Structure of the human taurine transporter TauT reveals substrate recognition and mechanisms of inhibition

24 Taurine, a sulfur-containing amino acid, plays a crucial role in human health due to its 25 antioxidant, anti-inflammatory, osmoregulatory properties 1,2. Taurine levels are 26 primarily regulated by the sodium-/chloride-dependent taurine transporter (TauT) 3. 27 Inhibitors of TauT have potential in the treatment of certain diseases, including 28 neurological disorders and cancer 4,5. Here, we present five structures of TauT: in the 29 substrate-free apo state, in complex with taurine and in complex with taurine-mimetic 30 inhibitors β -alanine, GABA and guanidinoethyl sulfonate (GES), each with varying 31 length of linear structure, in both inward-facing and occluded conformations. The 32 taurine-, β -alanine- and GABA-bound hTauT structures, in the presence of NaCl, 33 adopt an occluded conformation, with ligands binding in the central pocket. In the 34 presence of KCl, GES-bound hTauT adopts an inward-facing conformation, with two 35 molecules positioned along the substrate translocation pathway with one into the deep 36 central cavity and the other precluded conformational change from inward-facing to 37 occluded state. Combined with function analysis, our structures provided insights into 38 the overall architecture, substrate coordination and inhibitor recognition mechanisms 39 of TauT. 40

41 Taurine is a sulfur-containing non-protein amino acid that is found in high 42 concentration in various excitatory and oxidative tissues, particularly in the brain, 43 heart, skeletal muscles and retinal1. Taurine primarily resides in the intracellular fluid 44 of these tissues, where its concentration can reach 10–15 mM, compared to 45 approximately 50 μ M in plasma 6. Taurine participates in numerous physiological 46 activities, including regulating cell volume, maintaining cell integrity 7,8, forming bile 47 salts9, reducing the risk of cardiovascular diseases 10, promoting retinal 48 differentiation11 and reducing neuronal apoptosis and inflammation 2, through its 49 membrane stabilizing, anti-oxidative and anti-inflammatory properties. Dysregulated 50 taurine levels are associated with cardiomyopathy 12, retinopathy 13, neurological 51 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint abnormalities14, weakened immune response15 and cancer16,17. 52 In most species, the biosynthesis of taurine occurs primarily in the liver and its 53 biosynthetic capacity declines with aging 3,18. Thus, the high intracellular taurine 54 levels are maintained by taurine intake through the action of the major taurine 55 transporter, TauT (encoded by the SLC6A6 gene)3,19. The expression level of TauT is 56 high in the placenta and skeletal muscle, moderate in the heart, brain, lung and 57 pancreas, and low in the liver 5,20. Loss-of-function mutations in TauT have been 58 associated with dilated cardiomyopathy and retinal degeneration 12,21,22. On the other 59 hand, overexpression of TauT has been observed in several types of cancers, such as 60 gastric and colorectal cancer 23,24. High expression levels of TauT correlate with poor 61 prognosis and advanced tumor stages, suggesting that TauT may be a potential target 62 for anticancer therapies 5,25. 63 TauT is a chloride- and sodium-dependent transporter that belongs to the solute carrier 64 6 (SLC6) family. This family can be divided into four subgroups : GABA 65 (γ -aminobutyric acid), monoamine, neurotransmitter amino acid, and nutrient amino 66 acid transporters. TauT belongs to the first subgroup based on the similarity of its 67 substrates (Extended Data Fig. 1a) 5. Progress in structural studies of SLC6 68 transporters has revealed competitive and allosteric mechanism of their inhibitors 26-32. 69 Inhibition of GAT1 (SLC6A1) dependent GABA clearance by tiagabine in the 70 synaptic cleft is an established strategy for treating epilepsy and structural studies 71 elucidate the mode of action of these inhibitors and provides blueprints for the design 72 of neuromodulators 33,34. RGX-202, an inhibitor of the creatine transporter (SLC6A8), 73 is under clinical trials for colon cancer treatment 5,35. However, there is limited 74 information about TauT inhibitors. Although several substrate-mimetics inhibitors of 75 TauT, with simple or more complex structure, have been reported, high resolution 76 structure of TauT may largely contributed to the development of highly selective and 77 potent TauT inhibitors with anti-neoplastic property 36,37. 78 Here we report five structures of TauT in the apo state, in complex with its substrate 79 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint taurine and in complex with its inhibitors with different length of linear structure in 80 the inward-facing and occluded conformations. Combine with function analysis, these 81 structures reveal the overall architecture, substrate binding and inhibitor recognition 82 mechanism of TauT. 83

