could be interchanged); FT-IR (ATR) v max : 3317, 2916, 2842, 1585, 1502, 1477, 1330, 1247, 1147, 1107, 1083, 102, 999, 958, 887, 823, 746, 688, 638 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 14 NO 6 376.0827, found 376.0827. Analytical HPLC (λ =280 nm): purity: 95 %; t R =19.7 min. 2.4.11. 2-(3′,4′-Dihydroxyphenyl)-3-methyl-6-nitrochromane- (4 → 4,2 → O-3)-resorcinol (37) Method C was followed by using the flavylium salt 14 (0.200 g) and resorcinol (22, 0.059 g, 0.5 mmol). Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM-EtOH (97:3) yielded pure analogue 37 as a white amorphous solid (0.122 g, 60 % from 2). Melting point: 182 ◦C (decomposes); 1 H NMR (400 MHz, CD 3 OD) δ 8.22 (d, J =2.7, 1H, H-5(A)), 8.01 (dd, J =9.0, 2.7, 1H, H-7 (A)), 7.15 (d, J =8.1, 1H, H-5′′(D)), 7.08 (d, J =2.3, 1H, H-2′(B)), 6.97 (m, 2H, H-8(A), H-6′(B)), 6.84 (d, J =8.3, 1H, H-5′(B)), 6.46 (d, J =2.3, 1H, H-2′′(D)), 6.44 (dd, J =8.1, 2.3, 1H, H-6′′(D)), 3.95 (d, J =2.5, 1H, H-4(C)), 2.45 (dd, J =13.4, 2.5, 2H, H-3(C)), 0.78 (d, J =6.8, 3H, CH 3 ); 13 C NMR (100 MHz, CD 3 OD) δ 159.2 (C-9(A)), 158.9 (C-1′′(D)), 152.6 (C-3′′(D)), 147.0 (C-4′(B)), 146.0 (C-3′(B)), 142.8 (C-6(A)), 132.0 (C-1′ (B), C-10(A)), 130.5 (C-5′′(D)), 124.8 (C-7(A)), 123.6 (C-5(A)), 119.0 (C- 6′(B)), 117.5 (C-8(A)), 115.8 (C-5′(B)), 115.3 (C-4′′(D)), 114.7 (C-2′(B)), 111.0 (C-6′′(D)), 104.0 (C-2′′(D)), 103.8 (C-2(C)), 40.5 (C-4(C)), 35.6 (C- 3(C)), 13.8 (CH 3 ); FT-IR (ATR) v max : 3346, 2960, 2910, 1591, 1502, 1477, 1434, 1330, 1247, 1143, 1101, 1008, 974, 916, 889, 823, 779, 744, 688 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 22 H 16 NO 7 406.0932, found 406.0937. Analytical HPLC (λ =280 nm): purity: 94 %; t R =18.3 min. 2.4.12. 2-(4′-Hydroxyphenyl)-3-methyl-6-nitrochromane-(4 → 4,2 → O- 3)-resorcinol (38) Method C was followed by using the flavylium salt 15 (0.200 g) and resorcinol (22, 0.059 g, 0.5 mmol) Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM-MeOH (98:2) yielded pure analogue 38 as a white amorphous solid (0.131 g, 64 % from 2). Melting point: 250 ◦C (decomposes); 1 H NMR (400 MHz, CD 3 OD) δ 8.28 (d, J =2.8, 1H, H-5(A)), 8.06 (dd, J =9.0, 2.8, 1H, H-7 (A)), 7.53–7.49 (m, 2H, H-2′(B)), 7.20 (d, J =8.2, 1H, H-5′′(D)), 7.03 (d, J =9.0, 1H, H-8(A)), 6.92–6.87 (m, 2H, H-3′′(D)), 6.51 (d, J =2.4, 1H, H-2′′(D)), 6.48 (dd, J =8.2, 2.4, 1H, H-6′′(D)), 4.03 (d, J =2.5, 1H, H-4 (C)), 2.54 (qd, J =6.9, 2.5, 1H, H-3(C)), 0.82 (d, J =6.9, 3H, CH 3 ). 13 C NMR (100 MHz, CD 3 OD) δ 159.5 (C-4′(B), C-9(A)), 159.3 (C-1′′(D)), 152.9 (C-3′′(D)), 143.2 (C-6(A)), 132.3 (C-10(A)), 131.7 (C-1′(B)), 130.9 (C-5′′(D)), 129.0 (C-2′(B)), 125.1 (C-7(A)), 123.9 (C-5(A)), 117.9 (C-8 (A)), 116.1 (C-3′(B)), 115.6 (C-4′′(D)), 111.4 (C-6′′(D)), 104.4 (C-2′′(D)), 104.2 (C-2(C)), 40.8 (C-4(C)), 35.9 (C-3(C)), 14.09 (CH 3 ); FT-IR (ATR) v max : 3396, 3170; 2916, 2842, 1585, 1502, 1477, 1330, 1247, 1147, 1107, 1083, 102, 999, 958, 887, 823, 746, 688, 638 cm −1 ; HRMS (ESITOF) m/z [M−H] - Calcd. for C 22 H 16 NO 6 390.0983, found 390.0982. Analytical HPLC (λ =280 nm): purity: 98 %; t R =20.2 min. 2.4.13. 