Amino Acids Bearing Aromatic or Heteroaromatic Substituents as a New Class of Ligands for the Lysosomal Sialic Acid Transporter Sialin

This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes. pubs.acs.org/jmc Article Amino Acids Bearing Aromatic or Heteroaromatic Substituents as a NewClassofLigandsfortheLysosomalSialicAcidTransporterSialin ́ Lilian Dubois, Nicolas Pietrancosta, Alexandre Cabaye, Isabelle Fanget, Cecile Debacker, Pierre-André Gilormini, Patrick M. Dansette, Julien Dairou, Christophe Biot, Roseline Froissart, Anne Goupil-Lamy, Hugues-Olivier Bertrand, Francine C. Acher, Isabelle McCort-Tranchepain,* Bruno Gasnier,* and Christine Anne* Cite This: J. Med. Chem. 2020, 63, 8231−8249 Read Online Metrics & More Article Recommendations sı Supporting Information ACCESS * ABSTRACT: Sialin, encoded by the SLC17A5 gene, is a lysosomal sialic acid transporter defective in Salla disease, a rare inherited leukodystrophy. It also enables metabolic incorporation of exogenous sialic acids, leading to autoanti- bodies against N-glycolylneuraminic acid in humans. Here, we identiﬁed a novel class of human sialin ligands by virtual screening and structure−activity relationship studies. The ligand scaﬀold is characterized by an amino acid backbone with a free carboxylate, an N-linked aromatic or heteroaromatic substituent, and a hydrophobic side chain. The most potent compound, 45 (LSP12-3129), inhibited N-acetylneuraminic acid 1 (Neu5Ac) transport in a non-competitive manner with IC ≈2.5μM,avalue400-foldlowerthantheK forNeu5Ac.Invitroandmoleculardockingstudies 50 M attributed the non-competitive character to selective inhibitor binding to the Neu5Ac site in a cytosol-facing conformation. Moreover, compound 45 rescued the traﬃcking defect of the pathogenic mutant (R39C) causing Salla disease. This new class of cell-permeant inhibitors provides tools to investigate the physiological roles of sialin and help develop pharmacological chaperones for Salla disease. ■ INTRODUCTION Owing to a mutation in the cytidine monophosphate- SLC17 transporters, a subgroup from the major facilitator Neu5Ac hydroxylase (CMAH) gene, humans do not superfamily (MFS) of secondary active transporters, are synthesize this sialic acid, which is abundant in other species.12 involved in the excretion of urate and other organic anions Neu5Gcisthus incorporated into human cell glycans, inducing Downloaded via 62.96.59.114 on February 6, 2024 at 08:54:30 (UTC).in the kidney, the uptake of anionic neurotransmitters the production of anti-Neu5GC autoantibodies associated with 13,14 (glutamate and ATP) into synaptic vesicles, and the export chronic inﬂammation, cancer, and atherosclerosis. This of sialic acids and acidic hexoses from lysosomes for reuse in endocytic route is also exploited in glycoengineering 1−7 + approaches to alter cell-surface glycans with synthetic sialic metabolism. The latter function is ensured by sialin, a H / 5−7 acid derivatives and thereby modulate immune responses, See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.sialic acid symporter encoded by the SLC17A5 gene.Owing to its H+ coupling and the steep pH gradient (∼2.5 units) inhibit pathogen binding, target toxins to tumors, or track cells across the lysosomal membrane, sialin actively exports sialic in vivo after transplantation.15,16 acids from the lysosomal lumen, thus contributing with de The essential role of sialin in sialic acid homeostasis is also novo biosynthesis to their cytosolic availability. This in turn highlighted by its inactivation in two autosomal recessive impacts the sialylation of cell surface glycoconjugates and their genetic diseases, Salla disease and infantile free sialic acid role in cell−cell, cell−extracellular matrix, and cell−pathogen 5,17,18 storage disease (ISSD). Both diseases are associated with interactions. Sialin has also been implicated in the uptake of defective lysosomal export of sialic acids,6,7,19 leading to their 8 nitrate at the plasma membrane and the uptake of aspartate, accumulation in diverse tissues and their excretion in urine. glutamate, and N-acetyl-aspartyl-glutamate into synaptic However, they strongly diﬀer in their clinical course and 9,10 vesicles. In addition to its role in recycling endogenous sialic acids, Received: March 20, 2020 lysosomal export by sialin enables the metabolic incorporation Published: July 1, 2020 of exogenous sialic acids internalized by ﬂuid-phase endocy- tosis,11 thus circumventing the absence of a sialic acid transporter at the plasma membrane. This endocytic route impacts human health by incorporating dietary-derived N- glycolylneuraminic acid (Neu5Gc) into glycoconjugates. ©2020 American Chemical Society https://dx.doi.org/10.1021/acs.jmedchem.9b02119 8231 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 1. Identiﬁcation of novel sialin inhibitors (A) structures (2 to 10) of the best hits from a previous virtual screening.33 (B) Whole-cell assay of human sialin. Mutation of its lysosomal sorting motif redirects sialin to the plasma membrane to facilitate transport measurements. (C) Human 3 sialin was assayed for [ H]Neu5Ac uptake at pH 5.0 in the absence (control) or presence of these compounds (means ± SEM of two to four independent experiments). severity. While ISSD aﬀects multiple organs from birth, causing The pathophysiology of Salla disease remains poorly death within a few months or years, Salla disease selectively understood. Sialin-defective mice studies have shown that aﬀects the white matter of the brain in a progressive brain hypomyelination results from defective maturation of 17,20,21 oligodendrocytes.23,24 However, the link between this defect manner. Salla patients show hypotonia, ataxia, nystag- and defective lysosomal sialic acid export or other transport mus, and delayed motor development during the ﬁrst year. 9 Psychomotor milestones, including speech, progressively activities of sialin (see ref 25) is unclear. Decreased worsen during infancy and childhood, leading to severe downregulation of the polysialylated neural cell adhesion molecule (PSA-NCAM), which could delay oligodendrocyte/ motor and cognitive deﬁcits in adulthood. Magnetic resonance 23 axon contacts, may be involved. However, this decrease is imaging shows nonspeciﬁc brain hypomyelination with a limited, indicating that other mechanisms, such as impaired 22 26 thinning of the corpus callosum. Current treatment is ganglioside metabolism, should contribute to the myelination supportive. defect. 8232 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article On the other hand, genotype−phenotype relationship and 4 are both amino acid derivatives with common features: studies have shown a clear correlation between the level of the α amine is protected by the same hydrophobic carbamate, Neu5Ac transport and the clinical severity induced by (ﬂuoren-9-yl)methoxycarbonyl (Fmoc); the carboxylic acid is SLC17A5 mutations. Salla disease is almost exclusively free; and a 4 to 6-carbon side chain bears a distal amine associated with one missense mutation, R39C, in the protected by a benzyloxycarbonyl (Cbz) or tert-butoxycarbonyl homozygous or compound-heterozygous state, which partially (Boc) group. impairs lysosomal targeting of sialin6,27 and decreases its These similarities suggested a novel scaﬀold recognized by transport activity to ∼20% of the wild-type level,6,7 whereas sialin. To test this hypothesis, we evaluated a series of 14 N- ISSD-associated mutations abolish either the expression or the Fmoc amino acids (11−24) in our transport assay (see 6,7,18 structures in the Supporting Information, pp. S3−S4). N-Fmoc transport activity of sialin. The association of Salla disease with residual sialic acid transport has been conﬁrmed in a rare amino acids 11, 12, 16, 22, and 24 were prepared according to 28 classical methods (see the Supporting Information), whereas case of a patient with a homozygous K136E mutation, which also partially preserves sialic acid transport and lysosomal compounds 13−15, 17−21,and23 were commercially localization.6,7 The traﬃcking defect may result from available. Interestingly, 10 out of 14 compounds inhibited destabilization of the R39C and K136E mutants and their Neu5Ac uptake with an eﬃciency greater than or similar to retention by the protein quality control system. This feature that of compounds 3 and 4 (Figure 2), thus conﬁrming the along with the residual transport activity and the overwhelming predominance of the R39C mutation suggests pharmacological chaperone therapy as a promising option to treat Salla disease. In contrast with other SLC17 transporters such as 29−31 32 VGLUT and VNUT proteins, we lack pharmacological tools to study the cellular and physiological roles of sialin or to help rescue the R39C mutant causing Salla disease. In a 33 previous study, we characterized the interaction of human sialin with synthetic sialic acids and identiﬁed one compound with IC ≈ 40 μM, a value 25-fold lower than the K for 50 M Neu5Ac. We also built a cytosol-facing homology model of human sialin based on a distantly homologous MFS crystallo- graphic structure, the glycerol-3-phosphate transporter (GlpT) Figure 2. N-Fmoc amino acids (30 μM) were tested for inhibition of 34 from Escherichia coli (E. coli). After validating the sialic acid- human sialin as in Figure 1. The dotted lines correspond to the binding site of this model by site-directed mutagenesis, virtual inhibition by compounds 3 and 4. screening against the ZINC database and Neu5Ac transport studies led us to identify the artiﬁcial tripeptide FR139317 as a existence of a novel scaﬀold characterized by an N-Fmoc 33 new sialin ligand unrelated to sialic acids. In this study, we group, a free carboxylate, and a hydrophobic side chain. The disclose other virtual hits from this screen and identify a novel lower activity of compounds 22−24 may result from the scaﬀold of the sialin ligand well suited for chemical presence of a bulkier side chain (compare Fmoc-Lys(Fmoc)- modiﬁcations. After optimization, the new ligands obtained OH24with3)oraconstrainedringstructure of the side chain inhibit sialin with a micromolar aﬃnity in a non-competitive in Fmoc-Pro-OH 23. All tested amino acids except Fmoc-DL- manner and one of them could partially rescue the traﬃcking Leu-OH 12 belonged to the L-series. Comparison of the defect of the pathogenic R39C mutant. racemic 12 and L-13 forms of Fmoc-Leu-OH suggests that the RESULTS α carbon conﬁguration is not critical to binding sialin (Figure ■ 2). Screening and Scaﬀold Selection. Our previous virtual To characterize further this sialin ligand scaﬀold, we screening of a GlpT-based homology model of human sialin examined the inﬂuence of the proximal amine substitution. 35 The 9-ﬂuorenylmethoxycarbonyl group of compound 3 was against a subset of the ZINC database (∼12,000 com- pounds) identiﬁed the endothelin-A receptor antagonist thus replaced by a coumarinyl, quinolinyl, xanthenyl, or FR139317 2 and eight other molecules33 (Figure 1A) as anthracenyl group linked to the α amine by an amide or candidate sialin ligands. Transport studies conﬁrmed 2 as a carbamate bond. In another set of compounds, we examined competitive inhibitor of N-acetylneuraminic acid 1 (Neu5Ac) the inﬂuence of substituents attached to the side chain. For this transport by human sialin with a Ki of 20 μM, a value ∼50-fold purpose, we ﬁrst introduced 4-methylcarbonylcoumarin lower than the K for Neu5Ac.33 We thus tested the other instead of Cbz in compound 3 and then selected compound M compoundsusingthesametransport assay, which is based on a 16 among those with good activity to alter the side chain sorting mutant (L22G/L23G) redirected to the plasma because the sulfanyl group allows alkylation with substituted 6 coumarins in a metabolically stable manner (resistance to membrane. In this approach, the poorly tractable lysosomal 3 cytosolic and lysosomal hydrolases). The synthesis and export is replaced by a whole-cell [ H]Neu5Ac uptake in acidic extracellular buﬀer to mimic the lysosomal lumen and facilitate biological activity of the resulting compounds are described transport measurements (Figure 1B). This technique provides in the following sections. higher signal-to-noise ratios than lysosomal assays with highly Chemistry. All 7-hydroxy, 4-substituted coumarins were similar Neu5Ac transport properties.6 synthetized via Pechmann condensation in good yields using Among the novel virtual hits, four compounds inhibited either a Brønsted acid with resorcinol or Lewis acid with an Neu5Ac uptake and three of them, Fmoc-Lys(Cbz)-OH 3, aminophenol derivative (Scheme 1). Compounds 25−28 were Fmoc-D-Phe(pCH NHBoc)-OH 4, and calcein 5 were more prepared according to conventional methods with slight 2 36−38 eﬃcient inhibitors than Neu5Ac (Figure 1C). Interestingly, 3 improvements (see the Supporting Information). 