84 Overall architecture of human TauT 85 Xenopus laevis oocytes injected with human TauT cRNA exhibited time-dependent 86 accumulation of [ 3H]-taurine (Extended Data Fig. 1b) with an apparent IC 50 of 2.63 87 ±1.04 µM (Extended Data Fig. 1c). Full-length wild-type hTauT, fused with a 88 C-terminal affinity tag, was expressed in HEK293F cells 38. Recombinant hTauT 89 protein was extracted from membranes using lauryl maltose neopentyl glycol (LMNG) 90 and cholesteryl hemisuccinate (CHS). The affinity tag was removed, and hTauT was 91 purified using size exclusion chromatograph (SEC) (Extended Data Fig. 1d). Protein 92 from the peak fractions was used to prepare cryo-EM samples, leading to the 93 determination of hTauT structures in the apo state (hTauT APO), in complex with 94 substrate taurine (hTauT TAU), and in complex with inhibitors GABA (hTauT GABA), 95 β -alanine (hTauTβ A) and guanidinoethyl sulfonate (GES) (hTauTGES). The resolutions 96 of these structures range from 2.9 /i4 Å and to 3.3 /i4 Å (Fig. 1, Extended Data Fig. 2-4 97 and Extended Data Table 1). The structures of hTauT in the apo state, in complex with 98 taurine, β -alanine and GABA, were prepared in the presence of 150 mM NaCl, while 99 the GES-bound structure was prepared in the presence of 120 mM KCl. These five 100 structures exhibit inward-open and occluded conformations. 101 The structure of hTauT APO features a canonical LeuT fold with pseudo-twofold 102 symmetry organization of TMs 1–5 relative to TMs 6–10 and the transmembrane 103 domain displays an inward-opening (Fig. 1a and 1b) conformation. The central 104 substrate- binding site is accessible from the cytoplasmic side (Extended Data Fig. 5a). 105 Density of the N-terminal sequence and TM1a (residue 1-57) was less clear in the 106 inward-opening state due to flexibility (Extended Data Fig. 4b). Intracellular loops 107 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint (ILs) and extracellular loops (ELs) except for EL2 (residue 180-188), connecting the 108 TMs were unambiguously modelled into the density. The closed extracellular gate is 109 formed by interactions between residues from TM1b, TM3, TM6a and TM10 110 (Extended Data Fig. 5b). Density of Cl − was observed at the conserved 111 chloride-binding site (Extended Data Fig. 5c). Ligand-bound hTauT structures in the 112 presence of NaCl all adopt occluded states, with one molecule binding in the central 113 binding pocket formed by the unwound regions of TM1 and TM6, together with 114 adjacent residues in TM3 and TM8 (Fig. 1c-1e). The GES-bound structure in the 115 presence of KCl adopts an inward-facing conformation, with two molecules 116 positioned within the central cavity (Fig. 1f). 117 Taurine and ion binding pockets 118 Human TauT (hTauT) exhibited taurine transporter activity with a Km of 7.62 ± 119 1.96/i4 µM (Fig. 2a). The structure of hTauT in complex with taurine adopts an 120 occluded conformation at an overall resolution of 3.3 Å (Fig. 2b and 2c, Extended 121 Data Fig. 3b). The superposition of taurine-bound hTauT with hTauT APO shows 122 overall root mean-square deviation (r.m.s.d.) values of 1.86 Å, with TM1a undergoing 123 an obvious conformational change that occludes the intracellular permeation pathway 124 (Fig. 2b and 2c). TM1, TM3, TM6 and TM8 enclosed the central substrate-binding 125 pocket S1 (Fig. 2d). Density in the pocket was identified as taurine molecule and 126 additional densities of Na + and Cl − ions were observed in the conserved Na1 and 127 chloride binding site 39 (Fig. 2d). According to the three subsite (A, B and C) 128 representation of the central binding pocket, taurine occupies subsite A with its sulfate 129 group and a water molecule was observed in subsite C (Fig. 2d). One oxygen atom of 130 the sulfate group of taurine coordinates Na +, one oxygen atom forms hydrogen bond 131 with nitrogen atom of Gly62 and the other oxygen atom forms hydrogen bond with 132 hydroxyl group oxygen of Ser 402 and nitrogen atom of Gly60 (Fig. 2e). The amino 133 group of taurine forms a hydrogen bond with carbonyl group of F300, while the 134 carbon linker is stabilized by the side chain of Tyr138 and Phe300 (Fig. 2e). Na + at 135 Na1 site is coordinated by the side chains of Asn63, Asn333, Ser301 and carbonyl 136 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint group of Phe58 and Ser301 (Fig. 2f, Extended Data Fig. 6a). The Cl − is chelated by 137 the side chains of Tyr83, Gln297, Ser337 and Ser301 (Fig. 2g, Extended Data Fig. 6d). 138 No density corresponding to Na+ ion was observed at the conserved Na2 site, however 139 a taurine like density was identified near the Na2 site, surrounded by TM1 (Val59), 140 TM5 (Ala250 and Phe254) and TM8 (Leu398 and Ser402) (Extended Data Fig. 6g). 141 To validate the function of residues surrounded taurine and ions, mutagenesis 142 experiments were performed by measuring the uptake of radiolabeled taurine of TauT 143 mutants. Substitutions of S402A, Y138A and F300A which coordinate taurine leads to 144 complete loss of the transport activity (Fig. 2h). Substitution of Ser301, which 145 coordinates Cl− and Na+ at Na1 site, to alanine also abolished the transport of taurine 146 (Fig. 2h). E406A mutation at subsite B do not affect the transport activity of TauT, 147 however, substitution of L134A and N135A completely abolish taurine transportation 148 (Fig. 2h). 