2-(3′,4′-Dihydroxyphenyl)-6-chlorochromane-(4 → 4,2 → O-3)- resorcinol (39) Method C was followed by using the flavylium salt 16 (0.185 g) and resorcinol (22, 0.055 g, 0.5 mmol). Then, the solvent was removed and the crude purified by semipreparative HPLC. Purification eluting with MeOH-H 2 O (60:40) yielded pure analogue 39 as a brown syrup (0.065 g, 36 % from 3). 1 H NMR (400 MHz, CD 3 OD) δ 7.27 (d, J =2.5, 1H, H-5 (A)), 7.11 (d, J =2.1, 1H, H-2′(B)), 7.09–7.04 (m, 2H, H-7(A), H-5′′(D)), 6.99 (dd, J =8.3, 2.1, 1H, H-6′(B)), 6.87 (d, J =8.7,1H, H-8(A)), 6.81 (d, J =8.3, 1H, H-5′(B)), 6.38–6.36 (m, 2H, H-2′′(D), H-5′′(D)), 3.99 (t, J =3.0,1H, H-4(C)), 2.28* (dd, J =13.5, 3.0, 1H, H-3b (C)), 2.25* (dd, J =13.5, 3.0, 1H, H-3a (C)); 13 C NMR (100 MHz, CD 3 OD) δ 158.9 (C-1′′ (D)), 154.2 (C-3′′(D)), 152.4 (C-9(A)), 147.1 (C-4′(B)), 146.3 (C-3′(B)), 134.8 (C-1′(B)), 131.1 (C-10 (A)), 129.3 (C-5′′(D)), 128.2 (C-5 (A)), 128.8 (C-7(A)), 127.1 (C-6(A)), 119.0 (C-8(A)), 118.8 (C-6′(B)), 116.2 (C-5′(B)), 114.5 (C-2′(B)), 114.1 (C-4′′(D)), 110.4 (C-6′′(D)), 104.4 (C-2′′ (D)), 100.4 (C-2(C)), 34.8 (C-3(C), C-4(C)) (*these signals could be interchanged); FT-IR (ATR) v max : 3319, 1606, 1508, 1479, 1334, 1249, 1126, 1085, 1008, 813 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 15 ClO 5 381.0535, found 381.0538. Analytical HPLC (λ =280 nm): purity: 98 %; t R =20.6 min. 2.4.14. 2-(3′,4′-Dihydroxyphenyl)-6-bromochromane-(4 → 4,2 → O-3)- resorcinol (40) Method C was followed by using the flavylium salt 18 (0.208 g) and resorcinol (22, 0.055 g, 0.5 mmol). Then, the solvent was removed and the crude purified by semipreparative HPL. Purification eluting with MeOH-H 2 O (60:40) yielded pure analogue 40 as a brown reddish syrup (0.054 g, 35 % from 4). 1 H NMR (400 MHz, CD 3 OD) δ 7.40 (d, J =2.5, 1H, H-5(A)), 7.18 (dd, J =8.7, 2.5, 1H, H-7(A)), 7.11 (d, J =2.1, 1H, H- 2′(B)), 7.07 (d, J =8.9, 1H, H-5′′(D)), 6.98 (dd, J =8.3, 2.1, 1H, H-6′(B)), 6.82 (d, J =8.7, 1H, H-8(A)), 6.80 (d, J =8.3, 1H, H-5′(B)), 6.38–6.36 (m, 2H, H-2′′(D), H-6′′(D)), 3.99–3.97 (m, 1H, H-4(C)), 2.27* (dd, J =13.7, 2.9, 1H, H-3b (C)), 2.23* (dd, J =13.7, 2.9, 1H, H-3a (C)); 13 C NMR (100 MHz, CD 3 OD) δ 158.9 (C-1′′(D)), 154.2 (C-3′′(D)), 152.9 (C-9 (A)), 147.1 (C-4′(B)), 146.3 (C-3′(B)), 134.7 (C-1′(B)), 131.7 (C-10 (A)), 131.8 (C-7(A)), 131.1 (C-5 (A)), 129.4 (C-5′′(D)), 119.5 (C-8(A)), 119.0 (C-4′′(D)), 118.6 (C-6′(B)), 116.2 (C-5′(B)), 114.5 (C-2′(B)), 114.3 (C-6 (A)), 100.3 (C-2(C)), 110.4 (C-6′′(D)), 104.4 (C-2′′(D)), 34.8 (C-3(C)), 34.7 (C-4(C)) (*these signals could be interchanged); FT-IR (ATR) v max : 3300, 1606, 1514, 1475, 1116, 1062, 1008, 811 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 21 H 15 BrO 5 425.003, found 425.0031. Analytical HPLC (λ =280 nm): purity: 96 %; t R =25.8 min. 2.4.15. 