8233 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article Scheme 1. Synthesis of Functionalized Coumarinsa a Reagents and conditions: (i) ethyl-4-chloroacetoacetate, 70% aq H SO ,0°C, 1.5 h, 97%; (ii) K CO , acetone, Me SO , RT, overnight, 94%; (iii) 2 4 2 3 2 4 HO,reﬂux, 48 h, 69%; (iv) ethyl-4-chloroacetoacetate, CH SO H RT, 4 h, 97%; (v) (a) ClTi(OiPr) , dimethyl 3-oxoglutarate, toluene, reﬂux, 2 3 3 3 overnight; (b) NaOH, MeOH, RT, overnight, 35−66% (two steps); (vi) succinic anhydride, 70% aq H SO , 140 °C, 4 h, 92%; (vii) MOM-Cl, 2 4 DIEA, CH Cl ,0°C, 45 min, 71%; (viii) Fmoc-Cl, pyridine, RT, 2 h, 34%; and (ix) TFA, CH Cl ,0°C then RT, 1.5 h, 83%. 2 2 2 2 Using chlorotriisopropoxytitanium(IV)39 was more eﬃcient Alkylation of Fmoc-Cys-OH by coumarins 25, 26, and 28 than zinc(II) chloride40 for the synthesis of 7-(dialkylamino)- proceeded at room temperature overnight in the presence of coumarins 29 and 30. The xanthenyl derivative 32 was an excess of base to give compounds 45−47 at approximately 41 30% yields (Scheme 3). obtained according to a known procedure as the 3-(3,6- Structure−Activity Analysis of the Heterocycle-Sub- dihydroxy-9H-xanthen-9-ylidene)propionic acid tautomer, un- stituted Amino Acids. Compounds 35−42 and 44 (see ambiguously shown by NMR in aprotic solvent (DMSO). structures in the Supporting Information, pp. S3−S4) were Regioselective protection of the 7-hydroxy-4-hydroxyme- tested in our whole-cell assay of sialin at 30 μM(Figure 3)as thylcoumarin 27 with methoxymethyl chloride before for the Fmoc-amino acids 11−24. When 9-ﬂuorenylmethyl introducing the 9H-ﬂuoren-9-yl methyl carbonate in the carbonate is directly linked to coumarin 27, the resulting benzylic position and deprotection of the 7-hydroxyl led to 35. compound 35 does not signiﬁcantly inhibit sialin, showing that Amide bond formation between the amine of a lysine-OMe removal of the amino acid backbone and free carboxylate derivative and acids 29−32 as well as commercially available 7- abolishes binding. Replacing the 9-ﬂuorenylmethyl carbamate hydroxy-4-carboxymethylcoumarin and 4-carboxyquinolin-2- linked to the α amine in 3 by diverse heterocycles in one was carried out42 (Scheme 2). Attempts to condense previously activated 7-hydroxy-4-carboxymethylcoumarin as an compounds 36−41 reduced activity, ranging from a moderate acid chloride or HOBt ester in the presence of EDC with Lys loss with the bulky coumarin derivative 41 to virtually no (Cbz)-OMe failed. In contrast, amide bond formation inhibition with the quinolinyl compound 39. No signiﬁcant occurred in the presence of HBTU and HOBt via uronium diﬀerence was observed among coumarinyl derivatives 36−38 43,44 with this 4-carboxymethylcoumarin or bearing diﬀerent substituents at the 7-position (hydroxy, salt activation dimethyl, or diethylamino). Of note, when the amide bond xanthene 32, leading to, after saponiﬁcation of the methyl of 36 was converted into a more ﬂexible carbamate bond in 42, ester, 36 and 40 in poor yields (19 and 8%, respectively). The the aﬃnity for sialin increased to a level similar to that of condensation of 29, 30, 4-carboxyquinolin-2-one, and 31 with compound 3. Lys(Cbz)-OMe performed in the presence of HBTU in The inﬂuence of substituting the side chain was tested in CHCl also led to the expected products, generally in better compounds 44−47. Replacing the benzyloxycarbonyl group 2 2 45 yields (12−50%). In contrast, under these conditions, the present in 3 by a substituted coumarin in 44 did not coupling of 29 to the distal amine of lysine 43 was poorly signiﬁcantly alter activity. In contrast, the Fmoc-Cys- eﬃcient (5%, not optimized). The carbamate analogue 44 of (substituted coumarin)-OH 45−47 had better activities than 36 was obtained from 27 with carbonyl diimidazole activation. the cysteine analogue Fmoc-Cys(pMeOBzl)-OH 20 (Figure 3, 8234 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article a Scheme 2. Synthesis of N-Acyl and N-Carbamate Diprotected Lysine a Reagents and conditions: (i) (a) HBTU, HOBt, DMF, DIEA, RT, overnight and (b) aq LiOH, THF, RT, 2 h, 19% for 36, 8% for 40 (two steps); (ii) (a) HBTU, CH Cl , DIEA, RT, 2 h and (b) aq LiOH, THF, RT, 2 h, 50% for 37, 19% for 38, 43% for 39, 12% for 41, 5% for 44 (two steps); 2 2 (iii) (a) CDI, DMF, DIEA, RT then 80 °C, 1.5 h, 21% and (b) aq LiOH, THF, RT, 2 h 21% (two steps). Scheme 3. Synthesis of S-Alkyl Cysteinea red dotted line) or the lysine analogue 3 (brown line), showing that the length of the side chain is not critical to recognition. The most active compounds 11, 13, and 45−47 inhibited Neu5Acuptake by sialin in a concentration-dependent manner (Figure 4) in agreement with a speciﬁc interaction. Fmoc- Cys[Coum-7-OH]-OH45(LSP12-3129) was the most potent inhibitor with an IC of 2.4 ± 0.7 μM(n = 3), a value ∼400- 50 33 fold lower than the K for Neu5Ac. To assess its selectivity, M + compound 45 was tested on two other H -driven lysosomal transporters, LYAAT1 and cystinosin, measured in whole-cell assay similar to that of sialin.46,47 Interestingly, it had no eﬀect on LYAAT1 and partially inhibited cystinosin (43.7 ± 9.1% inhibition, n = 3) at a concentration (30 μM) that fully inhibits a sialin (Figure 4C,D). Reagents and conditions: (i) from 25 and 28; DIEA, THF, RT, 16 h, Compounds 45 and 13 (Fmoc-Leu-OH) Are Non- 35 and 32%, respectively; from 26:EtN, DMF, RT, 16 h, 28%. 3 competitive Inhibitors. We next studied how 45 and, for comparison, 13 (Fmoc-Leu-OH) interact with human sialin. 8235 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article were incubated with [3 3 H]Neu5Ac or [ H]Fmoc-Leu-OH at similar saturation levels (IC /10, 100 and 2 μM, respectively), 3 50 no sialin-dependent [ H]Fmoc-Leu-OH uptake could be detected in acidic medium in contrast with the strong 3 [ H]Neu5Ac uptake signal (Figure S1). Fmoc-Leu-OH thus acts as a blocker of sialin rather than a translocated substrate. To extend these experiments to 45, we developed another whole-cell assay where sialin activity is detected by the co- + transport of H . We co-expressed the ratiometric ﬂuorescent 2 48 pHprobeE GFP withanmRFP-tagged humansialin L22G/ L23G construct in HEK293 cells to detect the cytosolic acidiﬁcation associated with the uptake of sialin substrates. Neu5Ac (10 mM) applied at pH 5.5 induced a signiﬁcant 2 E GFP ﬂuorescence response consistent with the expected Figure 3. Screening of heterocycle-substituted amino acids. acidiﬁcation. In contrast, a saturating concentration (30 μM) Compounds (30 μM) were tested for inhibition of human sialin as of 45 did not induce any detectable acidiﬁcation (Figure S2). in Figure 1. Each derivative is colored according to and compared Therefore, neither Fmoc-Leu-OH nor 45 is signiﬁcantly with (dotted lines) its reference compound, 3 (brown) or 20 (red). translocated by human sialin, implying that they both act as blockers. Fmoc-Leu-OH was chosen because it eﬃciently inhibits We next determined the inhibition mode of these Neu5Ac uptake (Figure 4B; IC = 23.3 ± 3.5 μM, n =4) compounds and studied the saturation kinetics of [3H]Neu5Ac 50 uptake at varying inhibitor concentrations. Unexpectedly, and is commercially available as a radiolabeled compound. Of note, Fmoc-Leu-OH had limited eﬀects on LYAAT1 and despite the “active site” focus of our initial virtual screening, cystinosin at a concentration (100 μM) that fully inhibits sialin both Fmoc-Leu-OH and 45 acted as non-competitive rather (Figure 4C,D). We ﬁrst asked whether these compounds are than competitive inhibitors since they decreased the Vmax of substrates (translocated ligands) or blockers (not translocated) uptake while leaving the K for Neu5Ac essentially unchanged M of the lysosomal transporter. When sialin-expressing cells (see representative experiments in Figure 5A,B). At 10 μM, (EGFP-Sialin L22G/L23G construct) and mock cells (EGFP) Fmoc-Leu-OHdecreased Vmax by 56 ± 5% (n = 5 independent Figure 4. Dose−response relationship and selectivity of selected compounds. (A) Inhibition of human sialin. Means ± SEM of two to three independent experiments. (B) Representative sialin inhibition curves for compounds 13 (blue) and 45 (green) (triplicate measurements). (C, D) + Eﬀect of 13 and 45 on two other H -driven transporters, LYAAT1 and cystinosin. (C) Representative experiments (triplicate measurements). (D) Means ± SEM of three independent experiments where the three transporters were assayed in parallel. 8236 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article 3 Figure 5. Inhibition mode of Fmoc-Leu-OH and 45. (A, B) Representative Eadie−Hofstee graphs (left) of [ H]Neu5Ac uptake saturation kinetics in the presence or absence of Fmoc-Leu-OH 13 and 45 (means ± SEM of triplicate measurements). The reciprocal maximal velocity is plotted against the inhibitor concentration (right graphs) to determine inhibition constants. (C) Potential mechanism accounting for the non-competitive inhibition by aromatic N-substituted amino acids (inhibitor 2). Black arrows depict the Neu5Ac transport cycle in our assay. Red arrows depict two inhibition mechanisms. experiments) with K values of 1.6 ± 0.3 and 1.3 ± 0.2 mM in conformation (Figure 5C, inhibitor 1), the inhibitor and M the absence and presence of inhibitors, respectively. Similarly, Neu5Ac should exclude each other and an excess of Neu5Ac 45 (3 μM) decreased V by 31 ± 8% (n = 3) with K values should displace the inhibitor binding equilibrium and rescue max M of 1.4 ± 0.5 and 1.7 ± 0.5 mM in the absence and presence of inhibition. In contrast, if the inhibitor permeates the cell inhibitors, respectively. Mean K values of 10.2 ± 3.0 μM(n = membrane (for instance, by diﬀusing across the lipid bilayer) i 3) and 7.8 ± 1.8 μM(n = 4) were obtained for Fmoc-Leu-OH and binds to the inward-facing state, the competition between and 45, respectively. Neu5Ac (on the extracellular side) and the inhibitor (on the Potential Mechanism for the Non-competitive In- cytosolic side) should be biased toward inhibitor binding hibition. Conceivably, the non-competitive inhibition by (Figure 5C, inhibitor 2) because structural transitions are Fmoc-Leu-OH and 45 could reﬂect a decrease in the H+ much slower than binding equilibriums: when the cytosolic electrochemical gradient driving Neu5Ac uptake by sialin. inhibitor dissociates, the protein in the apo, inward-facing state However, the lack of an eﬀect of 45 on LYAAT1 and the will rebind with another inhibitor molecule before it can switch moderate inhibition of cystinosin by Fmoc-Leu-OH (Figure to the apo, outward-facing state. Extracellular Neu5Ac thus 4C,D) ruled out this possibility. As our virtual screening was cannot displace the inhibitor binding equilibrium in this case, conducted on a cytosol-facing model of sialin,33 we reasoned resulting in non-competitive inhibition. The non-competitive that the lack of competition between Neu5Ac, on one hand, inhibition of sialin by Fmoc-Leu-OH and 45 may thus result and 45 or Fmoc-Leu-OH, on the other hand, might result from from their selective or preferential binding to the cytosol-facing a selective blockade of sialin in the cytosol-facing state. Like state. It may be noted that Fmoc-Leu-OH does enter cells in a any secondary active transporter,49,50 sialin should operate sialin-independent manner (Figure S1), enabling its action on through alternating-access structural transitions, which expose the cytosolic side of the transporter. the substrate-binding site to either the luminal (extracellular in We used two approaches to test this model. First, we 51 our assay) or cytosolic side of the membrane. If an inhibitor examined how fast sialin recovers its transport activity upon binds to the Neu5Ac-binding site of the outward-facing washing of N-substituted amino acid inhibitors. If these 8237 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article inhibitors act on the extracellular side, washing should quickly (log P = 4.1 and 3.91, respectively) is also consistent with this restore Neu5Ac transport. On the contrary, if they model. predominantly act on the cytosolic side, recovery should be Second, we examined the docking of 45 to a novel 3D rate-limited by the cellular eﬄux of cytosolic inhibitors. Cells homology model of human sialin based on a closer template. Very recently, two crystal structures of E. coli H+ expressing sialin at the plasma membrane were thus /D-galactonate preincubated with Neu5Ac or N-substituted amino acids at symporter (DgoT) in outward-facing occluded and inward- 3 51 pH 5.0, washed at neutral pH, and tested for [ H]Neu5Ac facing open conformations have been determined. DgoT is a uptake at pH 5.0 in the absence of inhibitors. For comparison, closer homologue of human sialin (25.5% sequence identity 3 another set of cells was inhibited during [ H]Neu5Ac uptake and 45.0% similarity over 420 amino acids; BlastP alignment) as done previously. When this protocol was applied to Neu5Ac than E. coli GlpT (21.0% identity and 37.9 similarity over 309 1, the neutral wash fully restored [3H]Neu5Ac transport to a amino acids). Moreover, the DgoT structures have a higher level (130%) above that of the uninhibited control, probably resolution and better global model quality estimation than the 34,51 reﬂecting the trans-stimulation of sialin by the cytosolic pool of GlpT structure. Accordingly, we built two new sialin Neu5Ac that accumulated during the pre-incubation step (see homology models based on the inward-facing open (6E9N) ref 52 for a similar