149 Binding mode of β -alanine 150 Previous results have revealed that the best inhibitor for TauT is linear substrate 151 analogues especially β -alanine which have a three-carbon linker and a amine group 152 identical to taurine 5. β -alanine shows inhibitory activity with IC 50 values of 31.65 ± 153 10.76/i4 µM (Fig. 3a), which is similar to the value measured using HEK293 cells 36. 154 The change of acidic moiety to carboxy group decreased the inhibitory effect 12-fold 155 compared with sulfate group (Extended Data Fig. 1c). The structure of hTauT β A 156 adopts occluded conformation which is almost identical to hTauT TAU with the overall 157 r.m.s.d. values of 1.16 Å (Fig. 3b-3c and 3e-3f). β -alanine binds at the same site as 158 taurine at subsite A and densities of water, Na + and Cl − were observed at conserved 159 subsite B, Na1 and chloride-binding site, respectively (Fig. 3d, Extended Data Fig. 6b 160 and 6e). One oxygen atom of the carboxy group forms hydrogen bond with carbonyl 161 group of Ser301 and the other oxygen coordinates Na + and form hydrogen bond with 162 nitrogen atom of Gly62 (Fig. 3d). The amino group of β -alanine also forms hydrogen 163 bond with carbonyl group of F300 and the carbon linker is stabilized by the side chain 164 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint of Tyr138 and Phe300 (Fig. 3d). At conserved Na2 site, a β -alanine like density was 165 observed similar to the density of hTauTTAU (Extended Data Fig. 6h). 166 The major differences between the substrate-free and substrate-bound hTauT is TM1a 167 and N-terminal loop (Fig. 3f). The conformational change of TM1a of hTauT APO 168 swings F58 away from the pocket, makes the central pocket accessible from the 169 cytoplasmic side (Fig. 3f-3g). This movement disrupts the Na1 binding site and leaves 170 the chloride binding site unaffected (Extended Data Fig. 6j and 6l). Hydrophobic 171 interaction between residues of TM1 (Phe50, V al51, Val54, Ala55 and Phe58), TM2 172 (Phe95, Phe96 and Ile99) and TM6 (Met309 and L312) and TM7 (L329) stabilized 173 the occluded conformation (Fig. 3h). The clear N-terminal and C-terminal densities of 174 hTauTTAU enabled us to identify the salt bridges formed at the occluded conformation 175 (Extended Data Fig. 4a). Interaction between side chain of Arg41-Asp416 and 176 Lys317-Glu593-Lys43 further stabilized the occluded conformation (Fig. 3i). 177 Coordination of GABA 178 GABA, the linear inhibitor TauT, is different from β -alanine with a one atom longer 179 carbon linker (Fig. 4a). Inhibitory effect of GABA has been reported with IC 50 values 180 of 1014/i4 µM and the interaction of GABA-mimetics with TauT has been investigate 181 37. The IC 50 value of GABA for inhibiting hTauT is 66.42 ± 53.56 /i4 µM, around 15 182 times more potent than previous reported (Fig. 4a). Comparing with β -alanine, longer 183 carbon linker does not significantly decrease the inhibitory effect of GABA (Fig. 3a 184 and 4a). GABA-bound hTauT is in the occluded conformation with GABA molecule 185 binds at the central cavity (Fig. 4b and 4c). GABA bind at subsite A with its carboxy 186 group and its amino group extends toward subsite B, and densities of water molecules 187 were observed at subsite B and C (Fig. 4d). Densities of Na+ and Cl− were observed at 188 conserved Na2 site and chloride-binding site (Fig. 4d, Extended Data Fig. 6f and 6i). 189 No clear Na+ density and ligand-like density were observed at Na1 site and near Na2 190 site as observed in maps of hTauTTAU and hTauTβ A (Extended Data Fig. 6c and 6g-6i). 191 The carboxy group of GABA forms a hydrogen bond with nitrogen of Gly62 and the 192 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint hydroxyl group of Tyr138 (Fig. 4e). The amino group of GABA forms a hydrogen 193 bond with the side chain of Glu406 and a water molecule (Fig. 4e). Na + at the Na2 194 site is coordinated by the side chains of Asp401 and Ser402, the carbonyl oxygens of 195 Gly56, Val59 and Leu398 (Fig. 4f). Superposition of hTauT APO with hTauT GABA 196 shows the carbonyl group of Gly56 swings away from the binding pocket during 197 occluded to inward-facing conformational change (Extended Data Fig. 6k). 198 Comparing with the central binding pocket of GAT1 shows that residues consist of 199 subsite A are identical (Extended Data Fig. 1a and 7a). Substitution of Y60G and 200 G279A from GAT1 to TauT may reduce the binding affinity of GABA which is 201 consistent with previously reported functional experiments 33. Substitution of Glu406 202 to alanine, which does not affect the binding of taurine, reduced the inhibitory effect 203 of GABA (Fig. 4a). The GABA coordination function of negatively charged glutamate 204 is supported by the fact that Tyr60 (corresponding to Gly57 near Glu406) is 205 substituted to glutamate in GAT2 and GAT3. Inhibitory effect of nipecotic acid, a 206 GABA-mimetic, was reported with IC 50 values of 2.02 /i4 mM 37. Comparison of 207 hTauTGABA with occluded nipecotic acid-bound GA T1 demonstrates that nipecotic 208 acid may be coordinated by hTauT in a similar mode (Extended Data Fig. 7b). 