2-(3′,4′-Dihydroxyphenyl)-6-carboxymethylchromane- (4 → 4,2 → O-3)-resorcinol (41) and 2-(3′,4′-dihydroxyphenyl)-6- carboxychromane-(4 → 4,2 → O-3)-resorcinol (42) Method C was followed by using the flavylium salt 20 (0.290 g) and resorcinol (22, 0.090 g, 0.82 mmol). Then, the solvent was removed and the crude purified by silica gel CC. Purification eluting with DCM-MeOH A. Alejo-Armijo et al. Bioorganic Chemistry 129 (2022) 106127 7 (97:3) yielded pure analogue 41 as white foam (0.169 g, 44 % from 5) and 42 as a brownish syrup (0.026 g, 7 % from 5). When THF was used as solvent instead of MeOH, the pure analogue 42 was formed as the sole product (0.087 g, 30 % from 5). Compound 41: 1 H NMR (400 MHz, CD 3 OD) δ 7.98 (d, J =2.1, 1H, H-5(A)), 7.80 (dd, J =8.5, 2.1, 1H, H-7 (A)), 7.15 (d, J =2.3, 1H, H-2′(B)), 7.13 (d, J =7.9, 1H, H-5′′(D)), 7.03 (dd, J =8.3, 2.3, 1H, H-6′(B)), 7.00 (d, J =8.5, 1H, H-8(A)), 6.85 (d, J =8.3, 1H, H-5′(B)), 6.43–6.38 (m, 2H, H-2′′(D), H-6′′(D)), 4.12 (t, J =2.9, 1H, H-4(C)), 3.88 (s, 3H, COOCH 3 ), 2.38–2.30 (m, 2H, H-3(C)); 13 C NMR (100 MHz, CD 3 OD) δ 168.4 (COOMe), 158.7 (C-1′′(D)), 157.7 (C-9(A)), 153.7 (C-3′′(D)), 147.0 (C-4′(B)), 146.1 (C-3′(B)), 134.2 (C-1′ (B)), 130.6 (C-7(A)), 130.2 (C-5 (A)), 129.4 (C-10 (A)), 129.1 (C-5′′(D)), 124.2 (C-6(A)), 118.9 (C-4′′(D)), 118.3 (C-6′(B)), 117.4 (C-8(A)), 116.0 (C-5′(B)), 114.2 (C-2′(B)), 110.2 (C-6′′(D)), 104.2 (C-2′′(D)), 100.5 (C-2 (C)), 52.4 (COOMe), 34.6 (C-3(C)), 30.7 (C-4(C)). FT-IR (ATR) v max : 3319, 2974, 1681, 1606, 1504, 1436, 1284, 1259, 1236, 1195, 1153, 1147, 1107, 1018, 968, 896, 840, 769 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 23 H 18 O 7 405.0980, found 405.0981. Analytical HPLC (λ =280 nm): purity: 99 %; t R =28.7 min. Compound 42: 1 H NMR (400 MHz, CD 3 OD) δ 7.97 (d, J =2.1, 1H, H-5(A)), 7.80 (dd, J =8.5, 2.1, 1H, H-7(A)), 7.13–7.11 (m, 2H, H-2′(B), H-5′′(D)), 7.01 (dd, J =8.3, 2.2,1H, H-6′(B)), 6.98 (d, J =8.5, 1H, H-8(A)), 6.82 (d, J =8.3,1H, H-5′ (B)), 6.40–6.37 (m, 2H, H-2′′(D), H-6′′(D)), 4.11 (t, J =2.9, 1H, H-4(C)), 2.35* (dd, J =14.0, 2.9, 1H, H-3b (C)), 2.31* (dd, J =14.0, 2.9, 1H, H-3a (C)); 13 C NMR (100 MHz, CD 3 OD) δ 170.0 (COOH), 158.9 (C-1′′(D)), 157.9 (C-9(A)), 154.0 (C-3′′(D)), 147.2 (C-4′(B)), 146.4 (C-3′(B)), 134.6 (C-1′(B)), 131.2 (C-7(A)), 130.7 (C-5 (A)), 129.5 (C-10 (A)), 129.4 (C-5′′ (D)), 125.1 (C-6(A)), 119.2 (C-4′′(D)), 118.6 (C-6′(B)), 117.6 (C-8(A)), 116.3 (C-5′(B)), 114.5 (C-2′(B)), 110.5 (C-6′′(D)), 104.5 (C-2′′(D)), 100.7 (C-2(C)), 35.0 (C-3(C)), 34.9 (C-4(C)) (*these signals could be interchanged); FT-IR (ATR) v max : 3300, 1684, 1610, 1506, 1440, 1153, 1107, 1020, 970 cm −1 ; HRMS (ESI-TOF) m/z [M−H] - Calcd. for C 22 H 16 O 7 391.0823, found 391.0828. Analytical HPLC (λ =280 nm): purity: 93 %; t R =15.9 min. 2.5. Human lactate dehydrogenase A enzymatic activity assay Enzymatic activity of hLDHA was determined with recombinant human LDHA (95 %, specific activity >300 units/mg and concentration of 0.5 mg/mL, Abcam, Cambridge, United Kingdom) in the presence of sodium pyruvate (96 %, Merck) as substrate and β-NADH (≥97 %, Merck) as cofactor in 100 mM potassium phosphate buffer, pH 7.4. The enzymatic assay was performed on 96-well microplates and the decrease in the β-NADH absorbance (340 nm) was measured in a Synergy HT Multi-Detector Microplate Reader (BioTeK Instrument, Inc.) for 10 min at 28 ◦C. The activity was determined using the method employed by Li et al. [34] and modified as described here: in each well, the final volume was set to 200 µL and the final concentrations were 100 mM potassium phosphate buffer, 0.041 units/mL LDHA, 155 µM β-NADH, 1 mM pyruvate (saturated