trans-stimulation in lysosomal membrane and outward-facing occluded (6E9O) conformations of DgoT vesicles). In contrast, when sialin was pre-inhibited by Fmoc- and performed docking experiments with ﬂexible side chains Leu-OH or 45, only partial recovery was achieved after a 15 53 located at the potential binding site using GOLD software as min wash (55 or 49% persisting inhibition, respectively) implemented in Discovery Studio 2019. (Figure 6), in agreement with our hypothesis of the action on 45 could not be docked to the outward-facing occluded sialin model because the binding site was too small to accommodate the inhibitor. In contrast, 45 was well accommodated in the binding site of the inward-facing model (Figure 7A), providing a putative structural explanation Figure 6. Persistence of 13- and 45-induced inhibition upon washing. HEK293 cells expressing sialin at the plasma membrane were pre- incubated or not with Neu5Ac 1 (5 mM), Fmoc-Leu-OH 13 (120 μM), or 45 (12 μM) for 15 min at pH 5. After a 15 or 30 min wash at Figure 7. Docking of 45 to a DgoT-based homology model of human 3 sialin. (A) Side view of the inward-facing homology model with the pH 7 at room temperature, [ H]Neu5Ac uptake was measured and compared to the signals obtained in the absence or presence of docked inhibitor (in green). (B, C) Residues involved (purple in (B)) competitors at the same concentrations (means ± SEM of four and nature of their interaction with 45 shown in a 2D depiction. independent experiments). t-Test relative to control: ** = p < 0.01; *** = p < 0.001; **** = p < 0.0001; NS, ≥0.05). for the non-competitive inhibition by this compound. The selected pose was further submitted to a 20 ns molecular the cytosolic, rather than the extracellular, side of the dynamics simulation in a membrane environment to optimize transporter. Of note, when this experiment was repeated the 3D structure. with FR139317 2, which inhibits sialin in a competitive The interaction of 45 with the inward-facing sialin model is manner,33 the 15 min wash fully restored [3H]Neu5Ac depicted in Figure 7B,C. The carboxylic group of sialic acids transport without trans-stimulation (Figure S3). The com- and sialin inhibitors, a feature shared with all substrates of petitive versus noncompetitive nature of sialin inhibitors thus SLC17 transporters, is bound to a conserved arginine (R57 in correlates with the rate of Neu5Ac transport recovery upon sialin) and two tyrosines (Y119 and Y306). Similar interactions washing, in agreement with an action on distinct sides of the of D-galactonate with this conserved arginine (R47) and two membrane. The lower lipophilicity of FR139317 (log P = 2.29 other tyrosines (Y44 and Y79) are essential for substrate in the protonated state) as compared to Fmoc-Leu-OH and 45 51 recognition and selectivity in DgoT. 8238 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 8. Eﬀect of compound 45 on the R39C mutant. (A, B) Wild-type or R39C human sialin tagged with EGFP was transiently expressed in HeLa cells by electroporation in the presence or absence of 45 (70 μM). After two days, cells were ﬁxed and analyzed under ﬂuorescence microscopy using LAMP1 immunostaining (red) to detect late endosomes and lysosomes. Sialin localization was categorized as illustrated in (A) in a blind manner. The graph (B) synthesizes the outcome of three independent experiments (162 to 184 cells per condition; χ2 test: ***, p < 0.001; ****, p < 0.0001). (C) In independent experiments, colocalization was quantitated using scatter plots of the sialin and LAMP1 pixel intensities. The graphs show the distribution of Pearson’s correlation coeﬃcient across 20 to 25 cells per condition in a representative example of three independent experiments (Kruskal−Wallis one-way ANOVA with post hoc Dunn’s test: **, p < 0.01;***, p < 0.001). (D) Skin ﬁbroblasts from a healthy subject (WT) and a compound-heterozygote (p.Arg39Cys/Leu336Trpfs) Salla patient were cultured for 2 days with or without 45 (170 μM)andassayed for free sialic acid level by micro-LC/ESI MRM-MS3 (means ± SEM of four independent experiments; t-test: * = p < 0.05; ** = p < 0.01). Two hydrophobic moieties of 45 are expected to block the as molecular hinges for the rotation of the N and C domains alternating-access transition of the transporter: an Fmoc group 49 In DgoT, TM4 has a during alternating-access transitions. on the amino function of the cysteine core and a coumarinyl kink at P135 (P180 in sialin) and TM10 is continuous in the group on the side chain of that amino acid. On one side, the outward-facing conformation whereas TM4 is continuous and 9H-ﬂuorene part of the Fmoc group makes hydrophobic 51 TM10 has a kink in the inward-facing conformation. The interactions with F179, P180, and L426 and several van der tight interactions of 45 with residues close to these kinks Waals contacts with I124, H183, G427, and N430. The (Y179, P180 in TM4; S407 in TM10) in the sialin model carbonyl of the Fmoc group establishes a hydrogen bond with should block alternating-access movements, providing an Q123. On the other side of the ligand, the coumarinyl group explanation for its inhibition of Neu5Ac transport. forms hydrogen bonds with S407, a Π stacking with Y203, and Compound 45 Partially Rescues the Sialin R39C van der Waals contacts with L199 and Y301. In between, F179 Mutant. The R39C mutation causing Salla disease induces a 6,27 contacts with the sulfur atom and carboxylate (Figure 7B,C). traﬃcking defect of sialin possibly due to its destabilization In MFS transporters, helical discontinuities in speciﬁc TMs act and recognition by chaperones of the protein quality control 8239 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article system. We reasoned that 45 binding to the cytosolic side of Two ligands with this scaﬀold, Fmoc-Leu-OH 13 and 45, sialin might stabilize nascent R39C polypeptides in the were characterized in more detail. They inhibited Neu5Ac endoplasmic reticulum and rescue their delivery to lysosomes. transport with IC of 23 and 2.4 μM as compared to an IC 50 50 45 (12 μM) eﬃciently inhibited the residual transport of ∼1 mM for Neu5Ac. Neither compound was translocated activity of sialin R39C in our whole-cell assay (Figure S4), by sialin (Figures S1 and S2), and somewhat surprisingly, both showing that 45 binding is not impaired by the pathogenic inhibited Neu5Ac transport in a non-competitive, rather than a mutation. We thus transiently expressed wild-type (WT) or competitive, manner (Figure 5), notwithstanding the substrate- 33 R39C human sialin fused to EGFP (with an intact sorting binding pocket focus of our virtual screening. In contrast, motif) in HeLa cells in the absence or presence of 45 (70 μM) another sialin inhibitor identiﬁed in this screen, the endothelin- and examined their intracellular distribution by ﬂuorescence A receptor antagonist FR139317, inhibited Neu5Ac transport 33 microscopy. As this distribution varies across cells in a given in a competitive manner. condition (Figure 8A), sialin was compared to a lysosomal This apparent paradox could be explained in a model where marker (LAMP1) and the dual staining in individual cells was the new compounds permeate biological membranes and bind classiﬁed into three categories: lysosomal (≥70% overlap to sialin in a cytosol-facing (inward-facing) rather than lumen- between sialin and LAMP1 puncta), non-lysosomal (≤30% facing (outward-facing) conformation. The protonation of the overlap), and mixed. In agreement with earlier studies, the carboxylate group at pH 5.0 and the high partition coeﬃcient lysosomal category predominated in untreated cells expressing of these compounds facilitate their passive diﬀusion into the WTsialin (83%) whereas this category dropped to 11% for the cells in our assay of sialin transport activity. As structural R39C mutant with a predominance of the mixed (55%) and transition equilibriums are slower than ligand binding non-lysosomal (34%) categories. However, when sialin R39C equilibriums, this mechanism should trap sialin in an inward- was expressed in the presence of 45 (72 μM), the lysosomal facing state (Figure 5C) and prevent extracellular Neu5Ac category increased to 32% at the expense of the non-lysosomal binding to the lumen-facing state, thus decreasing the transport category (15%), while this treatment did not alter the capacity. A similar non-competitive inhibition of a transporter distribution of WT sialin (Figure 8B). Similar results were by ligands of the substrate pocket has been reported for the observed when sialin was expressed by lipofection instead of interaction between the glucose transporter GLUT1 and electroporation (Figure S4). 45 thus partially rescues the cytochalasin B or forskolin.55 An analogous mechanism also traﬃcking defect of the R39C mutant. To conﬁrm this eﬀect, occurs in enzymes undergoing a conformational change when we repeated these experiments and quantitated the colocaliza- inhibitors selectively bind to the product-favoring conforma- tion between EGFP-sialin and LAMP1 using scatter plots and tion.56 Selective or predominant binding of Fmoc-Leu-OH and Pearson’s correlation coeﬃcient analysis of the pixel 45toaninward-facing state of sialin may thus account for their ﬂuorescence intensities. This analysis showed a much lower non-competitive inhibition. level of sialin R39C/LAMP1 colocalization relative to the wild- We provide two pieces of evidence in support of this type. When sialin R39C-expressing cells were treated with a hypothesis. First, inhibition of sialin by Fmoc-Leu-OH and 45 high dose (300 μM) of 45, colocalization was rescued to a level in our whole-cell assay persisted after a 15 min wash (Figure close to that of the wild-type (Figure 8C). 6), in agreement with an action in the cytosolic rather than Next, we tested whether this increased delivery of sialin extracellular compartment. Second, we built a more accurate R39Ctolysosomes could decrease sialic acid storage in patient inward-facing homology model of human sialin based on the + cells. We used Salla ﬁbroblasts from a compound-heterozygote recent crystallographic structures of the H /D-galactonate patient (R39C and L336W + frameshift alleles) because these symporter DgoT.51 Molecular docking of 45 to this model cells accumulate sialic acid to higher levels than homozygous showed that it binds well to the substrate pocket of the inward- R39C cells.54 Compound-heterozygote Salla ﬁbroblasts and facing state, in agreement with our model. However, other control ﬁbroblasts were thus treated or not with 45 (30 to 170 potential mechanisms for the non-competitive inhibition such μM)for2days,andtheir level of free sialic acid was assayed by as the existence of another 45-binding site cannot be excluded. mass spectrometry. Salla ﬁbroblasts cultured in standard Further studies are needed to distinguish between these medium accumulated free sialic acid by ∼8-fold as compared possibilities. to control ﬁbroblasts. The 45 treatment did not reduce but The best docked pose of 45 showed several interesting instead slightly increased this accumulation (Figure 8C), features, including an interaction of its carboxylate group with probably reﬂecting an inhibition of sialin R39C by 45 at the a conserved arginine in TM1 (R57 in human sialin) that is 51 lysosomal membrane. To circumvent this eﬀect, we used a required for D-galactonate binding in DgoT. 45 may thus pulse-chase protocol in which the drug was applied for 2 days share common interactions with the anionic substrates of to correct the traﬃcking defect followed by a 6 h chase in drug- sialin. Another notable feature is the interaction of 45 with free medium to remove lysosomal inhibition. However, this residues in TM4 and TM10 that act as hinge regions for 51 treatment did not reverse sialic acid storage (data not shown). alternating-access structural transitions in DgoT. 45 binding to human sialin should thus impair its structural transitions, ■ DISCUSSION explaining why this compound blocks Neu5Ac transport. The identiﬁcation of 45 as a cell-permeant ligand of the In this study, we exploited our previous virtual screening of “active” site (sialic acid pocket) of sialin prompted us to test 33 human sialin and report chemical substitutions of validated whether it might act as a pharmacological chaperone proof-of- virtual hits. This led to the identiﬁcation of a novel ligand principle for the treatment of Salla disease. Pharmacological scaﬀold unrelated to sugar substrates, which is characterized by chaperones are selective ligands that bind to and stabilize an amino acid backbone, a free carboxylate, a N-linked misfolded mutant proteins to rescue their retention by the aromatic or heteroaromatic substituent, and a hydrophobic 57,58 protein quality control system. This approach is a side chain. promising option for Salla disease as this condition is almost 8240 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article exclusively caused by a single mutation, R39C,18 which method A-2 using solvent A (KH PO buﬀer, 50 mM, pH 5.5) and 2 4 solvent B (CH CN): 0% B for 5 min, a linear increase from 0 to 50% partially preserves the lysosomal targeting and the