209 Inhibition of guanidinoethyl sulfonate 210 Guanidinoethyl sulfonate (GES), a substrate-mimetics of TauT, acts as a competitive 211 inhibitor of taurine transport and has been reported effectively reduced taurine level in 212 brain and plasma 40,41. GES shows inhibitory activity with IC 50 values of 5.27 ± 213 1.46/i4 µM , similar to the effect of taurine (Fig. 5a). GES-bound hTauT sample was 214 prepare in the presence of KCl and the structure adopts an inward-facing 215 conformation with the central pocket accessible from the cytoplasmic side (Fig. 5b 216 and 5c). The binding mode of GES is different from other substrate analogues with 217 two molecules binding at the central pocket (Fig. 5b and 5c). 218 Superposition of hTauTGES with hTauT TAU shows that the conformational change of 219 TM1a leads to the opening of the binding pocket of the second GES molecule (Fig. 220 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint 5d). Guanidine group of GES1 occupies at the same site as the amino group of taurine, 221 while the sulfate group of GES2 overlaps with the sulfate group of taurine (Fig. 5d). 222 Several densities of water molecules were observed surrounded GES molecules (Fig. 223 5e). Density of GES1 located within the central pocket with sulfate group binds at 224 subsite C and guanidine group coordinate by water molecule at subsite B (Fig. 5e). 225 GES2 is parallel to the substrate translocation pathway with its sulfate group 226 occupying subsite A and guanidine group facing the cytoplasmic side (Fig. 5e). 227 Sulfate group of GES1 forms hydrogen bonds with the back bone of Phe300 and 228 Ala463, guanidine group interact with Glu 406, sulfate group of GES2 and water 229 molecule coordinated by the side chain of Asn135 and Gln403, and carbon linker is 230 coordinated by Phe300, Ala303 and Leu306 (Fig. 5f). Similar to the binding mode of 231 acidic group of taurine, sulfate group of GES2 occupies subsite A. Oxygens of GES2 232 form hydrogen bonds with GES1, Gly62, Tyr138 and Ser402 (Fig. 5g). Guanidine 233 group is coordinated by hydrogen bond network formed between Ser301, Asn333, 234 Gly57, Phe58 and water molecules (Fig. 5g). The substitution of E406A reduced the 235 inhibitory effect of GES, similar to GABA, demonstrates the function of Glu406 in 236 the coordination of ligand with longer linker between acidic group and basic group 237 (Fig. 5a). Comparison with tiagabine-bound GAT1 in the inward-facing conformation 238 shows that sulfate group of GES2 occupies the same site as carboxy group of 239 tiagabine and the movement of TM1a opens the binding pocket of the groups on the 240 other end of these inhibitors (Extended Data Fig. 7c). 241

242 Here we presented the structures of human TauT in the apo state, in complex with 243 taurine and in complex with substrate-mimetic inhibitors with different acidic groups, 244 basic groups and lengths of carbon linkers. Comparing to the inward-facing 245 conformation of substrate-free hTauTAPO, the structure of hTauTTAU reveals the central 246 binding pocket and the coordination mode of Na+ at the Na1 site and Cl− at conserved 247 chloride-binding site, similar to GABA-bound GAT133. Residues in subsites B and C, 248 which do not directly coordinate taurine, may participate in shaping the central cavity 249 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint and the coordination of water molecules. Water molecules in the cavity help 250 coordinate the substrate by forming hydrogen bonds and acting as proton donors, 251 facilitating the hydrolysis of the basic tail and the formation of cation- π interaction 252 between taurine and Tyr138 and Phe300. The presence of Na + and the substrate 253 stabilizes hTauT in the occluded state, as the replacement of NaCl with KCl leads to 254 the inward-facing conformation of hTauT GES. Taurine- and β -alanine-like densities 255 observed near the conserved Na2 site probably stabilize the occluded conformation, 256 mimicking the function of Na + at Na2 site and preventing the release of substrates to 257 the cytoplasmic side, as high intracellular taurine levels reduce transport activity of 258 TauT 3,42. 259 The elevated expression level of the TauT transporter is associated with poor 260 prognosis and advanced tumor stage in various cancers, suggesting the importance of 261 taurine in cancer development 5,23-25. Recent study reported the pro-tumoral role of 262 taurine and indicated that its anticancer effects in earlier studies might have been due 263 to the restoration of antitumor immunity, which is impaired by the exhaustion of 264 environmental taurine depleted by tumor cells 43. Inhibition of taurine transportation 265 could be a promising strategy in anticancer therapy. The structures of hTauT β A, 266 hTauTGABA and hTauTGES reveals the binding modes of these linear inhibitors. All of 267 these molecules occupy subsite A with their acidic heads. Taurine includes an ethane 268 backbone with one amino group attached to