conditions), and DMSO solutions (1 %, v/v) of pure compounds at concentrations in the range of 0.09–200 µM. The reaction was initiated by the addition of pyruvate and a suitable linear timeframe was selected to calculate the slope of each concentration. Controls for the establishment of the 0 % and 100 % enzymatic activity were introduced in the assay and the compound 3-[[3[(cyclopropylamino)sulfonyl]- 7-(2,4-dimethoxy-5-pyrimidinyl)-4-quinolinyl]amino]-5-(3,5-difluor- ophenoxy)benzoic acid (GSK 2,837,808 A, Tocris, Minneapolis, USA) was used as a positive control at 1 µM in well [28]. All measurements were made in triplicate and data were expressed as the mean ±SD and plotted in GraphPad Prism version 5.00 for Windows (GraphPad Software, La Jolla, California, USA). Nonlinear regression analysis was used for dose response curve, fitting of logarithm of inhibitor concentration vs normalized enzymatic activity, to calculate IC 50 values (separate graphs, used to calculate IC 50 values for each compound, have been included in the Supplementary Material section). 2.6. Human lactate dehydrogenase B enzymatic activity assay Enzymatic activity of hLDHB was determined with recombinant human LDHB (95 %, specific activity >300 units/mg and concentration of 1.0 mg/mL, Abcam, Cambridge, United Kingdom) following the same protocol described in subsection 2.5. 2.7. Lipophilicity and cell membrane permeability Calculated log P (ClogP) has been estimated using the MOE computational tool [52]. 2.8. Mouse hepatocytes isolation and culture Hepatocytes were isolated in situ by collagenase perfusion method from male C57BL/6 Agxt1 -/- Grhpr -/- and Hoga1 -/- mouse livers [53]. A total of 2.5 ×10 6 cells/well were cultured in six-well plates with Williams E medium (Thermo Fisher, Whaltham, USA) supplemented with 5 % fetal bovine serum (Thermo Fisher), 2 mM L-glutamine, 100 U/mL penicillin, 100 μ g/mL streptomycin, 2.2 mUI/mL insulin, 2.5 μ g/mL amphotericin B and 0.3 μ g/mL hydrocortisone. After 24 h, medium was changed to serum-free Williams E medium. High levels of metabolic substrates (5 mM glycolate, 10 mM glycolate and 10 mM hydroxyproline) were used in these short-term cultures to enhance production of oxalate in Agxt1 -/- , Grhpr -/- and Hoga1 -/- primary hepatocyte cultures, respectively [7]. Under these conditions, oxalate accumulates in the media of hyperoxaluric hepatocytes in a time-dependent manner. Culture media was collected at 24 and 48 h after addition of inhibitors to measure the amount of oxalate excreted by the Agxt1 -/- , Grhpr -/- or Hoga1 -/- hepatocytes. 2.9. Cell viability and cytotoxicity 96-Well plates were seeded with 1.0 ×10 4 cells/well and treated with the same concentrations of compounds as in 6-well plates. At each time point, 20 μ L of Cell Titer 96® Aqueous One Solution Reagent (Promega, Madison, WI, USA) was added to the medium, incubated 2 h at 37 ◦C, 5 % CO 2 , and absorbance measured at 490 nm. Relative cellular viability at a concentration of 10 μ M for each inhibitor was calculated from the concentration of colored MTS formazan (3-(4,5-dimethylth- iazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)–2H-tetrazo- lium) found in wells treated with the inhibitor by comparison with the concentration of formazan found in wells without added inhibitor (negative controls). Each relative cellular viability value for every inhibitor was calculated as an average of three replicates. Values showed a cell death below 3 % at 24 h and 48 h [7]. 