transport 3 6,7 B between 5 and 20 min. HPLC-MS analyses were performed on a activity of sialin. We thus applied 45 to sialin R39C- Surveyor HPLC system coupled to a LCQ Advantage Thermo transfected cells and Salla patient ﬁbroblasts to test its potential Finnigan LCQ Advantage Instrument. HPLC was equipped with a beneﬁts. Interestingly, this treatment partially rescued the Gemini C18 column (100 mm × 2.1 mm, 3 μm), ﬂow: 220 μL/min. traﬃcking defect of sialin R39C. However, it did not rescue The dead volume was approximately 800 μL. Products were eluted sialic acid storage in patient ﬁbroblasts, presumably because with method B using solvent A (H O/0.1% HCO H) and solvent B the rescue of the traﬃcking defect remained insuﬃcient and/or 2 2 (MeOH): 40% B for 1 min, a linear increase from 40 to 100% B this eﬀect was masked by the inhibition of sialin R39C at the between 1 and 9 min, 100% B from 10 to 13 min, and with method C lysosomal membrane. The slight increase of sialic acid storage using solvent A (H O/0.1% HCO H) and solvent B (CH CN/0.1% 2 2 3 upon 45 treatment supports the latter hypothesis. In this HCO2H): 40% B for 1 min, a linear increase from 40 to 100% B respect, the lack of competition between Neu5Ac accumulated between 1 and 6 min, 100% B from 7 to 13 min. The purity of the in the lumen of patient lysosomes and 45 binding at the tested compounds was established by analytical HPLC-MS and 1 13 cytosolic face of sialin would be a disadvantage for therapeutic HPLC and was at least 95%. Spectroscopic ( H and CNMR,MS) applications. and/or analytical data were obtained using chromatographically In summary, our study identiﬁes a new class of cell-permeant homogeneous samples. The photophysical properties of the ﬁnal inhibitors with a micromolar aﬃnity for sialin, providing compounds were obtained by diﬀerent measurements on the Greiner and Nonchuk 96-well plates using a Tecan spectroﬂuorometer valuable tools to study the diverse transport activities of sialin (Saﬁre). These measurements were carried out in solution in ethanol and their physiological roles. These tools may also help at 25 °C. mechanistic and structural studies of sialin. They may also be Syntheses. General Procedure A for the Synthesis of Compounds 29, 30, and 31. A 0.5 M solution of ClTi(OiPr) (1.5 used to prevent Neu5Gc incorporation into therapeutic 3 glycoproteins produced by cell culture. In the case of Salla equiv) in dry toluene was added to aminophenol (1 equiv) and β- disease, however, further studies are needed to explore the cetoester (1 equiv) derivatives. The reaction mixture was reﬂuxed overnight then cooled to room temperature and diluted with CH Cl potential of pharmacological chaperone therapy. 2 2 (7.5 mL/mmol). The whole solution was poured into H O (10 mL/ 2 EXPERIMENTAL SECTION mmol) and stirred for few minutes. The aqueous layer was extracted ■ twice with CH Cl (2 × 30 mL/mmol). The combined organic layers 2 2 General Chemistry Information. All the amino acids belong to were dried (Na SO ), and the solvent was removed under vacuum. 2 4 the L-series unless speciﬁed otherwise. Amino acid derivatives are Thecrude product was puriﬁed or used directly to another reaction as designated according to the three-letter code and the recommenda- mentioned in each case. tions from the IUPAC-IUB Commission on Biochemical Nomencla- General Procedure B for the Synthesis of Compounds 37−39, ture. Fmoc-Lys(Cbz)-OH 3, Fmoc-Leu-OH 13, Fmoc-Asp(tBu)-OH 41, and 44. Heteroaromatic derivatives (1 equiv), amino acid 14, Fmoc-Cys(tBu)-OH 15, Fmoc-Tyr(tBu)-OH 17, Fmoc-Phe-OH derivatives (1−1.2 equiv), and HBTU (1.1−1.2 equiv) were 18, Fmoc-Ile-OH 19, Fmoc-Cys(pMeOBzl)-OH 20, Fmoc-Thr(tBu)- suspended in CH Cl (0.1 M) then DIEA (4−6 equiv) was added 2 2 OH 21, and Fmoc-Pro-OH 23 were purchased from Novabiochem dropwise at 0 °C, and the reaction mixture was stirred for 2 h at room and used for biological test. Glu(OMe)-OH, DL-Leu-OH, Fmoc- temperature. The solution was diluted with CH Cl (50 mL/mmol) 2 2 Cys(Trt)-OH, Cys(Bzl)-OH, Fmoc-Lys(Boc)-OH, and HCl·Lys- and washed three times with aqueous HCl solution (0.1 M, 20 mL/ (Cbz)-OMe were purchased from Novabiochem and functionalized. mmol). Aqueous layers were extracted twice with CH Cl (50 mL/ 2 2 The other reagents were purchased from Aldrich or Acros. Prior to mmol) then the organic layers were pooled, dried (Na SO ), ﬁltered, 2 4 use, tetrahydrofuran (THF) was distilled from sodium benzophenone and evaporated. The mixture was treated with aqueous LiOH solution and dichloromethane (CH Cl ) from CaH . All reactions were carried (0.5 M, 1.2 equiv)/THF (1/1) and stirred for 2 h at room 2 2 2 temperature. The solution was evaporated and diluted with H O then out under an argon atmosphere and monitored by thin-layer 2 chromatography with Merck 60F-254 precoated silica (0.2 mm) on acidiﬁed to pH 2 with HCl (1 M). The precipitate was ﬁltered and, glass. Flash chromatography was performed with Merck Kieselgel 60 when necessary, puriﬁed by column chromatography on silica gel (200−500 mm); the solvent systems were given in v/v. 1H NMR (CH Cl /MeOH/AcOH: 95/5/0.1 to 85/15/1). 2 2 13 4-Carboxymethyl-7-dimethylamino-coumarin (29). General pro- (500 MHz) and C NMR (126 MHz) spectra were recorded on a Bruker AVANCEII-500 spectrometer. Chemical shifts (δ)are cedure A was followed using ClTi(OiPr) (1.791 mL, 7.5 mmol) in 3 reported in parts per million. Multiplicity was given using the toluene (12 mL), 3-dimethylaminophenol (0.685 g, 5 mmol), and following abbreviations: s (singlet), brs (broad singlet), d (doublet), dimethyl 3-oxoglutarate (1.791 mL, 5 mmol). The crude product was dd (doublet of doublets), t (triplet), q (quadruplet), and m dissolved in MeOH (12.5 mL) and aqueous NaOH solution (1 M, (multiplet). During acquisition, the spectral window covers a proton 12.5 mL) and stirred overnight at room temperature. The solution chemical shift range from −1 to +12 ppm or from −5 to +20 ppm. was acidiﬁed to pH 5 with aqueous HCl solution (2 M), and the 13 1 resulting precipitate was recovered by ﬁltration, washed with HCl (0.1 C chemical attributions were assigned using H-decoupled spectra. For clarity, in some case, Greek letters are used as locants for NMR M), and dried under vacuum. Compound 29 (0.487 g) was obtained attribution of the side chain of the amino acid, while the heterocycle as a green-yellow powder in a 40% yield. 1H NMR (500 MHz, moiety is numbered according to the IUPAC nomenclature. Melting (CD ) SO): δ 7.47 (d, J = 9.0 Hz, 1H, H-5), 6.72 (dd, J = 9.0 Hz, J = 3 2 ̈ 2.5 Hz, 1H, H-6), 6.56 (d, J = 2.5 Hz, 1H, H-8), 6.05 (s, 1H, H-3), points were determined with a Buchi 530 apparatus and are 13 uncorrected. Mass spectra (MS) were recorded on a Thermo 3.77 (s, 2H, CH ), 3.01 (s, 6H, CH ); C NMR (126 MHz, 2 3 Finnigan LCD Advantage spectrometer and HRMS on an Exactive (CD ) SO): δ 170.7 (CO H), 160.6 (CO), 155.4 (Cq-O), 152.8 3 2 2 + (Cq-Ar), 150.2 (Cq-N), 126.0, 109.6, 109.1 (CH-Ar), 108.0 (Cq-Ar), (Thermo Scientiﬁc) spectrometer with positive (ESI ) or negative − + (ESI ) electrospray ionization. HPLC analyses were carried out on a 97.5 (CH-Ar), 40.1 (2 × NCH ), 37.2 (CH ); MS: (ESI ), m/z (%): + 3 + 2 max Prominence Shimadzu instrument with an LC20A pump, C18 column [M + H] = 248.3 (100%); [2M + Na] = 517.0 (80%); λ = 373 max −1 −1 abs (250 mm × 4.6 mm, 5 μm), ﬂow: 1 mL/min; eluted peaks were nm, λ = 443 nm, ε (λ ) = 23,300 M cm , Φ = 0.677. em max detected by a PDA detector (SPD-M20A), and retention times are 4-Carboxymethyl-7-diethylamino-coumarin (30). General proce- reported in minutes. The dead volume is approximately 200 μL. dure A was followed using ClTi(OiPr)3 (3.582 mL, 15 mmol) in Products were eluted with method A-1 using solvent A (KH PO4 toluene (25 mL), 3-diethylaminophenol (1.652 g, 10 mmol), and 2 buﬀer, 50 mM, pH 5.5) and solvent B (CH3CN): 0% B for 10 min, a dimethyl 3-oxoglutarate (1.470 mL, 10 mmol). The crude product linear increase from 0 to 50% B between 10 and 25 min, or with was dissolved in MeOH (25 mL) and aqueous NaOH solution (1 M, 8241 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article 25 mL) then stirred overnight at room temperature. The solution was (Cq-7), 156.4 (Cq-4), 154.4 (Cq-8a), 125.4 (CH-5), 113.1 (CH-6), acidiﬁed to pH 5 with aqueous HCl solution (2 M), and the resulting 111.5 (Cq-4a), 103.2, 103.1 (CH-3, CH-8), 93.9 (OCH O), 59.0 2 + + precipitate was recovered by ﬁltration, washed with HCl (0.1 M), and (CH OH), 58.9 (CH ); MS: (ESI ), m/z (%): [M + H] = 237.1 2 3 dried under vacuum. Compound 30 (0.947 g) was obtained as a green (100%). powder in a 35% yield. 1H NMR (500 MHz, (CD ) SO): δ 7.44 (d, J 7-(Methoxymethoxy)-4-(Fmoc-oxymethyl)-coumarin (34). 3 2 = 9.0 Hz, 1H, H-5), 6.69 (dd, J = 9.0 Hz, J = 2.5 Hz, 1H, H-6), 6.51 Fmoc-Cl (0.206 g, 0.8 mmol, 1.2 equiv) was added by portion over (d, J = 2.5 Hz, 1H, H-8), 5.99 (s, 1H, H-3), 3.75 (s, 2H, CH CO H), 30 min to a solution of compound 33 (0.160 g, 0.68 mmol, 1 equiv) 2 213 3.41 (q, J = 7.0 Hz, 4H, NCH ), 1.11 (t, J = 7.0 Hz, 6H, CH ); C in dry pyridine (3 mL). After stirring for 2 h at room temperature, the 2 3 NMR (126 MHz, (CD ) SO): δ 170.7 (CO H), 160.7 (CO), reaction mixture was diluted with EtOAc (20 mL) and washed with 3 2 2 155.8 (Cq-O), 150.4 (Cq-Ar), 150.1 (Cq-N), 126.3, 109.1, 108.7 brine (3 × 20 mL). The organic layer was dried, ﬁltered, and (CH-Ar), 107.6 (Cq-Ar), 96.8 (CH-Ar), 44.0 (2 × NCH ), 37.2 evaporated under vacuum. The residue was puriﬁed by silica-gel + 2 + (CH CO H), 12.3 (2 × CH ); MS: (ESI ), m/z (%): [M + H] = column chromatography (cyclohexane/EtOAc: 9/1 to 7/3). Com- 2 2 3 + + 276.3 (100%); [2M + Na] = 573.0 (85%); [2M + H] = 550.9 pound 34 (0.108 g) was obtained as a white powder in a 34% yield. (70%). 1H NMR(500 MHz, CDCl ): δ 7.75 (d, J = 7.5 Hz, 2H, H-Ar), 7.59 2-(11-Oxo-2,3,6,7-tetrahydro-1H,5H,11H-pyrano[2,3-f]pyrido- 3 (d, J = 7.5 Hz, 2H, H-Ar), 7.40 (m, 3H, H-Ar), 7.31 (t, J = 6.5 Hz, [3,2,1-ij]quinolin-9-yl)acetic acid (31). From ClTi(OiPr)3 (2.262 2H, H-Ar), 7.03 (d, J = 2.5 Hz, 1H, H-8), 6.97 (dd, J = 9.0 Hz, J = 2.5 mL, 6.75 mmol) in toluene (9 mL), 8-hydroxyjulodin (0.851 g, 4.5 Hz, 1H, H-6), 6.38 (t, J = 1.5 Hz, 1H, H-3), 5.29 (d, J = 1.5 Hz, 2H, mmol), and dimethyl 3-oxoglutarate (681 μL, 4.5 mmol), compound CH-Coum), 5.22 (s, 2H, OCH O), 4.48 (d, J = 7.0 Hz, 2H, 31asmethylester was obtained according to general procedure A and 2 2 13 CHCH), 4.26 (t, J = 7.0 Hz, 1H, CH CH), 3.47 (s, 3H, CH ); C used without further puriﬁcation in the next step. 1H NMR (500 2 2 3 NMR (126 MHz, CDCl3): δ 160.7, 160.6 (CO, Cq-O), 155.5 MHz, CDCl3): δ 6.99 (s, 1H, H-8), 5.98 (s, 1H, H-10), 3.70 (s, 3H, (OCO2), 154.7 (Cq-OMOM), 148.4 (Cq-Ar), 143.2, 141.5 (2 × Cq- CH), 3.61 (s, 2H, CH CO CH ), 3.26 (q, J = 6.0 Hz, 4H, CH -N), 3 2 2 3 2 Ar), 128.2, 127.4, 125.3 (2 × CH-Ar), 124.6 (CH-Ar), 120.3 (2 × 2.89 (t, J = 6.5 Hz, 2H, CH -Ph), 2.79 (t, J = 6.5 Hz, 2H, CH -Ph), 2 2 CH-Ar), 113.8 (CH-Ar), 111.5 (Cq-Ar), 111.1, 104.5 (CH-Ar), 94.6 1.98 (m, 4H, CH -CH -CH ). Methyl ester 31 was then dissolved in 2 2 2 (OCHO), 70.7 (CH CH), 64.7 (CH -Coum), 56.6 (CH ), 46.9 2 2 2 3 MeOH(15mL)andaqueousNaOHsolution(1M,15mL),andthe + + (OCHCH); MS: (ESI ), m/z (%): [M + H] = 459.3 (100%). mixture was stirred overnight at room temperature. The pH was 2 7-Hydroxy-4-(Fmoc-oxymethyl)-coumarin (35). Compound 34 adjusted to 2 with aqueous HCl solution (1 M), and the resulting (0.090 g, 0.2 mmol) was suspended in CH Cl (1.5 mL) at 0 °C, and precipitate was recovered by ﬁltration and washed with aqueous HCl 2 2 TFA (1.5 mL) was added. The solution was stirred 1.5 h at room solution (0.1 M, 30 mL). Compound 31 (0.890 g) was obtained as a temperature. After evaporation of the solvents, the residue was yellow powder in a 66% yield. mp: 186−187 °C; 1H NMR (500 MHz, puriﬁed by silica gel column chromatography (cyclohexane/EtOAc: (CD ) SO): δ 12.67 (brs, 1H, CO H), 7.05 (s, 1H, H-8), 5.94 (s, 1H, 3 2 2 9/1 to 7/3). Compound 35 (0.069 g) was obtained as a white powder H-10), 3.71 (s, 2H, CH CO H), 3.24 (m, 4H, CH -N), 2.71 (m, 4H, 1 2 2 2 in an 83% yield. H NMR (500 MHz, (CD ) SO): δ 10.63 (s, 1H, CH-Ph), 1.88 (m, 4H, CH -CH -CH ); 13C NMR (126 MHz, 3 2 2 2 2 2 OH), 7.89 (d, J = 7.5 Hz, 2H, H-Ar), 7.66 (d, J = 7.5 Hz, 2H, H-Ar), (CD ) SO): δ 170.8 (CO H), 160.7 (CO), 150.7, 150.1, 145.5 3 2 2 7.51 (d, J = 8.5 Hz, 1H, H-5), 7.41 (t, J = 7.5 Hz, 2H, H-Ar), 7.32 (t, J (Cq-Ar), 122.0 (CH-Ar), 117.7 (Cq-Ar), 108.3 (CH-Ar), 107.4, =7.5 Hz, 2H, H-Ar), 6.81 (dd, J = 8.5 Hz, J = 2.5 Hz, 1H, H-6), 6.75 105.6 (Cq-Ar), 49.2, 48.6 (CH -N), 37.2 (CH CO H), 27.0, 20.9, 2 2 2 (d, J = 2.5 Hz, 1H, H-8), 6.08 (s, 1H, H-3), 5.33 (s, 2H, CH -Coum), + + 2 20.0, 19.9 (CH2); MS: (ESI ), m/z (%): [M + H] = 300.1 (80%); 4.61 (d, J = 6.0 Hz, 2H, OCH CH), 4.34 (t, J = 6.0 Hz, 1H, [2M + Na]+ = 621.1 (100%). 13 2 OCHCH); CNMR(126MHz,(CD)SO):δ161.4,159.9, 155.0 3-(3,6-Dihydroxy-xanthen-9-ylidene)propanoic Acid (32). A 2 3 2 (Cq-O), 153.9 (OCO ), 149.7 (Cq-Ar), 143.2 (2 × Cq-Ar), 140.8 (2 stirred mixture of succinic anhydride (2.50 g, 25 mmol, 0.5 equiv) 2 ×Cq-Ar), 127.7, 127.1 (2 × CH-Ar), 125.9 (CH-Ar), 124.8, 120.2 (2 and resorcinol (2.75 g, 25 mmol, 1 equiv) in aqueous H SO solution 2 4 × CH-Ar), 113.0 (CH-Ar), 108.9 (Cq-4a), 108.4 (CH-3), 102.4 (70%, 30 mL) was heated to 140 °C for 4 h. The reaction mixture was (CH-8), 69.2 (OCH CH), 64.6 (CH -Coum), 46.3 (OCH CH); then cooled to room temperature and poured into H O (500 mL). 