one carbon and one sulfate group 269 attached to the other carbon. The length of β -alanine is similar to taurine, and they 270 bind at the same site in the central pocket of hTauT. Functional studies indicate that 271 the sulfate group probably enhance the affinity of ligands by donating protons, which 272 facilitates the coordination of Na + at Na1 site and interact with more residues in 273 subsite A. In comparison, GABA, with a four-carbon backbone, has a similar 274 inhibitory effect to β -alanine. The amino tail of GABA is coordinated by Tyr138 and 275 Phe300 through cation- π interaction and Glu406 facilitates the binding of zwitterion 276 with longer carbon linkers. GES, a more complex structure, shows potent inhibition of 277 hTauT and may have unexplored antitumor properties. Base on the binding mode of 278 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint GES, additional groups attached to the amino group of taurine could be 279 accommodated in subsite C, providing a blueprint for developing more effective 280 inhibitors. Our findings provide a detailed structural basis for understanding TauT 281 function and its inhibition, offering valuable insights for future therapeutic 282 development. 283

284 Cell lines 285 FreeStyle 293 F (Thermo Fisher Scientific) suspension cells were cultured in 293 286 Expression Medium (Gibco) supplemented with 1% FBS at 37 °C, in an atmosphere 287 of 6% CO2 and 60% humidity. Sf9 insect cells (Thermo Fisher Scientific) were 288 cultured in Sf-900 III SFM medium (Thermo Fisher Scientific) at 27 °C. 289 Oocyte expression systems and flux experiments 290 Xenopus laevis oocytes were isolated and maintained as previously described44. 291 Briefly, ovarian fragments were isolated from the animal and digested with 292 collagenase (2mg/ml, type IA, Sigma-Aldrich) for 2 h in OR2 buffer containing 82.5 293 mM NaCl, 2 mM KCl, 1 mM MgCl2 and 5 mM Hepes (pH 7.4). Oocytes were then 294 washed in ND96 buffer containing 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM 295 MgCl2 and 5 mM HEPES (pH 7.8). Stage VI oocytes were transferred to fresh OCM 296 containing 60% Leibovitz L15, 100 μ g/ml gentamycin and 10 mM Hepes (pH 7.8) 297 and were maintained at 18 - 20 °C prior to injection. cDNAs of wild-type and mutant 298 hTauT were transferred into the pKSM vector. cRNA was obtained using the T3 299 mMessage mMachine Kit (Thermo Fisher Scientific) and 25 ng injected into selected 300 oocytes. Oocytes were incubated at 20/i4 °C for 40-48 h in ND-96 solution with 100 301 μ g/ml gentamycin to allow expression of TauT. Uptake was measured in buffer 302 containing 100 mM NaCl, 2 mM KCl, 1 mM CaCl2, 1 mM MgCl2 and 1 mM HEPES 303 (pH 7.4) and medium containing the desired amino acid (Thermo Fisher Scientific) 304 concentration and 10 μ Ci/ml [3H]taurine (PerkinElmer). For the assessment of 305 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint time-dependent accumulation of taurine, oocytes were incubated in buffer containing 306 4 μ M taurine (10 μ Ci/ml). For assessment of mutant TauT activity relative to 307 wild-type as well as the inhibitory effects of inhibitors, oocytes were incubated in 308 buffer containing 0.25 μ M taurine (10 μ Ci/ml). Oocytes were solubilized by adding 1% 309 Triton followed by liquid scintillation fluid for to quantitate radioactivity. Activity of 310 TauT were calculated by subtracting the uptake of the non-injected group from that of 311 the cRNA-injected groups. 312 Expression and purification of TauT 313 cDNA encoding human wild-type TauT was cloned into the BacMam expression 314 vector with C-terminal GFP-His-Strep tags45. The baculoviruses were produced using 315 the Bac-to-Bac system and amplified in Sf9 cells. For protein expression, HEK293F 316 cells cultured in Freestyle 293 medium at a density of 2.5/i4 ×/i4 106/i4 ml−1 were 317 infected with 10% volume of virus. Sodium butyrate (10 mM) was added to the 318 culture 8 h post-infection and cells were harvested 60 h post-infection. Cells were 319 collected by centrifugation at 3,724g for 10 min and washed with 20 mM HEPES (pH 320 7.4) containing 150 mM NaCl. 321 Cell pellets were solubilized in buffer containing 150 mM NaCl, 20 mM Hepes at pH 322 7.4 (HBS), 1% (w/v) lauryl maltose neopentyl glycol (LMNG; Anatrace) and 0.1% 323 (w/v) cholesteryl hemisuccinate tris salt (CHS; Anatrace) at 4/i4 °C for 1/i4 h. Insoluble 324 components were removed by centrifugation at 70,000g for 30 min. The supernatant 325 was supplemented with 10 mM imidazole and loaded onto a 5 ml Ni-NTA column 326 (GE Healthcare). Beads were first washed with HBS containing 0.01% (w/v) LMNG 327 and 0.001% (w/v) CHS then with HBS containing 0.01% (w/v) LMNG, 0.001% (w/v) 328 CHS and 20 mM imidazole. Proteins were eluted using HBS with 0.01% (w/v) 329 LMNG, 0.001% (w/v) CHS and 250 mM imidazole, and tags were removed using 330 TEV protease. For purification of TauT, the eluate was subjected to SEC using a 331 Superose 6 Increase 10/300 GL column (GE Healthcare) in buffer containing 20 mM 332 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint HEPES pH 7.4, 150 mM NaCl, 0.01% (w/v) LMNG and 0.001% (w/v) CHS. The 333 eluted peak corresponding to TauT was collected for cryo-EM sample preparation. 