2.10. Oxalate determination Oxalate excreted to the medium was measured with an oxalate oxidase assay kit (Greiner Diagnostic GmbH, Bahlingen, Germany), following manufacturer’s instructions. The method involves oxidation of oxalate (1 equiv) by oxalate oxidase with formation of H 2 O 2 (1 equiv) and subsequent utilization of the generated H 2 O 2 for the formation of a dye (absorbance at 590 nm) in a HRP (horseradish peroxidase) catalyzed reaction with the substrates 3-methyl-2-benzothiazolinone hydrazone (MBTH) and 3-(dimethylamino)benzoic acid (DMAB). For oxalate quantification, a standard curve was constructed using aqueous dilutions of oxalate containing 0, 0.025, 0.05, 0.1, 0.2 and 0.25 nmol/ μ L (R 2 =0.9978). For the standard curve, absorbance was measured at 590 nm following the same protocol described above [9]. Relative oxalate for each inhibitor at the concentration of 10 μ M was calculated from the concentration of extracellular oxalate found in wells treated with the inhibitor, as a percentage of the extracellular oxalate found in wells without added inhibitor (negative controls). Each relative oxalate value for every inhibitor was calculated as an average of three replicates. A. Alejo-Armijo et al. Bioorganic Chemistry 129 (2022) 106127 8 3. Results and discussion 3.1. Chemical synthesis According to our experience in the synthesis of flavylium salts [46] and the influence of the electronic features of flavylium salts on their reactivity with nucleophiles to yield bicyclic adducts [38], we have selected for this work several electron-withdrawing groups at A-ring, such as NO 2 , halogens, COOH, and COOMe. Thus, the two-step synthetic method followed to obtain 2,8-dioxabicyclo[3.3.1]nonane derivatives (23–42) is based on the addition of a commercial oxygenated aromatic nucleophile, phloroglucinol (21) or resorcinol (22), to flavylium salts (10–20), electrophilic compounds previously synthesized (Scheme 1). The synthesis of those flavylium (1-benzopyrylium) salts was performed by aldol condensation in acid media between salicylic aldehyde (1) and derivatives (2–5) and acetophenone or propiophenone derivatives (6–9) according to procedures previously used by us [38,41,46]. Most of the flavylium salts (10–19) were obtained with moderate to high yield (63–91 %) using CH 3 CO 2 H/conc. H 2 SO 4 4:1 (v/ v) as acid media (Method A, entries 1–10 in Table 1). However, salt 20 could not be prepared in proper yield under these conditions, and a second method using ethanol saturated with gaseous hydrogen chloride had to be followed to get a high yield of 20 (96 %, Method B, entry 11 in Table 1). The substitution pattern of rings A, B, and C in the flavylium salts was designed taking into account two key structural features: (1) hydroxyl groups are present in all polyphenolic scaffold-based LDHAi’s [25], (2) electron-withdrawing groups