2 2 + 2 2 HPLC-MS (method B): tR = 14.1 min; MS: (ESI ), m/z (%): [M + The stirred solution was alkalinized to pH 12 with aqueous NaOH + + − H] = 415.1 (100%); [2M + Na] = 850.9 (60%); HRMS (ESI ) solution (50%), while the temperature was kept at 0 °C. Acetic acid − max calcd for [C H O − H] : 413.1030; found: 413.1040; λ = 324 was added to the solution until pH 4, and the resulting brown max 25 18 6 −1 −1 abs nm, λ = 399 nm, ε (λ ) = 4760 M cm , Φ = 0.243. precipitate was ﬁltered. The ﬁltrate was washed with H O(3× 25 6 em max 2 2 N -((Benzyloxy)carbonyl-N -(2-(7-hydroxy-2-oxo-2H-chromen-4- mL) and acetone (15 mL) and dried under reduced pressure to give yl)acetyl)-L-lysine (36). 7-Hydroxy-4-carboxymethyl-coumarin (0.220 pure compound 32 (3.28 g) as an orange solid in a 92% yield from g, 1 mmol, 1 equiv), HBTU (0.417 g, 1.1 mmol, 1.1 equiv), and resorcinol. 1H NMR (500 MHz, (CD ) SO): δ 7.51 (d, 1H, J = 8.5 HOBt(0.148g,1.1mmol,1.1equiv) werediluted in anhydrous DMF 3 2 Hz, H-Ar), 7.41 (d, 1H, J = 8.5 Hz, H-Ar), 6.65−6.47 (m, 4H, H-Ar), (3 mL). DIEA (380 μL, 2.2 mmol, 2.2 equiv) was added dropwise, 5.87 (t, 1H, J = 7.0 Hz, CCH), 3.31 (d, 2H, J = 7.0 Hz, CH2-CH); and the resulting solution was stirred for a few minutes then a solution 13 C NMR (126 MHz, (CD ) SO): δ 173.5 (CO H), 158.2, 157.9 of HCl·Lys(Cbz)-OMe (0.331 g, 1 mmol, 1 equiv) in DMF (1 mL) 3 2 2 (Cq-OH), 153.0, 151.3 (Cq-Ar), 128.6 (CH-Ar), 125.7 (Cq-Ar), was added. The reaction mixture was stirred overnight at room 124.3, 116.1 (CH-Ar), 114.0, 113.1 (Cq-Ar), 112.0 (CH-Ar), 110.8 temperature, and DMF was evaporated. The residue was diluted in + (CH), 102.6, 102.3 (CH-Ar), 36.3 (CH ); MS: (ESI ), m/z (%): EtOAc (50 mL), washed with brine (3 × 40 mL), dried (Na SO ), + max 2 max 2 4 [M + H] = 285.1 (100%); λ = 455 nm, λ = 516 nm, ε (λ )= ﬁltered, and evaporated. The ester compound was diluted with THF −1 −1 abs em max 5140 M cm , Φ = 0.243. (4 mL) and a solution of aqueous LiOH (0.5 M, 4 mL) and stirred 2 4-Hydroxymethyl-7-methoxymethoxy-coumarin (33). Com- h at room temperature. After evaporation of THF, the pH was pound 27 (0.576 g, 3 mmol, 1 equiv) was suspended in CH Cl at adjusted to 2 with aqueous HCl solution (1 M) and extracted with 2 2 0°C.DIEA(272μL,3.6mmol,1.2equiv)andMOM-Cl(627μL,3.6 EtOAc (2 × 40 mL). Organic layers were dried (Na SO ), ﬁltered, 2 4 mmol, 1.2 equiv) were added dropwise, and the solution was stirred and evaporated. The residue was puriﬁed by column chromatography for 45 min at 0 °C. After evaporation under reduced pressure, H O on silica gel (CH Cl /MeOH/AcOH: 85/15/1). Compound 36 2 2 2 (10 mL) was added to the residue and the resulting precipitate was (0.090 g) was obtained as a white powder in a 19% yield. mp: 78−79 1 recovered by ﬁltration and dried under vacuum. Compound 33 °C; HNMR(500MHz,CD OD):δ7.47(d,J=8.5Hz,1H,H-Ar), 1 3 (0.511 g) was obtained as a white solid in a 71% yield. H NMR (500 7.17 (m, 5H, H-Ph), 6.64 (d, J = 8.5 Hz, 1H, H-Ar), 6.54 (s, 1H, H- MHz, (CD ) SO): δ 7.63 (d, J = 8.5 Hz, 1H, H-5), 7.04 (d, J = 2.5 Ar), 6.06 (s, 1H, H-Ar), 4.89 (s, 2H, CH -Ph), 4.19 (m, 1H, H-α), 3 2 2 Hz, 1H, H-8), 7.00 (dd, J = 8.5 Hz, J = 2.5 Hz, 1H, H-6), 6.32 (s, 1H, 3.60 (s, 2H, CH2-Coum), 2.93 (m, J = 6.5 Hz, 2H, H-ε), 1.70, 1.56 13 H-3), 5.30 (s, 2H, OCH O), 4.72 (s, 2H, CH OH), 3.39 (s, 3H, (2m, 2H, H-β), 1.34−1.25 (m, 4H, H-γ,H-δ); CNMR(126MHz, 13 2 2 CH); C NMR (126 MHz, (CD ) SO): δ 160.3 (CO), 159.4 CDOD): δ 175.2 (CO H), 170.8 (CONH), 163.1, 163.0, 158.9 3 3 2 3 2 8242 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article 13 (CO-Coum), 156.8 (NHCO ), 152.6, 138.7 (Cq-Ar), 129.7 (2 × (2m, 2H, H-β), 1.46−1.34 (m, 4H, H-γ,H-δ); CNMR(126MHz, 2 CH-Ar), 129.1 (CH-Ar), 128.9 (2 × CH-Ar), 127.9, 114.4 (CH-Ar), (CD ) SO): δ 173.0 (CO H), 165.3 (CONH-Ar), 161.4 (CO-Ar) 3 2 2 113.2 (CH-Ar, Cq-Ar), 103.8 (CH-Ar), 67.3 (OCH -Ph), 66.9 (CH - 156.0 (NHCO ), 146.6 (Cq-Ar), 139.2 (Cq-Ar), 137.3 (Cq-Ar), 2 2 2 Coum), 54.0 (C-α), 41.6 (C-ε), 32.2 (C-β), 30.6 (C-δ), 24.3 (C-γ); 130.6 (CH-Ar), 128.3 (2 × CH-Ar), 127.6 (3 × CH-Ar), 126.2 (Cq- HPLC (method A-2): tR = 13.2 min; HPLC-MS (method B): tR = Ar), 121.9 (CH-Ar), 119.5 (CH-Ar), 116.4 (CH-Ar), 115.5 (CH-Ar), + + 11.8 min; MS: (ESI ), m/z (%): [M + H] = 483.1 (100%), [M − 65.1 (OCH -Ph), 53.6 (C-α), 40.2 (C-ε), 31.1 (C-β), 29.2 (C-δ), + − − 2 CO +H] =439.4(20%);MS:(ESI ),m/z(%):[M−H] =481.3 23.0 (C-γ); HPLC (method A-2): t = 12.8 min; HPLC-MS (method 2 − − R (100%), [M − C H O − H] = 373.2 (50%); HRMS (ESI ) calcd + + 7 8 − max B): tR = 11.8 min; MS: (ESI ), m/z (%): [M + H] = 452.3 (100%), for [C H N O − H] : 481.1616; found: 481.1619; λ = 324 nm, + − − max 25 26 2 8 −1 −1 abs [M−CO2+H]=408.3(20%);MS:(ESI ),m/z(%):[M−H] = λ = 399 nm, ε (λ ) = 9880 M cm , Φ = 0.0904. 450.6 (20%); [2M − H]− = 902.0 (100%); HRMS (ESI−) calcd for em 6 max 2 N -((Benzyloxy)carbonyl-N -(2-(7-(dimethylamino)-2-oxo-2H- − max [C H NO − H] : 450.1670; found: 450.1669; λ = 332 nm, 24 25 3 6 abs chromen-4-yl)acetyl)-L-Lysine (37). From 7-dimethylamino-4-car- λmax= 434 nm, ε (λ ) = 16,700 M−1 cm−1, Φ = 0.0101. boxymethyl-coumarin 29 (0.125 g, 0.5 mmol, 1 equiv), HCl· em 6 max 2 N -((Benzyloxy)carbonyl-N -(3-(3,6-dihydroxy-9H-xanthen-9- Lys(Cbz)-OMe (0.165 g, 0.6 mmol, 1.1 equiv), HBTU (0.220 g, ylidene)propanoyl)-L-lysine (40). From compound 32 (0.384 g, 1.0 0.6 mmol, 1.1 equiv), and DIEA (346 μL, 2.0 mmol, 4 equiv), mmol, 1 equiv), HBTU (0.417 g, 1.1 mmol, 1.1 equiv), and HOBt compound37(0.128g) was obtained by ﬁltration as a yellow powder (0.148 g, 1.1 mmol, 1.1 equiv) were diluted with anhydrous DMF (3 1 in a 50% yield according to general procedure B. H NMR (500 MHz, mL). DIEA (380 μL, 2.2 mmol, 2.2 equiv) was added dropwise, and (CD ) SO): δ 8.55 (d, J = 7.0 Hz, 1H, NH), 7.56 (d, J = 9.0 Hz, 1H, 3 2 this solution was stirred for a few minutes then a solution of HCl· H-5), 7.30 (m, 6H, H-Ph, NH), 6.66 (d, J = 9.0 Hz, 1H, H-6), 6.54 (s, Lys(Cbz)-OMe (0.331 g, 1.0 mmol, 1 equiv) in DMF (1 mL) was 1H, H-8), 6.04 (s, 1H, H-3), 5.00 (s, 2H, CH -Ph), 4.16 (m, 1H, H- 2 added. The solution was stirred overnight at room temperature then α), 3.67 (s, 2H, CH -Coum), 3.00 (m, 8H, H-ε,CH), 1.71, 1.61 2 13 3 the solvent was evaporated, and the residue was diluted in EtOAc (50 (2m, 2H, H-β), 1.39−1.30 (m, 4H, H-γ,H-δ); CNMR(126MHz, mL). The organic layer was washed with brine (3 × 40 mL), dried (CD ) SO): δ 173.4 (CO H), 168.0 (CONH) 160.7, 156.0 (CO- 3 2 2 (Na SO ), ﬁltered, and evaporated. The mixture was diluted with a Coum), 155.3 (NHCO ), 152.8 (Cq-Ar), 151.3 (Cq-N(Me) ), 137.7 2 4 2 2 solution of aqueous LiOH (0.5 M)/THF (40 mL, 1/1) and stirred 2 (Cq-Ar), 128.3 (2 × CH-Ar), 127.7 (3 × CH-Ar), 126.0 (CH-Ar), h at room temperature. After evaporation of THF, the pH was 109.3 (CH-Ar), 108.9 (CH-Ar), 108.2 (Cq-Ar), 97.5 (CH-Ar), 65.1 adjusted to 2 with aqueous HCl solution (1 M) and extracted with (OCH-Ph), 52.00 (C-α), 40.1 (C-ε, 2xNCH ), 38.4 (CH -Coum), 2 3 2 EtOAc (2 × 40 mL). The organic layer was dried (Na SO ), ﬁltered, 30.6 (C-β), 29.0 (C-δ), 22.7 (C-γ); HPLC (method A-2): t = 14.4 2 4 + R and evaporated. The residue was puriﬁed by column chromatography min; HPLC-MS (method B): tR = 13.7 min; MS: (ESI ), m/z (%): on silica gel (CH Cl /MeOH/AcOH: 85/15/1). Compound 40 + + 2 2 [M + H] = 510.1 (100%), [M − CO +H] = 466.3 (20%); MS: − − 2 − (0.045 g) was obtained as a red powder in an 8% yield. mp: 106−107 (ESI ), m/z (%): [M − H] = 508.3 (100%): HRMS (ESI ) calcd °C; 1H NMR (126 MHz, (CD ) SO): δ 7.51 (d, J = 8.5 Hz, 1H, H- − max 3 2 for [C H N O − H] : 508.2089; found: 508.2089; λ = 373 nm, 27 31 3 7 abs Ar), 7.41 (d, J = 8.5 Hz, 1H, H-Ar), 7.15 (m, 5H, H-Ph), 6.59−6.45 max −1 −1 λ = 448 nm, Φ = 0.714, ε (λ ) = 16,800 M cm . em max (m, 4H, H-Ar), 5.78 (t, J = 7.0 Hz, 1H, CCH), 4.69 (s, 2H, CH - N6-((Benzyloxy)carbonyl-N2-(2-(7-(diethylamino)-2-oxo-2H- 2 Ph), 4.25 (m, 1H, H-α), 3.31 (d, J = 7.0 Hz, 2H, CH -CH), 2.88 (m, chromen-4-yl)acetyl)-L-lysine (38). From 7-diethylamino-4-carbox- 2 13 ymethyl-coumarin 30 (0.275 g, 1.0 mmol, 1 equiv), HCl·Lys(Cbz)- 2H, H-ε), 1.70, 1.55 (m, 2H, H-β), 1.29−1.18 (m, 4H, H-γ,H-δ); C NMR (126 MHz, (CD ) SO): δ 172.9 (CO H), 164.8 (CONH), OMe(0.360 g, 1.2 mmol, 1.2 equiv), HBTU (0.440 g, 1.2 mmol, 1.2 3 2 2 equiv), and DIEA (865 μL, 5.0 mmol, 5 equiv), compound 38 (0.128 158.2, 157.9 (Cq-3, Cq-6), 156.0 (NHCO), 153.0 (Cq-Ar), 151.3 g) was obtained by ﬁltration as a yellow powder in a 19% yield (Cq-Ar), 138.7 (Cq-Ar), 128.3 (CH-Ar), 128.1 (2 × CH-Ar), 127.2 according to general procedure B. 1H NMR (500 MHz, (CD ) SO): (3 × CH-Ar), 125.7 (Cq-Ar), 123.3, 115.7 (CH-Ar), 114.0, 113.1 3 2 (Cq-Ar), 112.0 (CH-Ar), 110.8 (CH), 103.0 (CH-Ar), 102.4 (CH- δ 8.54 (d, J = 7.5 Hz, 1H, NH), 7.52 (d, J = 9.0 Hz, 1H, H-5), 7.30 Ar), 66.1 (OCH Ph), 52.6 (C-α), 40.6 (C-ε), 36.3 (CH CO), 31.0 (m, 6H, H-Ph, NH), 6.66 (dd, J = 9.0 Hz, J = 2.5 Hz, 1H, H-6), 6.54 2 2 (d, J = 2.5 Hz, 1H, H-8), 5.99 (s, 1H, H-3), 5.00 (s, 2H, CH2-Ph), (C-β), 29.1 (C-δ), 22.6 (C-γ); HPLC (method A-2): tR = 13.7 and 15.1 min; HPLC-MS (method B): t = 12.1 and 12.9 min (2 4.15 (m, 1H, H-α), 3.65 (s, 2H, CH -Coum), 3.41 (q, J = 7.0 Hz, 4H, R 2 + + N(CH CH ) ), 2.96 (m, 2H, H-ε), 1.72, 1.60 (2m, 2H, H-β), 1.45− tautomers); MS: (ESI ), m/z (%): [M + H] = 547.2 (100%); MS: 2 3 2 13 − − − (ESI ), m/z (%): [M − H] = 545.3 (100%), [M − C H O − H] = 1.26 (m, 4H, H-γ,H-δ), 1.10 (t, J = 7.0 Hz, 6H, N(CH CH ) ); C 7 7 2 3 2 − − 438.2 (30%); HRMS (ESI ) calcd for [C H N O − H] : NMR (126 MHz, (CD ) SO): δ 173.4 (CO H), 168.0 (CONH), 30 30 2 8 3 2 2 max max 545.1929; found: 545.1956; λ = 455 nm, λ = 516 nm, ε (λ ) 160.7, 156.0 (CO-Coum), 155.7 (NHCO2), 151.2 (Cq-Ar), 150.3 −1 −1 abs em max (Cq-N(Me) ), 137.3 (Cq-Ar), 128.3 (2 × CH-Ar), 127.7 (3 × CH- = 5140 M cm , Φ = 0.243. 2 N6-((Benzyloxy)carbonyl-N2-(2-(11-oxo-2,3,6,7-tetrahydro- Ar), 126.3 (CH-Ar), 108.8 (CH-Ar), 108.5 (CH-Ar), 107.7 (Cq-Ar), 1H,5H,11H-pyrano[2,3-f]pyrido[3,2,1-ij]quinolin-9-yl))acetyl)-L-ly- 96.8 (CH-Ar), 65.1 (OCH -Ph), 52.0 (C-α), 44.0 (2 × NCH CH ), 2 2 3 sine (41). From compound 31 (0.150 g, 0.50 mmol, 1 equiv), HCl· 40.1 (C-ε), 38.4 (CH -Coum), 30.6 (C-β), 29.0 (C-δ), 22.7 (C-γ), 2 Lys(Cbz)-OMe (0.182 g, 0.55 mmol, 1.1 equiv), HBTU (0.209 g, 12.3 (2 × NCH CH ); HPLC (method A-2): t = 15.9 min; HPLC- 2 3 + R + 0.55 mmol, 1.1 equiv), and DIEA (346 μL, 2.0 mmol, 4 equiv), MS(method B): tR = 14.7 min; MS: (ESI ), m/z (%): [M + H] = + + compound 41 (0.033 g) was obtained by ﬁltration as a yellow powder 538.1 (100%), [2M + H] = 1074.9 (90%), [M − CO +H] = 494.4 − 2 − in a 12% yield according to general procedure B. 1H NMR (500 MHz, (20%); HRMS (ESI ) calcd for [C H N O − H] : 536.2402; max max 29 35 3 7 −1 CDOD): δ 7.40 (m, 5H, H-Ph) 7.15 (s, 1H, H-8), 6.01 (s, 1H, H- found: 536.2402; λ =379nm,λ =446nm,ε(λ )=18,600M 3 −1 abs em max 10), 5.06 (s, 2H, CH -Ph), 4.40 (m, 1H, H-α), 3.72 (s, 2H, CH - cm , Φ = 0.749. 2 2 6 2 Coum), 3.27 (m, 4H, CH N), 2.84−2.71 (m, 6H, 2 × CH -Ar, H-ε), N -((Benzyloxy)carbonyl-N -(2-oxo-1,2-dihydroquinoline-4-car- 2 2 bonyl)-L-lysine (39). From 4-carboxy-quinolin-2-one (0.189 g, 1.0 1.98−1.91 (m, 5H, 2 × CH -CH -CH ,H-β), 1.75 (m, 1H, H-β), 2 132 2 mmol, 1 equiv), HCl·Lys(Cbz)-OMe (0.364 g, 1.1 mmol, 1.1 equiv), 1.62−1.41 (m, 4H, H-γ,H-δ); C NMR (126 MHz, CD OD): δ 3 HBTU(0.420g,1.1mmol,1.1equiv), and DIEA (692 μL, 4 mmol, 4 175.6 (CO H), 171.6 (CONH), 165.3, 161.4 (CO-Coum), 159.1 2 equiv), compound 39 (0.168 g) was obtained after puriﬁcation by (NHCO2), 152.7, 152.3, 147.7 (Cq-Ar), 138.3 (Cq-Ar), 129.6 (2 × column chromatography on silica gel (CH Cl /MeOH/AcOH: 85/ CH-Ar), 129.1(CH-Ar), 128.8 (3 × CH-Ar), 123.3 (CH-Ar), 118.7 2 2 (Cq-Ar), 109.2 (CH-Ar), 108.8 (Cq-Ar), 67.5 (OCH -Ph), 51.0 (C- 15/1) as a white powder in a 43% yield, according to general 2 procedure B. mp: 214−215 °C; 1H NMR (500 MHz, (CD ) SO): δ α), 50.5, 50.4 (CH2-N), 41.5 (CH2-Coum), 40.1 (C-ε), 32.0 (C-β), 3 2 30.8 (C-δ), 28.8 (CH ), 24.1 (C-γ), 22.5, 21.6, 21.4 (CH ); HPLC 11.94 (brs, 1H, CO H), 8.51 (d, J = 7.5 Hz, 1H, H-Ar), 7.80 (d, J = 2 2 2 8.5 Hz, 1H, NH), 7.50 (t, J = 7.5 Hz, 1H, H-Ar), 7.35−7.21 (m, 7H, (method A-2): tR = 17.0 min; HPLC-MS (method B): tR = 15.0 min; − − H-Ar, H-Ph), 7.17 (t, J = 7.5 Hz, 1H, H-Ar), 6.53 (s, 1H, H-Ar), 4.98 MS: (ESI ), m/z (%): [M − H] = 560.5 (100%), [M − C H O − − − 7 8 − (s, 2H, CH -Ph), 4.27 (m, 1H, H-α), 2.99 (m, 2H, H-ε), 1.85, 1.64 H] = 452.4 (30%); HRMS: (ESI ) calcd for [C H N O − H] : 2 31 35 3 7 8243 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article max max 560.2402; found: 560.2424; λ = 395 nm, λ = 487 nm, ε (λ )= mL) then DIEA (785 μL, 4.5 mmol, 3 equiv) was added dropwise, −1 −1 abs em max 8870 M cm , Φ = 0.736. and the solution was stirred overnight at room temperature. After N6-((Benzyloxy)carbonyl-N2-(((7-hydroxy-2-oxo-2H-chromen-4- evaporation, the residue was puriﬁed by column chromatography on yl)methoxy)carbonyl)-L-lysine (42). 