334 Cryo-EM sample preparation and data acquisition 335 TauTAPO, TauTTAU (supplemented with 1 mM taurine), TauTGABA (with 1 mM GABA), 336 TauTβ A (supplemented with 1 mM β -alanine) and TauTGES(supplemented with 200 337 μ M GES) were diluted to/i4 9 mg/ml and aliquots (3/i4μ l) were applied to GF-0.6/1.0 338 300 Au-flat (Electron Microscopy Sciences). The cryo-EM grids were prepared using 339 Thermo Fisher Vitrobot Mark IV at 8/i4 °C and 100% humidity (Thermo Fisher 340 Scientific), and data were collected in a Titan Krios microscope (Thermo Fisher 341 Scientific) at 300 kV for data acquisition. Images were collected using a K3 camera 342 (Gatan) mounted post a Quantum energy filter with a 20/i4 eV slit and operated in 343 super-resolution mode with a pixel size of 0.67/i4 Å at the object plane. Defocus values 344 were set to range from -1.5 μ m to -2.0 μ m for data collection. Data were acquired 345 using EPU-2.9.0.1519REL. The dose rate on the detector was 15.0 e-s-1A-2 with a total 346 exposure of 56 e-A-2. Each 2.56 s movie was dose-fractioned into 32 frames. 347 Image processing 348 CryoSPARC software was used for structural analysis46. The original movies were 349 gain-corrected, motion-corrected and binned to a pixel size of 0.67 Å in a Patch 350 motion correction step. Dose-weighted micrographs were used for CTF estimation 351 using Patch-CTF in cryoSPARC. Particles were auto-picked using Topaz-0.2.347, 352 extracted with a pixel size of 2.68/i4 Å and subjected to two-dimensional (2D) 353 classification to remove contaminants. Particles showing obvious secondary structure 354 features were re-extracted with a pixel size of 1.34/i4 Å and subject to several rounds 355 of ab-initio reconstructions. Seed-facilitated 3D classification was performed to 356 improve the resolution48, by several rounds of heterogenous refinement and ab-initio 357 reconstructions. Final good particles were re-extracted with a pixel size of 0.67/i4 Å. 358 Non-uniform and local refinement were performed, and resolutions were estimated by 359 gold-standard Fourier shell correlation. 360 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint Model building 361 Alphafold49 was used to build initial models of TauT and docked into the cryo-EM 362 map with UCSF Chimera-1.1450. Models were manually rebuilt in Coot-0.9.251 and 363 further refined by Phenix1.19.2-415852. CIF files for ligands were generated in Phenix 364 using eLBOW. The residues contained in the final models are indicated in Extended 365 Data Table 1. Figures were prepared with Pymol-1.7.0.5 (Schrodinger, LLC.) and 366 UCSF ChimeraX-1.551. 367 Flow cytometry 368 293F cells were transfected with HA tagged wild-type and mutant constructs as 369 previously described53. Cells were collected 27 h post-transfection and incubated with 370 PE conjugated HA tag antibody (BioLegend, cat# 901518) for 30 min at room 371 temperature, followed by analysis on a flow cytometer (BD LSRFortessaTM). FACS 372 data were analyzed using FlowJo v10.6.2 software. 373 Quantification and statistical analysis 374 Global resolution estimations of cryo-EM density maps are based on the 0.143 375 Fourier Shell Correlation criterion54. The local resolution map was calculated using 376 cryoSPARC-4.6.146. Prism GraphPad software v9.3.0 was used for statistical analysis. 377 The number of biological replicates (N) and the relevant statistical parameters for 378 each set of experiments (such as mean or standard error) are described in the figure 379 legends. P values were calculated by two-sided unpaired t-tests. No statistical methods 380 were used to pre-determine sample sizes. 381

382 1 Santu l li, G. e t a l . Fun ctiona l R ole o f Taurine in Aging and Cardiovas cu lar 383 Health: An Updated Overview . Nutrie nts 15 , doi:10.3390/nu15 194236 (202 3). 384 2 Rafiee, Z., Garcia-Se rrano, A . M. & D uarte, J. M . N . T aurine Supplementation 385 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint as a Neuroprotective Str ategy upon Br ain D ysfunction in Metabol ic Syndrome 386 and Diabetes. Nutri ents 14, do i :10.3 3 90/nu14061292 (2022) . 387 3 Baliou, S . e t a l . S i gnifican ce of taur i n e transporter ( TauT ) i n homeostasis and 388 its l ayer s of regulat ion (Review) . M ol Med Rep 22 , 2163-21 73, 389 doi:10.3892/mmr .2020.11321 (2020). 390 4 Br oer, S. & Gether, U. Th e sol ute c a rrier 6 family of transpor ters. Br J 391 Pharmacol 167 , 256-278, doi:10.1111 /j.1476- 5381.2012.01975.x (2012) . 392 5 Star y , D. & B ajda, M. Taurine and Creatine Transpor ters as Potential Drug 393 Targe ts i n C ancer Therapy. Int J Mol Sci 24 , doi:10.33 90/ i jm s24043 788 394 (2023) . 395 6 Bhutia, Y . D ., Mathew, M., S i vapra k as am, S ., R amachandr an, S. & Ganapath y , 396 V. Unconv e nti ona l Functions of A mi no A c id Transpor ters: Role i n 397 Macropinoc ytos i s (S LC 38 A5/ SLC38 A 3) and D iet-Induced Obesity/Met a bolic 398 Syndr ome (SLC6 A19/S LC 6A14 /SL C 6 A 6). Biomolecules 12 , 399 doi:10.3390/biom12020235 (2022). 