at A-ring provide adequate electronic densities in the aromatic system in order to achieve the second step of the synthesis of dioxabicyclo derivatives [38]. Thus, flavylium salts with NO 2 , Cl, Br, COOH groups at the A-ring and one OH group or a catechol moiety at the B-ring have been synthesized (12–20) in addition to flavylium salts with non-substituted A-rings (10, 11). The addition of phloroglucinol (21) or resorcinol (22) to flavylium salts (10–20) was performed following a similar procedure to that described in the literature [54] and previously used by us [38,41]. Thus, the corresponding flavylium salts were mixed with one equivalent of phloroglucinol or resorcinol, and the reaction was carried out in absolute methanol or tetrahydrofuran at 50 ◦C (Method C, entries 12–31 in Table 1, the yields were calculated from starting aldehydes). The highest yields obtained in these reactions (45–64 %) corresponded to 2,8-dioxa- bicyclo[3.3.1]nonane derivatives with a NO 2 group at A-ring (25–28 and 35–38). Other derivatives (29–34 and 39–42) with electronwithdrawing groups at A-ring (Cl, Br, COOH or COOMe) were obtained with moderate yield (30–58 %), and other derivatives, like 23 and 24, with no substituents at A-ring, were obtained with lower yields (34 % and 22 %, respectively). The use of methanol as solvent in the reaction between flavylium salt 20 and phloroglucinol (21) or resorcinol (22) promoted, after the formation of 2,8-dioxabicyclo[3.3.1]nonane derivatives, the esterification reaction between the carboxylic group and methanol, catalyzed by traces of acid from the flavylium salts. In this reaction conditions, the corresponding methyl ester derivatives 33 and 41 were the main reaction products (58 % and 44 % yield, respectively; entries 22 and 30 in Table 1) and carboxylic derivatives 34 and 42 were obtained with low yields (9 % and 7 %, respectively). However, it was possible to obtain 34 and 42 with moderate yields (51 % and 30 %, respectively, entries 23 and 31 in Table 1) following the same procedure, but changing methanol by tetrahydrofuran. The largest nucleophilic character of phloroglucinol (21) versus resorcinol (22) was Scheme 1. Synthesis of flavylium salts (10–20) and 2,8-dioxabicyclo[3.3.1]nonane derivatives (23–42). Reagents and conditions: (i) H 2 SO 4 , HOAc; (ii) EtOH, HCl (g); (iii) MeOH, 50 ◦C, 24 h; (iv) MW, MeOH, 80 ◦C, 20 min; (v) MW, MeOH/aq buffer (pH 5.8), 100 ◦C. A. Alejo-Armijo et al. Bioorganic Chemistry 129 (2022) 106127 9 clearly translated into higher yields in the synthesis of 2,8-dioxabicyclo [3.3.1]nonane derivatives with a phloroglucinol moiety. In fact, some differences in yields were observed for dioxabicycles 29–34 (41–58 %) versus 39–42 (30–44 %), although they have not been observed for dioxabicycles with nitro group at A-ring, independently of the presence of a phloroglucinol moiety (25–28, entries 14–17 in Table 1) or a resorcinol one (35–38, entries 24–27 in Table 1). These results could be explained according to the highest electrophilic character of the corresponding flavylium salt that could be veiling the influence of the different nucleophilic properties of phloroglucinol (21) and resorcinol (22). In addition, no significant differences were observed in terms of yields between dioxabicycles with different substitution pattern at the C- ring, so yields for 25, 26, 35 and 36 (R 2 =H) (45–64 %) were similar to 27, 28, 37 and 38 (R 2 =CH 3 ) (50–64 %). The structures of the synthesized compounds were characterized by IR, 1 H NMR, 13 C NMR, and 2D NMR spectroscopy as well as by HRMS spectrometry. Compounds 15–20, 24, 26 and 28–42 have been synthesized for the first time in this work and their structural characterization is now reported ( 1 H NMR and 13 C NMR spectra of the new compounds are included in the Supplementary Material section). 3.2. Inhibitory activity of 2,8-dioxabicyclo[3.3.1]nonanes 23–42 against hLDHA The inhibitory effect of the 2,8-dioxabicyclo[3.3.1]nonane derivatives 23–42 over hLDHA catalytic activity, in the conversion of pyruvate to lactate, was evaluated by a UV spectrophotometric assay [34]. Briefly, the decrease in the co-substrate β-NADH absorbance at 340 nm (due to its oxidation into NAD + ) was measured for 10 min and the slope was compared with that obtained when no inhibitors were added (100 % enzymatic activity). The inhibitory activity of the commercial compound 3-[[3[(cyclopropylamino)sulfonyl]-7-(2,4-dimethoxy-5-pyr- imidinyl)-4-quinolinyl]amino]-5-(3,5-difluorophenoxy)benzoic acid (GSK-2837808A) (Fig. 1) was also assessed to validate the method [28]. Moreover, the compound stiripentol (Fig. 1) was synthesized by us, according to the literature [55], and used as a reference, because of its known ability to decrease the urine oxalate excretion [18]. In a first screening assay, all synthesized compounds with the 2,8- dioxabicyclo[3.3.1]nonane scaffold and stiripentol were evaluated, at a single concentration of 50 μ M, and the results were expressed as percentage of inhibition at that concentration (Table 2). Significant values, above 40 % of inhibition, were achieved for fifteen of them (23–27, 29–35, 39–41). However, compounds 28, 36–38 and 42 showed low percentages of inhibition (31.5 %–38.1 %). The reference compound Table 1 Reaction yields in the synthesis of flavylium salts 10–20 and dioxabicycles 23–42. entry compound R 1 R 2 R 3 R 4 synthetic method yield (%) a 1 10 H H OH – A 90 2 11 H H H – A 79 3 12 NO 2 H OH – A 77 4 13 NO 2 H H – A 76 5 14 NO 2 CH 3 OH – A 85 6 15 NO 2 CH 3 H – A 91 7 16 Cl H OH – A 79 8 17 Cl H H – A 63 9 18 Br H OH – A 90 10 19 Br H H – A 68 11 20 COOH H OH – B 96 12 23 H H OH OH C 34 13 24 H H H OH C 22 14 25 NO 2 H OH OH C 64 15 26 NO 2 H H OH C 45 16 27 NO 2 CH 3 OH OH C 63 17 28 NO 2 CH 3 H OH C 50 18 29 Cl H OH OH C 43 19 30 Cl H H OH C 42 20 31 Br H OH OH C 41 21 32 Br H H OH C 49 22 33 COOMe H OH OH C 58 23 34 COOH H OH OH C (9) 51 b 24 35 NO 2 H OH H C 50 25 36 NO 2 H H H C 64 26 37 NO 2 CH 3 OH H C 60 27 38 NO 2 CH 3 H H C 64 28 39 Cl H OH H C 36 29 40 Br H OH H C 35 30 41 COOMe H OH H C 44 31 42 COOH H OH H C (7) 30 c a Yields calculated from the starting aldehyde 1–5 (see Scheme 1). b Yield was improved up to 51% when THF was used as solvent instead of MeOH (9%). c Yield was improved up to 30% when THF was used as solvent instead of MeOH (7%). A. Alejo-Armijo et al.