7-Hydroxy-4-hydroxy methyl- silica gel (CH Cl /MeOH/AcOH: 85/15/1). Compound 45 (0.272 coumarin 27 (0.150 g, 0.78 mmol, 1 equiv) and CDI (0.156 g, 0.93 2 2 1 g) was obtained as a white powder in a 35% yield. H NMR (500 mmol, 1.2 equiv) were added in dry DMF (2 mL), and the solution MHz, (CD ) SO): δ 12.18 (brs, 1H, CO H), 11.01 (brs, 1H, OH), was stirred for 3 h at room temperature leading to a white precipitate. 3 2 2 7.88 (d, J = 7.5 Hz, 2H, H-Ar), 7.71 (d, J = 7.0 Hz, 2H, H-Ar), 7.66 HCl·Lys(Cbz)-OMe. (0.307 g, 0.93 mmol, 1.2 equiv) and DIEA (215 (d, J = 8.5 Hz, 1H, H-5), 7.40 (brt, J = 7.5 Hz, 3H, H-Ar, NH), 7.31 μL, 1.25 mmol, 1.6 equiv) were added in solution and heated to 80 (t, J = 7.5 Hz, 2H, H-Ar), 6.78 (dd, J = 8.5 Hz, J = 2.0 Hz, 1H, H-6), °C for 1.5 h. After cooling to room temperature, the pH was adjusted 6.71 (d, J = 2.0 Hz, 1H, H-8), 6.24 (s, 1H, H-3), 4.29−4.22 (m, 3H, to 3 with aqueous HCl (1 M) and the solution was extracted with OCHCH), 4.10 (m, 1H, H-α), 3.91, 3.88 (AB, 2H, CH -Coum), EtOAc(2×30mL).TheorganiclayerwaswashedwithH O(2×20 2 2 2 2.96 (dd, J = 13.5 Hz, J = 4.0 Hz, 1H, H-β), 2.80 (dd, J = 13.5 Hz, J = mL), dried (Na SO ), ﬁltered, and evaporated. After puriﬁcation by 13 2 4 4.5 Hz, 1H, H-β); C NMR (126 MHz, (CD ) SO): δ 172.0 column chromatography on silica gel (EtOAc), the ester compound 3 2 (CO2H), 161.3 (Cq-7), 160.1 (CO-Coum), 155.8, 155.4 (Cq-8a, was diluted with a solution of aqueous LiOH (0.5 M)/THF (3 mL, NHCO), 152.4 (Cq-4), 143.8 (2 × Cq-Fmoc), 140.7 (2 × Cq- 1/1) and stirred at room temperature for 2 h. THF was removed 2 Fmoc), 127.6, 127.0 (2 × CH-Fmoc), 126.9 (CH-5), 125.2 (2 × CH- under vacuum, and the pH was adjusted to 2 with aqueous HCl Fmoc), 112.8 (CH-6), 110.3 (CH-3), 110.0 (Cq-4a), 102.3 (CH-8), solution (1 M) then extracted with EtOAc (2 × 15 mL). Organic 65.6 (OCH CH), 54.2 (C-α), 46.6 (OCH CH), 33.4 (C-β), 31.3 layers were dried (Na SO ), ﬁltered, and evaporated. The residue was 2 2 2 4 (CH2-Coum); HPLC (method A-1): tR = 20.3 min; HPLC (method puriﬁed by column chromatography on silica gel (CH Cl /MeOH/ A-2): t = 16.2 min; HPLC-MS (method B): t = 16.6 min; HPLC- 2 2 R R AcOH: 85/15/1). Compound 42 (0.080 g) was obtained as a white + + MS (method C): t = 5.8 min; MS: (ESI ), m/z (%): [M + H] = gumina21%yield.mp:72−73°C;1HNMR(500MHz,CD OD):δ R − − 3 518.2 (100%); MS: (ESI ), m/z (%): [M − H] = 516.1 (100%); 7.44 (d, J = 9.0 Hz, 1H, H-5), 7.28 (m, 5H, H-Ph), 6.76 (dd, J = 8.5 − − HRMS (ESI ) calcd for [C H NO S − H] : 516.1122, found: Hz, J = 2.5 Hz, 1H, H-6), 6.68 (d, J = 2.5 Hz, 1H, H-8), 6.24 (s, 1H, max max 28 23 7 −1 −1 516.1126; λ =328nm,λ =399nm,ε(λ )=5960M cm ,Φ abs em max H-3), 5.25 (s, 2H, CH2-Ph), 5.02 (s, 2H, CH2-Coum), 4.18 (m, 1H, = 0.0922. H-α), 3.11 (t, J = 6.5 Hz, 2H, H-ε), 1.87, 1.72 (2 m, 2H, H-β), 1.54− N-(((9H-Fluoren-9-yl)methoxy)carbonyl)-S-((7-methoxy-2-oxo- 13 2H-chromen-4-yl)methyl)-L-cysteine (46). 4-Chloromethyl-7-me- 1.42 (m, 4H, H-γ,H-δ); C NMR (126 MHz, CD OD): δ 175.3 3 (CO2H), 163.5, 163.0, 158.9 (Cq-O), 157.7 (Cq-Ar), 156.7 thoxy-coumarin 26 (0.224 g, 1.0 mmol, 1 equiv) and Fmoc-Cys- (NHCO), 153.2, 138.4 (Cq-Ar), 129.5 (2 × CH-Ar), 128.9 (CH- OH16(0.343g,1.0mmol,1equiv)weredilutedindryDMF(8mL) 2 then Et N (278 μL, 2.0 mmol, 2 equiv) was added dropwise, and the Ar), 128.7 (2 × CH-Ar), 126.4 (CH-Ar), 114.6 (CH-Ar), 111.0 (Cq- 3 Ar), 108.9 (CH-Ar), 103.9 (CH-Ar), 67.4 (OCH -Ph), 62.0 (CH - solution was stirred overnight at room temperature. After evaporation, 2 2 Coum), 52.9 (C-α), 41.6 (C-ε), 32.4 (C-β), 30.4 (C-δ), 24.1 (C-γ); the residue was puriﬁed by column chromatography on silica gel HPLC (method A-2): t = 13.9 min; HPLC-MS (method B): t = (CH Cl /MeOH/AcOH: 85/15/1). Compound 46 (0.152 g) was R R 2 2 1 + + obtained as a white powder in a 28% yield. H NMR (500 MHz, 12.9 min; MS: (ESI ), m/z (%): [M + H] = 499.1 (50%); [2M + + − − (CD ) SO): δ 7.88 (d, J = 7.5 Hz, 2H, H-Ar), 7.75 (d, J = 8.0 Hz, 1H, Na] = 1019.0 (100%); MS: (ESI ), m/z (%): [M − H] = 497.2 3 2 − − H-Ar), 7.75 (d, J = 8.0 Hz, 1H, H-Ar), 7.70 (m, 2H, H-Ar), 7.40 (t, J (100%); HRMS (ESI ) calcd for [C H N O − H] : 497.1565; max max 25 26 2 9 −1 found: 497.1567; λ =324nm,λ =395nm,ε(λ )=11,200M =7.5 Hz, 2H, H-Ar), 7.30 (t, J = 7.5 Hz, 3H, H-Ar, NH), 6.96 (s, 1H, −1 abs em max H-Ar), 6.89 (d, J = 8.0 Hz, 1H, H-Ar), 6.32 (s, 1H, H-3), 4.33−4.17 cm , Φ = 0.12. 2 6 (m, 3H, OCH CH), 4.05 (m, 1H, NHCH), 3.92 (m, 2H, CH - N -(((9H-Fluoren-9-yl)methoxy)carbonyl)-N -(2-(7-(dimethylami- 2 2 no)-2-oxo-2H-chromen-4-yl)acetyl)-L-lysine (44). From 7-dimethy- Coum), 3.82 (s, 3H, CH ), 2.98 (dd, J = 13.5 Hz, J = 3.0 Hz, 1H, H- 3 13 lamino-4-carboxymethyl-coumarin 29 (0.247 g, 1.0 mmol, 1 equiv), β), 2.81 (dd, J = 12.5 Hz, J = 8.5 Hz, 1H, H-β); CNMR(126MHz, Fmoc-Lys-OMe43(0.442g,1.0mmol,1equiv),HBTU(416mg,1.1 (CD ) SO): δ 172.1 (CO H), 162.2 (CO Coum), 159.9 (Cq- mmol, 1.1 equiv), and DIEA (865 μL, 5.0 mmol, 5 equiv), compound 3 2 2 OMe),155.6, 155.3 (Cq-O, NHCO ), 152.3 (Cq-Ar), 143.8 (2 × Cq- 2 44 (0.030 g) was obtained after puriﬁcation by column chromatog- Ar), 140.6 (2 × Cq-Ar), 127.5, 127.0 (4 × CH-Ar), 126.7 (CH-Ar), raphy on silica gel (CH Cl /MeOH/AcOH: 85/15/1) as a yellow 125.2, 120.0 (4 × CH-Ar), 111.8 (CH-Ar), 111.3 (CH-Ar), 111.2 2 2 1 powder in a 5% yield, according to general procedure B. H NMR (Cq-Ar), 100.9 (CH-Ar), 65.6 (OCH CH), 55.8, 55.7 (C-α, OCH ), (500 MHz, CD OD): δ 7.76 (d, J = 7.5 Hz, 2H, H-Ar), 7.64 (d, J = 2 3 3 46.6 (OCH CH), 33.7 (C-β), 31.3 (CH -Coum); HPLC (method A- 2 2 7.5 Hz, 2H, H-Ar), 7.51 (d, J = 9.0 Hz, 1H, H-5), 7.36 (t, J = 7.5 Hz, 2): tR = 18.0 min; HPLC-MS (method B): tR = 18.0 min; HPLC-MS 2H, H-Ar), 7.28 (t, J = 7.5 Hz, 2H, H-Ar), 6.71 (dd, J = 9.0 Hz, J = + + (method C): t = 10.8; MS: (ESI ), m/z (%): [M + H] = 532.2 −R − − 2.5 Hz, 1H, H-6), 6.48 (d, J = 2.5 Hz, 1H, H-8), 6.02 (s, 1H, H-3), (100%); (ESI ), m/z (%): [M − H] = 530.2 (60%), [2M − H] = 4.33 (d, J = 6.0 Hz, 2H, OCH CH), 4.18 (t, J = 6.0 Hz, 1H, − − 2 1061.3 (100%); HRMS (ESI ) calcd for [C H NO S − H] : OCHCH),4.09(m,1H,H-α),3.64(s,2H,CH -Coum),3.21(t,J = max max 29 25 7 2 2 530.1279, found: 530.1277; λ = 324 nm, λ = 396 nm, ε (λ )= −1 −1 abs em max 6.5 Hz, 2H, H-ε), 3.01 (m, 6H, N(CH ) ), 1.84, 1.68 (2m, 2H, H-β), 8570 M cm , Φ = 0.0428. 13 3 2 1.56−1.38 (m, 4H, H-γ,H-δ); C NMR (126 MHz, CD OD): δ N-(((9H-Fluoren-9-yl)methoxy)carbonyl)-S-((6-chloro-7-hydroxy- 3 2-oxo-2H-chromen-4-yl)methyl)-L-cysteine (47). 4-Chloromethyl-6- 177.4 (CO H), 167.5 (CONH), 160.7 (CO-Coum), 155.5, 155.3 2 chloro-7-hydroxy-coumarin 28 (0.049 g, 0.2 mmol, 1 equiv) and (Cq-O, NHCO ), 152.7 (Cq-Ar), 151.4 (Cq-N(Me)2), 143.8, 140.6 2 Fmoc-Cys-OH 16 (0.068 g, 0.2 mmol, 1 equiv) were diluted in dry (2 × Cq-Fmoc), 127.5, 127.0 (2 × CH-Fmoc), 126.0 (CH-Ar), 125.2 (2 × CH-Fmoc), 120.0 (2 × CH-Fmoc), 109.3 (CH-Ar), 108.9 (CH- DMF (2 mL) then DIEA (108 μL, 0.6 mmol, 3 equiv) was added Ar), 108.2 (Cq-Ar), 97.4 (CH-Ar), 65.2 (OCH CH), 52.00 (C-α), dropwise, and the solution was stirred overnight at room temperature. 2 After evaporation, the residue was puriﬁed by column chromatog- 46.7 (OCH CH), 40.1 (C-ε), 40.0 (2 × NCH ), 38.7 (CH -Coum), 2 3 2 raphy on silica gel (CH Cl /MeOH/AcOH: 90/10/1). Compound 32.0 (C-β), 28.9 (C-δ), 22.7 (C-γ); HPLC (method A-2): t = 17.7 2 2 R 1 min; HPLC-MS (method B): tR = 15.2 min; MS: (ESI+), m/z (%): 47(0.035 g) was obtained as a white powder in a 32% yield. H NMR + − − (500 MHz, (CD ) SO): δ 12.85 (brs, 1H, CO H), 11.38 (brs, 1H, [M+H] =598.4(100%);MS:(ESI ),m/z(%):[M−H] =596.3 3 2 2 − − OH-Ar), 7.88 (m, 2H, H-Ar), 7.75 (d, J = 8.0 Hz, 1H, NH), 7.71 (d, J (100%); HRMS (ESI ) calcd for [C H N O − H] : 596.24022, 34 35 3 7 found: 596.24084; λmax= 373 nm, λmax= 447 nm, ε (λ ) = 17,600 = 7.5 Hz, 2H, H-Ar), 7.41 (t, J = 7.5 Hz, 2H, H-Ar), 7.31 (t, J = 7.5 −1 −1 abs em max Hz, 2H, H-Ar), 6.91 (s, 1H, H-Ar), 6.31 (s, 1H, H-3), 4.33−4.15 (m, M cm ,Φ=0.599. N-(((9H-Fluoren-9-yl)methoxy)carbonyl)-S-((7-hydroxy-2-oxo- 4H, OCH2CH, H-α), 3.96 (m, 2H, CH2-Coum), 2.95 (dd, J = 13.5 2H-chromen-4-yl)methyl)-L-cysteine (45). 4-Chloromethyl-7-hy- Hz, J = 4.0 Hz, 1H, H-β), 2.76 (dd, J = 13.5 Hz, J = 10.0 Hz, 1H, H- droxy-coumarin 25 (0.316 g, 1.5 mmol, 1 equiv) and Fmoc-Cys- 13 β); CNMR(126MHz,(CD)SO):δ172.0(CO H),159.6 (CO- 3 2 2 OH16(0.515 g, 1.5 mmol, 1 equiv) were diluted in dry THF (7.5 Coum),156.2, 155.9, 155.8, 153.6, 151.4, 143.8, 143.7 (Cq-Ar), 140.7 8244 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article 3 (2 × Cq-Ar), 127.5, 127.0 (2 × CH-Ar), 126.4 (CH-Ar), 125.2, 125.1 Radiotracer Flux Assays. N-acetyl [6- H]neuraminic acid (20 (CH-Ar), 120.0 (2 × CH-Ar), 116.7 (Cq-Ar), 111.6 (CH-Ar), 110.9 Ci/mmol) and [4,5-3H] Fmoc-L-leucine (50 Ci/mmol) were (Cq-Ar), 103.4 (CH-Ar), 65.7 (OCH CH), 53.6 (C-α), 46.6 14 2 purchased from American Radiolabeled Chemicals. L-[1,2,1′,2′- C]- 3 (OCHCH), 32.7 (C-β), 31.2 (CH -Coum); HPLC (method A-2): cystine (200 mCi/mmol) and L-[2,3,4,5- H]-proline (75 Ci/mmol) 2 2 3 t = 16.8 min; HPLC-MS (method B): t = 17.6 min; HPLC-MS were from PerkinElmer. [ H]Neu5Ac uptake into whole HEK293 R + R + 6 (method C): t = 8.9 min; MS: (ESI ), m/z (%): [M + H] = 552.1 cells was measured 2 days after transfection as described with minor R − − (100%); MS: (ESI ), m/z (%): [M − H] = 550.1 (100%). changes. Cells were brieﬂy washed and incubated for 15 min at room Modeling and Docking Computation. Homology Modeling. 3 temperature with [ H]Neu5Ac (12.5 nM; 0.05 μCi/well) in a The human sequence of sialin was retrieved from the UniProt + medium buﬀered with MES-Na pH 5.0. After two brief ice-cold database59 under the code Q9NRA2. The DgoT template structures washes, the cellular radioactivity in the cells was counted by liquid 6E9N and 6E9051 were retrieved form the Protein Data Bank. scintillation with a Tri-Carb 4910TR counter (PerkinElmer). For 60 Sequence alignment was carried out using CLUSTAL W as classical experiments, inhibitors were added simultaneously with ̀ 3 implemented in Discovery Studio (Dassault Systemes BIOVIA, [ H]Neu5Ac. However, for some experiments (Figure 6), inhibitors Discovery Studio, 2019) and further reﬁned manually. The model were pre-incubated for 15 min at pH 5.0 followed by 15 or 30 min 61 and the best out of 100 models + was generated using MODELER, washes in a medium buﬀered at pH 7.0 with MOPS-Na before was selected according to the PDF total energy. 3 measuring [ H]Neu5Ac transport. For saturation kinetics, incubation Molecular Docking. Flexible docking of 45 to the inward-facing was shortened to 10 min to keep measurements within the linear 53 A set of nine residues with 3 model was carried out using GOLD. phase of uptake at all [ H]Neu5Ac concentrations. ﬂexible side chains was used to deﬁne the binding site: F50, Y54, R57, 3 [ H]-Fmoc-Leu-OH uptake was measured similarly using 2 μM F115, F116, Y119, F179, Y306, and Y335. The Goldscore was used to (1/10 of IC ) and 0.05 μCi/well of the tracer. To compare Fmoc- 50 3 keep the best 10 out of 100 poses. Then, the pose that showed the Leu-OH and Neu5Ac in Figure S1,[H]Neu5Ac transport was best orientation according to the hypothesis made from the biological measured at a similar transporter occupancy (100 μM; 0.05 μCi/ data was selected and further reﬁned through molecular dynamic well). simulation. Human cystinosin and rat LYAAT1 were assayed at the plasma Molecular Dynamics Simulations. The system with proteins and membrane of HEK293 cells in MES-Na+ pH 5.0 buﬀer one day after ligands was prepared in the CHARMM-GUI web server62 in order to 14 3 lipofection using [ C]cystine (20 μM; 0.08 μCi/well) and [ H]- generate a membrane around the protein and solvate with water and proline (100 μM; 0.1 μCi/well) as substrates, respectively, as ions. A heterogeneous membrane made of 90% POPG and 10% 46,47 previously described. cholesterol was chosen, and a TIP3 water model with NaCl (0.15 M) Immunoﬂuorescence Analysis. Sialin distribution was analyzed counter ions was chosen for the solvation. The system was typed with 6 2 days after transfection as described. Cells were ﬁxed with 4% a CHARMM36mforce ﬁeld, and NAMD 2.13 was used. The system paraformaldehyde. After quenching with 50 mM NH Cl and several 4 was equilibrated through six constrained simulations for a total of 690 washes, cells were permeabilized and blocked with 0.05% saponin and ps by gradually diminishing the force constraints at each steps. The 2+ 2+ 0.2% BSA in PBS buﬀer containing Ca and Mg . Coverslips were following constraints were applied (each value represents an then incubated for ≥1 h with mouse anti-LAMP1 antibodies (H4A3; equilibration step): protein backbone (10/5/2.5/1/0.5/0.1 kcal/ Developmental Studies Hybridoma Bank) at 0.75 μg/mL in blocking mol), protein side chains (5/2.5/1.25/0.5/0.25/0.05 kcal/mol), lipid buﬀer, washed, and incubated with Cy3-conjugated donkey anti- heads (5/5/2/1/0.2/0 kcal/mol), and dihedral bonds (500/200/ mouse antibodies (Jackson ImmunoResearch) at 1.4 μg/mL in the 100/100/50/0 kcal/mol). Then, a production dynamics of 20 ns was samebuﬀer. Coverslips were then washed and mounted on glass slides carried out in NPT conditions at 303.15 K without any constraints. with Fluoromount-G (SouternBiotech). Epiﬂuorescence micrographs Cell Culture. HeLa and HEK293 cells were grown at 37 °C under were acquired under a 100× objective lens with a Nikon Eclipse TE- 5% CO2 in glucose-rich, Glutamax-I-containing Dulbecco’s modiﬁed 2000 microscope equipped with a CCD camera (Coolsnap). The Eagle medium (DMEM) supplemented with 10% fetal bovine serum intracellular distribution of recombinant sialin was classiﬁed into three (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin. categories (see Figure 8B) by an independent observer in a blind Fibroblasts from Salla patients were obtained from the Finnish manner. In other experiments, sialin/LAMP1 colocalization was Institute for Health and Welfare (THL), Helsinki, Finland, and used quantitated by assessing the spatial correlation between pixel in this study with ethical permission no. 78/13/03/00/16 (22 March intensities of the green and red channels. Dual-color images were 2016), issued by the Ethics Committee of the Helsinki and Uusimaa imported into the Fiji version (http://ﬁji.sc) of ImageJ,63 and the Hospital District, Finland. The ﬁbroblasts tested for sialic acid storage Substract Background and Coloc 2 plugins were used to calculate rescue carried compound-heterozygote mutations in the SLC17A5 Pearson’s correlation coeﬃcients across 20 to 25 cells/condition. gene: 115c→t/1007-1008del, corresponding to p.Arg39Cys/ Statistical analysis was made using Kruskal−Wallis nonparametric p.Leu336Trpfs at the protein level. Fibroblasts were grown at 37 one-way ANOVA with post hoc Dunn’s test. To test the eﬀect of 45 °C under 5% CO in glucose-rich, Glutamax-I-containing DMEM on sialin localization, the compound was added during the 2 supplemented with 20% FBS, 100 U/mL penicillin, and 100 μg/mL transfection step. The 45-supplemented culture medium was replaced streptomycin. by a fresh one every day until 4 h before cell ﬁxation. Expression of Recombinant Sialin. HeLa cells were transfected Quantiﬁcation of Sialic Acid Levels in Cells by Mass 6 Spectrometry. Two conﬂuent 75 cm2 ﬂasks of human ﬁbroblasts either by electroporation or lipofection. For electroporation, 2 × 10 HeLacells in 50 μL of ice-cold phosphate-buﬀer saline (PBS; pH 7.4) were used for each measurement. Cells were cultured for 2 days in the were mixed with 5 μg of wild-type or R39C pEGFP-C2-sialin presence of 30 or 168 μM 45 in 0.3% DMSO or with DMSO alone plasmid6 and immediately subjected to 10 square pulses (200 V, 3 for control at 37 °C under 5% CO . The 45-supplemented culture 2 ms) delivered at 1 Hz by a GHT 1287 electropulsator (Jouan) with 4 mediumwasreplaced by a fresh one twice. After the 2 days, cells were mm-spaced electrodes. Cells were then diluted with 7 mL of culture detached by trypsinization, washed with ice-cold PBS, and medium and distributed into 14 wells from a 24-well culture plate centrifuged. The resulting cell pellets were ﬂash-frozen and kept at containing glass coverslips. For lipofection, HeLa were plated −20 °C until measurement. (100,000 cells/well) into 24-well plates containing glass coverslips Sialic acid measurements were done with some diﬀerences in two and transfected on the following day with Lipofectamine 2000 laboratories. In one protocol, pellets were submitted to osmolysis by (Invitrogen) according to the manufacturer’s protocol. addition of 100 μL of ultrapure water (1 h, 4 °C) and subsequently HEK293cells were plated (250,000 cells/well) into poly-D-lysine- sonicated (3 × 20 s with 10 s resting intervals). At this stage, the coated 24-well plates and lipofected similarly with a construct carrying protein concentration was determined for further normalization using 6 a sorting motif mutation (pEGFP-C2-sialin L22G/L23G) to express Micro BCA Protein Assay Kit (ThermoFisher). After centrifugation human sialin at the plasma membrane. (3000 rpm, 10 min, 4 °C), three volumes of ice-cold EtOH were 8245 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article added to the supernatant and the mixture was kept overnight at −20 ■ AUTHORINFORMATION °C and centrifuged at 10,000 rpm for 10 min to precipitate glycoproteins. The new supernatant (containing free saccharides) Corresponding Authors was dried under a nitrogen ﬂow, while the new pellet was pooled with the previous one to extract protein- and lipid-bound saccharides. All Isabelle McCort-Tranchepain − Laboratoire de Chimie et samples were subsequently treated with 100 μL of triﬂuoroacetic acid Biochimie Pharmacologiques et Toxicologiques, CNRS, UMR ((TFA) 0.1 M, 2 h, 80 °C) for selective release of bound sialic acids 8601, Université de Paris, F-75006 Paris, France; orcid.org/ and dried overnight into a vacuum concentrator (Concentrator 5301, 0000-0001-7447-8806; Email: isabelle.mccort@ Eppendorf). Dried samples were incubated with 50 μL of the derivatization solution containing 1,2-diamino-4,5-methylenedioxy- parisdescartes.fr benzene dihydrochloride (DMB, 7 mM), 2-mercaptoethanol (1 M), ̀ Bruno Gasnier − SPPIN - Saints-Peres Paris Institute for the Na S O (18 mM), and TFA (20 mM) for 2 h at 50 °C in the dark. Neurosciences, CNRS, Université de Paris, F-75006 Paris, 2 2 4 Samples were kept at −20 °C before analysis. Quantitative analyses France; orcid.org/0000-0003-4458-6296; 3 were performed by micro-LC/ESI MRM-MS in the positive ion Email: [email protected] mode on an amaZon speed ETD ion trap mass spectrometer (Bruker ̀ Daltonics) equipped with a standard ESI source and controlled by Christine Anne − SPPIN - Saints-Peres Paris Institute for the 2 fragment ions Neurosciences, CNRS, Université de Paris, F-75006 Paris, Hystar software (ver. 3.2). The identiﬁcation of MS France; Email: [email protected] was based on previous papers.64,65 DMB-coupled sialic acid separation was achieved on Prominence LC-20AB micro LC system (Shimadzu). Samples were diluted 10-fold in formic acid (0.1%), and dilutions Authors were applied (5 μL) to a Luna 3 μm analytical column (C18, 100 Å, Lilian Dubois − Laboratoire de Chimie et Biochimie 150 × 1 mm, Phenomenex) with isocratic elution in acetonitrile/ methanol/water (4:6:90, v/v/v) at 60 μL/min. Multiple reaction Pharmacologiques et Toxicologiques, CNRS, UMR 8601, 3 was used for DMB-coupled sialic acid Université de Paris, F-75006 Paris, France monitoring (MRM) of MS quantiﬁcation (ion spray voltage of 4500 V; dry gas slow rate of 8 L/ ́ Nicolas Pietrancosta − Laboratoire des Biomolecules, LBM, min). Absolute quantiﬁcations were calculated by comparing ion ́ ́ ́ Sorbonne Universite,Ecole Normale Superieure, PSL University, intensities to a standard curve established for DMB-coupled sialic CNRS, F-75005 Paris, France; Neurosciences Paris Seine - acids (Neu5Ac, Neu5Gc, KDN). Results were normalized for the total Institut de Biologie Paris Seine (NPS - IBPS), Sorbonne protein amount (nmol of sialic acids/mg of protein). ́ In another protocol, 250 μLofa85μmol/L solution of Neu5Ac Universite, INSERM, CNRS, F-75005 Paris, France 13 Alexandre Cabaye − Laboratoire de Chimie et Biochimie 1,2,3- C3 (internal standard, IS; Sigma-Aldrich, ref no. 649694) was Pharmacologiques et Toxicologiques, CNRS, UMR 8601, added to each ﬁbroblast pellet. Samples were sonicated (3 × 10 s with Université de Paris, F-75006 Paris, France; BIOVIA, Dassault 5sresting intervals in ice) and assayed for protein concentrations. For ̀ free sialic acid, 100 μL of this homogenate was mixed with 150 μLof Systemes, F-78140 Velizy-Villacoublay, France ̀ acetonitrile (ACN), homogenized, and centrifuged. For the total sialic Isabelle Fanget − SPPIN - Saints-Peres Paris Institute for the acid, 100 μL of the homogenate was mixed with 200 μL of sulfuric Neurosciences, CNRS, Université de Paris, F-75006 Paris, acid (63 mM) and incubated for 1 h at 80 °C for hydrolysis. Samples France were supplemented with 450 μL of ACN, homogenized, and ́ ̀ Cecile Debacker − SPPIN - Saints-Peres Paris Institute for the centrifuged. The two supernatants (free and bound sialic acid) were Neurosciences, CNRS, Université de Paris, F-75006 Paris, 2 66 quantitated by LC/MS as described. Transitions 308.1 > 86.9 (for France Neu5Ac) and 311.1 > 89.9 (for IS) were monitored with a Pierre-André Gilormini − UMR 8576, UGSF, Unité de declustering potential of −50 V and a collision energy of −16 V. Glycobiologie et Fonctionnelle, Université de Lille, CNRS, F- Peak integration was performed with Analyst software (version 1.6.2, 59650 Lille, France Applied Biosystems; smoothing width: 3 points). Neu5Ac concen- Patrick M. Dansette − Laboratoire de Chimie et Biochimie trations were calculated from the Neu5Ac area/IS area ratio and the calibration curve (linear through zero) and normalized to protein Pharmacologiques et Toxicologiques, CNRS, UMR 8601, concentrations. Université de Paris, F-75006 Paris, France; orcid.org/0000- 0002-3694-3348 ■ ASSOCIATED CONTENT Julien Dairou − Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques, CNRS, UMR 8601, sı *Supporting Information Université de Paris, F-75006 Paris, France The Supporting Information is available free of charge at Christophe Biot − UMR 8576, UGSF, Unité de Glycobiologie https://pubs.acs.org/doi/10.1021/acs.jmedchem.9b02119. et Fonctionnelle, Université de Lille, CNRS, F-59650 Lille, Compound structure, abbreviations, and code numbers; France; orcid.org/0000-0002-7396-1959 synthesis scheme of Fmoc-amino acid 11, 12, 16, 22, 24, Roseline Froissart − Service de Biochimie et Biologie ́ and 43; HPLC conditions and retention time for Moleculaire Grand Est, Centre de Biologie et de Pathologie Est, commercially available compounds; HPLC and HPLC- Hospices Civils de Lyon, F-69677 Bron, France ̀ MS spectra for compounds 3, 13, 45, 46, and 47; Anne Goupil-Lamy − BIOVIA, Dassault Systemes, F-78140 experimental details and analytical data for compounds Velizy-Villacoublay, France ̀ 11, 12, 16, 22, 24−28, and 43; NMR spectra for Hugues-Olivier Bertrand − BIOVIA, Dassault Systemes, F- compounds 45−47; Figures S1 to S5; Supplementary 78140 Velizy-Villacoublay, France Methods; Supplementary References (PDF) Francine C. Acher − Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques, CNRS, UMR 8601, Molecular formula strings (SMILES) and associated Université de Paris, F-75006 Paris, France; orcid.org/0000- data (CSV). 0002-5413-4181 Sialin homology model docked with 45 (PDB) Complete contact information is available at: https://pubs.acs.org/10.1021/acs.jmedchem.9b02119 8246 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article Author Contributions Functions as a Nitrate Transporter in the Plasma Membrane. Proc. F.A., I.M.-T., B.G., and C.A. designed the research. L.D., N.P., Natl. Acad. Sci. U. S. A. 2012, 109, 13434−13439. A.C., I.F., C.D., P.-A.G., R.F., I.M.-T., and C.A. conducted the (9) Miyaji, T.; Echigo, N.; Hiasa, M.; Senoh, S.; Omote, H.; experiments. All authors analyzed the data. F.A., I.M.-T., B.G., Moriyama, Y. Identification of a Vesicular Aspartate Transporter. Proc. and C.A. wrote the paper. Natl. Acad. Sci. U. S. A. 2008, 105, 11720−11724. Funding (10) Lodder-Gadaczek, J.; Gieselmann, V.; Eckhardt, M. Vesicular Uptake of N-Acetylaspartylglutamate Is Catalysed by Sialin This work was supported by the Agence Nationale de la −38. (SLC17A5). Biochem. J. 2013, 454,31 Recherche (grant no. ANR-15-CE14-0001-02 to B.G.), the (11) Bardor, M.; Nguyen, D. H.; Diaz, S.; Varki, A. Mechanism of Association Nationale de la Recherche et de la Technologie Uptake and Incorporation of the Non-Human Sialic Acid N- ̀ (ANRT) and Dassault Systemes BIOVIA (CIFRE PhD Glycolylneuraminic Acid into Human Cells. J. Biol. Chem. 2005, ̀ 280, 4228−4237. scholarship no. 2018/0027 to A.C.), the Ministere de ́ (12) Chou, H.-H.; Takematsu, H.; Diaz, S.; Iber, J.; Nickerson, E.; l’Enseignement Superieur, de la Recherche et de l’Innovation ̂ ́ Wright, K. L.; Muchmore, E. A.; Nelson, D. L.; Warren, S. T.; Varki, (PhD fellowship to P.-A.G.), the Neuropole Region Ile-de- France (PhD fellowship to L.D.), and the Science Ambassador A. A Mutation in Human CMP-Sialic Acid Hydroxylase Occurred ̀ after the Homo-Pan Divergence. Proc. Natl. Acad. Sci. U. S. A. 1998, Program from Dassault Systemes BIOVIA (to A.C. and F.A.). 95, 11751−11756. Notes (13) Dhar, C.; Sasmal, A.; Varki, A. From “Serum Sickness” to The authors declare no competing ﬁnancial interest. “Xenosialitis”: Past, Present, and Future Significance of the Non- ACKNOWLEDGMENTS Human Sialic Acid Neu5Gc. Front. Immunol. 2019, 10, 807. ■ (14) Kawanishi, K.; Dhar, C.; Do, R.; Varki, N.; Gordts, P. L. S. M.; ̈ ̈ Varki, A. Human Species-Specific Loss of CMP-N-Acetylneuraminic WethankDr.AijaKyttala from the Finnish Institute for Health Acid Hydroxylase Enhances Atherosclerosis via Intrinsic and Extrinsic and Welfare (THL) for acquiring the ethical permissions in Mechanisms. Proc. Natl. Acad. Sci. U. S. A. 2019, 116, 16036−16045. Finland and providing Salla ﬁbroblasts for this study. C.B. ̈ thanks the glycoconjugate analysis facility PAGes (http:// (15) Bull, C.; Heise, T.; Adema, G. J.; Boltje, T. J. Sialic Acid plateforme-pages.univ-lille1.fr) and the FRABio research Mimetics to Target the Sialic Acid-Siglec Axis. Trends Biochem. Sci. federation. 