400 7 Pasantes-M orales, H . Tau rine H o meostas is and Volume Control. Adv 401 Neurobiol 16 , 33-53, doi:10.1007/97 8 -3-319-55769- 4_3 (2017). 402 8 Schousboe, A. & P asantes-Morales, H. Role of taur ine in neural c e l l v olu me 403 regulation. Can J Physiol P harmacol 70 Suppl , S356-361, 404 doi:10.1139/y92-283 ( 1992). 405 9 Chen, W., G uo, J. X. & C ha ng, P. T he effect of taurine on cho l este r ol 406 metabolis m. Mol Nutr Food Res 56, 681- 690, doi :1 0.1002/mnfr.201100 799 407 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint (2012) . 408 10 Qaradakhi, T. et al. The A nti-Inflammatory Effe c t of T aurine on Ca rdiovas c ular 409 Di sease. Nutrients 12 , doi:10.3390/ nu 12092 847 (2020). 410 11 Duan, H., Song, W ., Guo, J. & Yan, W . Taurine: A So urce and App l icatio n for 411 the Reli ef of Visua l Fatigue. Nu trients 15 , do i:10.3390/nu15081843 (2 023). 412 12 Pion, P . D ., K ittles on , M . D ., Rogers, Q. R. & Morris, J. G. Myocardial failur e i n 413 cats as sociat ed w ith low plasma taur i n e: a reversible c ar di o m y opa th y . 414 Science 237 , 764- 768, doi:10.1126/ s ci ence.3616 60 7 (1987). 415 13 Leon, A ., Levick, W. R. & Sa r os sy, M. G. Le sion t opography and n ew 416 histolog i ca l fea tures in fel ine taurine de fi cien cy ret inopathy. Exp Eye Res 61 , 417 731-741, do i :10.1 016/s0014-4835(05) 80024-7 (1995) . 418 14 Whar ton, B. A., Mo r l ey, R., Isaacs, E. B., Cole, T. J. & Lucas, A. Low p l asma 419 taurine and later n eurodevelopment. Arch Dis Child Fetal Neonatal Ed 89 , 420 F497- 498, doi:10.1136/adc.2003.048 389 (20 04). 421 15 Schuller-Levis, G., Mehta, P. D., Rudel li, R. & Sturman, J . I mmunol ogic 422 consequence s of taurine deficien cy in cats. J Leukoc B iol 47 , 321-3 31, 423 doi:10.1002/jl b.47.4.321 (1990 ). 424 16 Jong, C. J., Sandal, P . & Schaffer, S . W. The Rol e of Taurine in M itochon dr i a 425 Health: Mor e Th an Ju st an A ntioxi dant . Molecules 26 , 426 doi:10.3390/mole cule s26164913 (2021) . 427 17 Ma, N., H e, F., K a w ano k uchi, J ., Wa ng, G. & Yamashita, T. Taur i n e and I ts 428 Anticancer Function s: In V iv o and I n Vitro Study. Adv Exp Med B iol 1370 , 429 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint 121-128, do i :10.1 007/978-3-030- 93337-1_11 (2022) . 430 18 Baliou, S. e t a l . P rotective r ole of ta urine against ox i dati ve stres s (Review). 431 Mol Med Rep 24 , doi:10.3892/m mr.20 21.12242 (2021). 432 19 Miyamoto, Y. , Tiruppathi, C., Ganapa thy, V . & L ei bach, F. H. Acti ve transp ort 433 of taurine in rabbit jejuna l brush-border membrane vesic l es. Am J Physi ol 257 , 434 G65-72, doi:10.1152/a j pg i.1989.25 7. 1.G 65 (1989). 435 20 Ramamoorthy, S. e t a l . F uncti on al char ac ter iz a tion and chr omoso mal 436 local i zat ion of a cloned taurine tran spor ter fro m human pl acenta. Biochem J 437 300 ( Pt 3) , 893- 900, do i:10.1042/bj3 000893 (1994) . 438 21 Ito, T. & Mur ak am i , S. Tau rine defici en cy assoc iated w ith di lat ed 439 cardiomyopath y and aging. J Pharmacol Sci 154 , 175-181, 440 doi:10.1016/j. jphs.2023.12.006 (202 4 ). 441 22 Hayes , K. C., Carey, R. E . & Schmidt, S . Y. R etinal degene ration as socia ted 442 wi th taurine defic ien cy in t he cat. Science 188 , 949-951, 443 doi:10.1126/s cience. 1138364 (1975). 444 23 Dazhi , W ., Jing, D., C hunling, R., Mi, Z. & Z hix uan , X . Elevated SL C 6A6 445 expressio n drive s tumor i gen esi s an d affects clini cal out comes in ga str ic 446 cancer. Biomark M ed 13 , 95-104, doi: 10.2217/bm m-2018-0256 (2 019). 447 24 Yasunaga, M. & Matsumura, Y. Role of SLC6 A6 in p r omoting the surviv al and 448 multidrug resi stance of colore ctal can cer. Sc i R ep 4 , 4 852, 449 doi:10.1038/srep0485 2 (2014). 450 25 Baliou, S ., K y r i akopou l os, A. M., Spa ndidos, D. A. & Z oumpourlis, V. R ole of 451 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint taurine, its ha loamine s and its lncR N A TUG1 in both inflammation and ca ncer 452 progre s sion. On the road to therapeutics? (Rev i ew). Int J Oncol 57 , 631-6 64, 453 doi:10.3892/ij o.2020.5100 (20 20). 454 26 Zhang, H . e t a l. D imerization and antidepr ess ant recognition at noradr enaline 455 transporter . Natur e 630, 247-254, do i: 10.1038 /s 415 86-024-07437- 6 (2024). 456 27 Sr i vastava, D. K . e t a l . S tructure of the human dopamine transporter a nd 457 mechanis ms of i nhi bition. Nature 632, 672-6 77, 458 doi:10.1038/s41586-0 24-07739-9 (202 4). 459 28 Ji, W . e t a l . Substrate binding and inhibition me chani s m of norepinephrine 460 transporter . Natur e 633, 473-479, do i: 10.1038 /s 415 86-024-07810- 5 (2024). 461 29 Hu, T. e t a l. Tr ansport and inhibition mechanism s of the human noradr ena li ne 462 transporter . Natur e 632, 930-937, do i: 10.1038 /s 415 86-024-07638- z ( 2024). 463 30 Singh, I. e t a l . S tructure-based d isc o very of conformationa ll y se lect iv e 464 inhibitors of the serotonin trans porter. Cell 186, 2160- 2175 e2117, 465 doi:10.1016/j. cell.202 3.04.010 (2023) . 466 31 Shahsavar, A. e t a l . Str uctural insigh ts into the inhibit ion of glyc ine reuptake. 