2016, 41, 519−531. ̈ (16) Moons, S. J.; Adema, G. J.; Derks, M. T.; Boltje, T. J.; Bull, C. ABBREVIATIONS Sialic Acid Glycoengineering Using N-Acetylmannosamine and Sialic ■ Acid Analogs. Glycobiology 2019, 29, 433−445. Bz, benzoyl; Bzl, benzyl; Cbz, benzyloxycarbonyl; t-Bu, tert- (17) Aula, P.; Gahl, W. A. Disorders of Free Sialic Acid Storage. In butyl; Boc, tert-butoxycarbonyl; CDI, 1,1′-carbonyldimidazole; The Metabolic Molecular Basis of Inherited Disease; Eighth Edition; DIEA, diethylamine; DMF, dimethylformamide; Fmoc, ﬂuo- Scriver, C. R., Beaudet, A. L., Sly, W. S., Valle, D., Childs, B., Kinzler, ren-9-ylmethoxycarbonyl; HBTU, 2-(1H-benzotriazol-1-yl)- K. W., Vogelstein, B., Eds.; McGraw-Hill: New-York, 2001, pp 5109− 1,1,3,3-tetramethyluronium hexaﬂuorophosphate; MOMCl, 5120. ̈ methoxymethyl chloride; RT, room temperature; TFA, (18) Aula, N.; Salomaki, P.; Timonen, R.; Verheijen, F.; Mancini, G.; ̊ triﬂuoroacetic acid; THF, tetrahydrofuran; TMSCl, trimethyl- Mansson, J.-E.; Aula, P.; Peltonen, L. The Spectrum of SLC17A5- silyl chloride; Trt, trityl Gene Mutations Resulting in Free Sialic Acid−Storage Diseases Indicates Some Genotype-Phenotype Correlation. Am. J. Hum. Genet. REFERENCES 2000, 67, 832−840. ■ (19) Renlund, M.; Tietze, F.; Gahl, W. A. Defective Sialic Acid (1) Reimer, R. J. SLC17: A Functionally Diverse Family of Organic Egress from Isolated Fibroblast Lysosomes of Patients with Salla Anion Transporters. Mol. Aspects Med. 2013, 34, 350−359. Disease. Science 1986, 232, 759−762. (2) Iharada, M.; Miyaji, T.; Fujimoto, T.; Hiasa, M.; Anzai, N.; (20) Varho, T. T.; Alajoki, L. E.; Posti, K. M.; Korhonen, T. T.; Omote, H.; Moriyama, Y. Type 1 Sodium-Dependent Phosphate ̈̈ Renlund, M. G.; Nyman, S. R. G.; Sillanpaa, M. L.; Aula, P. P. Transporter (SLC17A1 Protein) Is a Cl–Dependent Urate Exporter. Phenotypic Spectrum of Salla Disease, a Free Sialic Acid Storage J. Biol. Chem. 2010, 285, 26107−26113. Disorder. Pediatr. Neurol. 2002, 26, 267−273. (3) Blakely, R. D.; Edwards, R. H. Vesicular and Plasma Membrane ̈ (21) Zielonka, M.; Garbade, S. F.; Kolker, S.; Hoffmann, G. F.; Ries, Transporters for Neurotransmitters. Cold Spring Harbor Perspect. Biol. M. A Cross-Sectional Quantitative Analysis of the Natural History of 2012, 4, a005595. Free Sialic Acid Storage Diseasean Ultra-Orphan Multisystemic (4) Sawada, K.; Echigo, N.; Juge, N.; Miyaji, T.; Otsuka, M.; Omote, Lysosomal Storage Disorder. Genet. Med. 2019, 21, 347−352. H.; Yamamoto, A.; Moriyama, Y. Identification of a Vesicular ̈ ̈ ̈ Nucleotide Transporter. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, (22) Sonninen, P.; Autti, T.; Varho, T.; Hamalainen, M.; Raininko, 5683−5686. R. Brain Involvement in Salla Disease. Am. J. Neuroradiol. 1999, 20, (5) Verheijen, F. W.; Verbeek, E.; Aula, N.; Beerens, C. E. M. T.; 433−443. Havelaar, A. C.; Joosse, M.; Peltonen, L.; Aula, P.; Galjaard, H.; Van (23) Prolo, L. M.; Vogel, H.; Reimer, R. J. The Lysosomal Sialic Der Spek, P. J.; Mancini, G. M. S. A New Gene , Encoding an Anion Acid Transporter Sialin Is Required for Normal CNS Myelination. J. Transporter , Is Mutated in Sialic Acid Storage Diseases. Nat. Genet. Neurosci. 2009, 29, 15355−15365. 1999, 23, 462−465. (24) Stroobants, S.; Van Acker, N. G. G.; Verheijen, F. W.; Goris, I.; ̀ Daneels, G. F. T.; Schot, R.; Verbeek, E.; Knaapen, M. W. M.; De (6) Morin, P.; Sagne, C.; Gasnier, B. Functional Characterization of ̈ Wild-Type and Mutant Human Sialin. EMBO J. 2004, 23, 4560− Bondt, A.; Gohlmann, H. W.; Crauwels, M. L. A.; Mancini, G. M. S.; 4570. Andries, L. J.; Moechars, D. W. E.; D’Hooge, R. Progressive (7) Wreden, C. C.; Wlizla, M.; Reimer, R. J. Varied Mechanisms Leukoencephalopathy Impairs Neurobehavioral Development in Underlie the Free Sialic Acid Storage Disorders. J. Biol. Chem. 2005, Sialin-Deficient Mice. Exp. Neurol. 2017, 291, 106−119. 280, 1408−1416. (25) Morland, C.; Nordengen, K.; Larsson, M.; Prolo, L. M.; (8) Qin, L.; Liu, X.; Sun, Q.; Fan, Z.; Xia, D.; Ding, G.; Ong, H. L.; Farzampour, Z.; Reimer, R. J.; Gundersen, V. Vesicular Uptake and Adams, D.; Gahl, W. A.; Zheng, C.; Qi, S.; Jin, L.; Zhang, C.; Gu, L.; Exocytosis of L-Aspartate Is Independent of Sialin. FASEB J. 2012, 27, He, J.; Deng, D.; Ambudkar, I. S.; Wang, S. Sialin (SLC17A5) 1264−1274. 8247 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article (26) Pitto, M.; Chigorno, V.; Renlund, M.; Tettamanti, G. Oxide Substituted γ-Amino Acids and Hybrid Peptides. Org. Biomol. Impairment of Ganglioside Metabolism in Cultured Fibroblasts Chem. 2014, 12, 8462−8472. from Salla Patients. Clin. Chim. Acta 1996, 247, 143−157. (45) Vasko, R. C.; Rodriguez, R. A.; Cunningham, C. N.; Ardi, V. C.; (27) Aula, N.; Jalanko, A.; Aula, P.; Peltonen, L. Unraveling the Agard, D. A.; McAlpine, S. R. Mechanistic Studies of Sansalvamide A- Molecular Pathogenesis of Free Sialic Acid Storage Disorders: Altered Amide: An Allosteric Modulator of Hsp90. ACS Med. Chem. Lett. Targeting of Mutant Sialin. Mol. Genet. Metab. 2002, 77,99−107. 2010, 1,4−8. ́ (28) Biancheri, R.; Rossi, A.; Verbeek, H. A.; Schot, R.; Corsolini, F.; (46) Sagne, C.; Agulhon, C.; Ravassard, P.; Darmon, M.; Hamon, Assereto, S.; Mancini, G. M. S.; Verheijen, F. W.; Minetti, C.; M.; El Mestikawy, S.; Gasnier, B.; Giros, B. Identification and Filocamo, M. Homozygosity for the p.K136E Mutation in the Characterization of a Lysosomal Transporter for Small Neutral Amino SLC17A5 Gene as Cause of an Italian Severe Salla Disease. Acids. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 7206−7211. Neurogenetics 2005, 6, 195−199. (47) Kalatzis, V.; Cherqui, S.; Antignac, C.; Gasnier, B. Cystinosin, + (29) Thompson, C. M.; Davis, E.; Carrigan, C. N.; Cox, H. D.; the Protein Defective in Cystinosis, Is a H -Driven Lysosomal Cystine Bridges, R. J.; Gerdes, J. M. Inhibitor of the Glutamate Vesicular Transporter. EMBO J. 2001, 20, 5940−5949. Transporter (VGLUT). Curr. Med. Chem. 2005, 12, 2041−2056. (48) Bizzarri, R.; Arcangeli, C.; Arosio, D.; Ricci, F.; Faraci, P.; (30) Pietrancosta, N.; Kessler, A.; Favre-Besse, F.-C.; Triballeau, N.; Cardarelli, F.; Beltram, F. Development of a Novel GFP-Based Quentin, T.; Giros, B.; El Mestikawy, S.; Acher, F. C. Rose Bengal Ratiometric Excitation and Emission PH Indicator for Intracellular Analogs and Vesicular Glutamate Transporters (VGLUTs). Bioorg. Studies. Biophys. J. 2006, 90, 3300−3314. Med. Chem. 2010, 18, 6922−6933. (49) Yan, N. Structural Biology of the Major Facilitator Superfamily (31) Tamura, Y.; Ogita, K.; Ueda, T. A New VGLUT-Specific Transporters. Annu. Rev. Biophys. 2015, 44, 257−283. Potent Inhibitor: Pharmacophore of Brilliant Yellow. Neurochem. Res. (50) Drew, D.; Boudker, O. Shared Molecular Mechanisms of 2014, 39, 117−128. Membrane Transporters. Annu. Rev. Biochem. 2016, 85, 543−572. (32) Moriyama, Y.; Nomura, M. Clodronate: A Vesicular ATP (51) Leano, J. B.; Batarni, S.; Eriksen, J.; Juge, N.; Pak, J. E.; Kimura- Release Blocker. Trends Pharmacol. Sci. 2018, 39,13−23. Someya, T.; Robles-Colmenares, Y.; Moriyama, Y.; Stroud, R. M.; ́ Edwards, R. H. Structures Suggest a Mechanism for Energy Coupling (33) Pietrancosta, N.; Anne, C.; Prescher, H.; Ruivo, R.; Sagne, C.; by a Family of Organic Anion Transporters. PLoS Biol. 2019, 17, Debacker, C.; Bertrand, H.-O.; Brossmer, R.; Acher, F.; Gasnier, B. No. e3000260. Successful Prediction of Substrate-Binding Pocket in SLC17 Trans- (52) Mancini, G. M.; Beerens, C. E.; Aula, P. P.; Verheijen, F. W. porter Sialin. J. Biol. Chem. 2012, 287, 11489−11497. Sialic Acid Storage Diseases: A Multiple Lysosomal Transport Defect (34) Huang, Y.; Lemieux, M. J.; Song, J.; Auer, M.; Wang, D.-N. for Acidic Monosaccharides. J. Clin. Invest. 1991, 87, 1329−1335. Structure and Mechanism of the Glycerol-3-Phosphate Transporter (53) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. from Escherichia Coli. Science 2003, 301, 616−620. Development and Validation of a Genetic Algorithm for Flexible (35) Irwin, J. J.; Shoichet, B. K. ZINC - A Free Database of Docking. J. Mol. Biol. 1997, 267, 727−748. Commercially Available Compounds for Virtual Screening. J. Chem. (54) Kleta, R.; Aughton, D. J.; Rivkin, M. J.; Huizing, M.; Strovel, E.; Inf. Model. 2005, 45, 177−182. Anikster, Y.; Orvisky, E.; Natowicz, M.; Krasnewich, D.; Gahl, W. A. (36) Jawalekar, A. M.; Meeuwenoord, N.; Cremers, J. S. G. O.; Biochemical and Molecular Analyses of Infantile Free Sialic Acid Overkleeft, H. S.; Van Der Marel, G. A.; Rutjes, F. P. J. T.; Van Delft, Storage Disease in North American Children. Am. J. Med. Genet. F. L. Conjugation of Nucleosides and Oligonucleotides by [3+2] 2003, 120A,28−33. Cycloaddition. J. Org. Chem. 2008, 73, 287−290. (55) Robichaud, T.; Appleyard, A. N.; Herbert, R. B.; Henderson, P. (37) Lin, W.; Long, L.; Feng, J.; Wang, B.; Guo, C. Synthesis of J. F.; Carruthers, A. Determinants of Ligand Binding Affinity and Meso-Coumarin-Conjugated Porphyrins and Investigation of Their Cooperativity at the GLUT1 Endofacial Site. Biochemistry 2011, 50, Luminescence Properties. Eur. J. Org. Chem. 2007, 2007, 4301−4304. 3137−3148. (38) Shibu, E. S.; Ono, K.; Sugino, S.; Nishioka, A.; Yasuda, A.; (56) Blat, Y. Non-Competitive Inhibition by Active Site Binders. Shigeri, Y.; Wakida, S. I.; Sawada, M.; Biju, V. Photouncaging Chem. Biol. Drug Des. 2010, 75, 535−540. Nanoparticles for MRI and Fluorescence Imaging in Vitro and in ̈̈ Vivo. ACS Nano 2013, 7, 9851−9859. (57) Morello, J.-P.; Bouvier, M.; Petaja-Repo, U. E.; Bichet, D. G. (39) Wirtz, L.; Kazmaier, U. A Mild Titanium-Catalyzed Synthesis Pharmacological Chaperones: A New Twist on Receptor Folding. of Functionalized Amino Coumarins as Fluorescence Labels. Eur. J. Trends Pharmacol. Sci. 2000, 21, 466−469. Org. Chem. 2011, 2011, 7062−7065. (58) Parenti, G.; Andria, G.; Valenzano, K. J. Pharmacological (40) Kim, J. H.; Sumranjit, J.; Kang, H. J.; Chung, S. J. Discovery of Chaperone Therapy: Preclinical Development, Clinical Translation, Coumarin Derivatives as Fluorescence Acceptors for Intrinsic and Prospects for the Treatment of Lysosomal Storage Disorders. Fluorescence Resonance Energy Transfer of Proteins. Mol. BioSyst. Mol. Ther. 2015, 23, 1138−1148. 2014, 10,30−33. (59) The Uniprot Consortium. UniProt: A Worldwide Hub of ̌ ̈ Protein Knowledge. Nucleic Acids Res. 2019, 47, D506−D515. (41) Sebej, P.; Wintner, J.; Muller, P.; Slanina, T.; Al Anshori, J.; (60) Thompson, J. D.; Higgins, D. G.; Gibson, T. J. CLUSTAL W: ́ Antony, L. A. P.; Klan, P.; Wirz, J. Fluorescein Analogues as Improving the Sensitivity of Progressive Multiple Sequence Alignment Photoremovable Protecting Groups Absorbing at ∼520 nm. J. Org. through Sequence Weighting, Position-Specific Gap Penalties and Chem. 2013, 78, 1833−1843. Weight Matrix Choice. Nucleic Acids Res. 1994, 22, 4673−4680. (42) Montalbetti, C. A. G. N.; Falque, V. Amide Bond Formation ̌ and Peptide Coupling. Tetrahedron 2005, 61, 10827−10852. (61) Sali, A.; Blundell, T. L. Comparative Protein Modelling by Satisfaction of Spatial Restraints. J. Mol. Biol. 1993, 234, 779−815. (43) Rudolf, K.; Eberlein, W.; Engel, W.; Pieper, H.; Entzeroth, M.; (62) Jo, S.; Kim, T.; Iyer, V. G.; Im, W. CHARMM-GUI: A Web- Hallermayer, G.; Doods, H. Development of Human Calcitonin Based Graphical User Interface for CHARMM. J. Comput. Chem. Gene-Related Peptide (CGRP) Receptor Antagonists. 1. Potent and 2008, 29, 1859−1865. 2-[3,5-Dibromo-N- Selective Small Molecule CGRP Antagonists. 1-[N (63) Schindelin, J.; Arganda-Carreras, I.; Frise, E.; Kaynig, V.; [[4-(3,4-Dihydro-2(1H)-Oxoquinazolin-3-Yl)-1-Piperidinyl]- Longair, M.; Pietzsch, T.; Preibisch, S.; Rueden, C.; Saalfeld, S.; Carbonyl]-D-Tyrosyl]-l-Lysyl]-4-(4-Pyridinyl)piperazine: The First Schmid, B.; Tinevez, J.-Y.; White, D. J.; Hartenstein, V.; Eliceiri, K.; CGRP Antagonist for Clinical Trials in Acute Migraine. J. Med. Tomancak, P.; Cardona, A. Fiji: An Open-Source Platform for Chem. 2005, 48, 5921−5931. Biological-Image Analysis. Nat. Methods 2012, 9, 676−682. (44) Mali, S. M.; Ganesh Kumar, M.; Katariya, M. M.; Gopi, H. N. (64) Baumann, A.-M. T.; Bakkers, M. J. G.; Buettner, F. F. R.; HBTU Mediated 1-Hydroxybenzotriazole (HOBt) Conjugate Addi- Hartmann, M.; Grove, M.; Langereis, M. A.; De Groot, R. J.; tion: Synthesis and Stereochemical Analysis of β-Benzotriazole N- ̈ Muhlenhoff, M. 9-O-Acetylation of Sialic Acids Is Catalysed by 8248 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249

Journal of Medicinal Chemistry pubs.acs.org/jmc Article CASD1 via a Covalent Acetyl-Enzyme Intermediate. Nat. Commun. 2015, 6, 7673. (65) Noel, M.; Gilormini, P.-A.; Cogez, V.; Yamakawa, N.; Vicogne, ́ D.; Lion, C.; Biot, C.; Guerardel, Y.; Harduin-Lepers, A. Probing the CMP-Sialic Acid Donor Specificity of Two Human β-D-Galactoside Sialyltransferases (ST3Gal I and ST6Gal I) Selectively Acting on O- and N-Glycosylproteins. ChemBioChem 2017, 18, 1251−1259. (66) Piraud, M.; Pettazzoni, M.; Menegaut, L.; Caillaud, C.; Nadjar, Y.; Vianey-Saban, C.; Froissart, R. Development of a New Tandem Mass Spectrometry Method for Urine and Amniotic Fluid Screening of Oligosaccharidoses. Rapid Commun. Mass Spectrom. 2017, 31, 951−963. 8249 https://dx.doi.org/10.1021/acs.jmedchem.9b02119 J. Med. Chem. 2020, 63, 8231−8249