467 Nature 591 , 677- 681, doi:10.1038/ s4 1586-021-0327 4-z (2021). 468 32 Col e man, J. A. e t a l . Se r otonin tra nsporter- i boga i ne comple xes i ll um in ate 469 mechanis ms o f inh ibit ion and transport. Nature 569, 141- 145, 470 doi:10.1038/s41586-0 19-1135-1 (201 9). 471 33 Zhu, A . et al. Molecu l ar bas i s for subs trate recogniti on and transport of hum an 472 GABA transporter G AT 1. Nat Struct Mol B iol 30 , 1012-1 022, 473 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint doi:10.1038/s41594-0 23-00983-z (2023) . 474 34 Motiwal a, Z. e t a l . S tructural basis of GAB A reu ptak e inhibiti on. Nature 606 , 475 820-826, do i :10.1 038/s41586-022-04814- x (2022). 476 35 Kur th, I. e t a l . Therapeutic targeting of SL C 6A 8 c r eatine transporter 477 suppresse s colon cancer progress io n and modulates human c r eatine le v els . 478 Sci Adv 7 , eabi7511, doi:10.1126/ scia dv.abi7511 ( 2021). 479 36 Ri chter , M., Moronia k, S. J . & M ichel, H . Iden ti fi cation of co mpetiti ve inhibit or s 480 of the human taurine transporter Tau T in a hum an ki dne y ce ll l ine. Pharmac ol 481 Rep 71 , 121-129, doi:10.101 6/ j .pha re p.2018.10.005 (2019 ). 482 37 Rasmus sen, R . N ., Lagunas, C., P lum, J., Holm, R. & Niel sen, C . U . 483 Interaction of GABA- mi me ti cs w ith the taur i ne transporter (T au T, Slc6a6) i n 484 hyperosmoti c treated Caco-2, LLC-P K 1 and rat renal SK P T cel ls. Eur J Ph arm 485 Sci 82 , 138- 146, doi:1 0.1016/j.e jps . 20 15.11.020 (2016). 486 38 Goehring, A . e t a l . Screening and large- s cale expr es s ion of membr ane 487 proteins in mam malia n ce lls for s tr uctural studies. Nat P rotoc 9 , 2574-2 585, 488 doi:10.1038/npro t.2014.173 (2014). 489 39 Wang , K . H., Penmatsa, A. & Gouau x , E . Neurotran s mitter and 490 psycho s timu lant recognit ion by th e dopamine transporter. Nature 521 , 491 322-327, do i :10.1 038/nature14431 (2 015). 492 40 Suar ez, L. M. e t a l. The ta urine trans porter s ubstr ate guanidinoethy l s u lfon ate 493 mimi cs the act ion of taurine on long- term s ynapt ic pote nti at ion. Amino A cids 494 48 , 2647- 2656, doi:10.1007 /s00726-0 16-229 8-9 (2016). 495 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint 41 Ejiri, K ., Akahori, S ., Kudo, K ., Sekiba, K . & Ubuka, T. E ffect of guan idinoet hyl 496 sulfonate on taurine concentr ati o ns and fetal growth in pregn ant r ats. Bi ol 497 Neonate 51 , 234-240, doi :10 .1159/ 00 0242658 (1987) . 498 42 Han, X., Budreau, A. M . & Chesney, R. W . The tau rine transporter gene and 499 its ro l e in r enal de velopme nt. Amino Acids 19 , 499-5 07, 500 doi:10.1007/s007260070002 (2000). 501 43 Cao, T. et al. C ancer SLC6 A6 -m edi at ed taurine uptake transactivate s imm une 502 checkpo i n t genes and induce s ex hausti on in CD 8( +) T cell s. Cell 187 , 503 2288-2304 e2227, doi:10.1016/j. cell .2 024. 03.011 (2024). 504 44 Plautz, C . Z., W ill iams, H . C . & Grai nger, R . M. Functional C loning U sin g a 505 Xenop us Oocy te Express ion S ystem . J V i s Exp , e535 18, doi:10.3791/53 518 506 (2016) . 507 45 Li, N. e t a l . S tructure of a Pancreat i c A T P-Sensitive Potassium Channel. Cell 508 168 , 101- 110 e110, doi:10.1 016/j.ce ll . 2016. 12.028 (2017). 509 46 Punjani, A., Rubinste in, J. L., Fleet, D . J. & Br ubaker, M. A. cryoSP A RC : 510 algorithms for r apid un supervi sed cryo-EM structure determination. Nat 511

14 , 290- 296, doi:10.1038/n meth.4169 (20 17). 512 47 Bepler , T. e t a l . P osi ti ve-unlabeled convolutiona l neu ral networks for parti cl e 513 pick i ng i n cryo-ele ctron mi c r ographs. Nat Methods 16 , 1153-11 60, 514 doi:10.1038/s41592-0 19-0575-8 (201 9). 515 48 Wang , N. e t a l . Stru c tur al ba sis o f human monocarbo xylate transporte r 1 516 inhibit i on by anti-can cer drug ca ndi da tes . Cel l 184 , 370- 383 e 313, 517 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint doi:10.1016/j. cell.202 0.11.043 (2021) . 518 49 Jumper, J . e t a l . H igh ly a ccurate pro tein structure predi ction w i th AlphaFold. 519 Nature 596 , 583- 589, doi:10.1038/ s4 1586-021-0381 9-2 (2021). 520 50 Petter sen, E. F. e t a l . UC S F Chimer a- -a visual iz at ion system for e xplorat or y 521 research and analys is. J Comput Chem 25 , 1605-1612 , doi :10 .1002/jcc.20 084 522 (2004) . 523 51 Emsley, P., Lohkamp, B., S c o tt, W . G . & Cowtan, K. Featu r es a nd 524 development of Coot. A cta Crysta l logr. D Biol. Crystallogr. 66 , 486-5 01, 525 doi:10.1107/S 0907444910007493 (20 10). 526 52 Adam s , P . D . e t a l . P H ENIX: a co mprehensive Python- based system for 527 macromolecu lar structure s olution. Acta Crystal logr. D B iol. Crystall ogr. 66 , 528 213-221, do i :10.1 107/S09074449090 52925 (2010). 529 53 Lu, Y . e t a l . St ructur al insights into the conformat i onal changes of 530 BT R 1/ S LC 4A11 in comple x with PIP(2). Nat Commun 14 , 61 57, 531 doi:10.1038/s41467-0 23-41924-0 (202 3). 532 54 Rosenthal, P. B. & Henderson, R . O p timal deter mination of pa rticle or i enta tion, 533 absolute hand, a nd contrast los s in s i n gl e- partic l e elec tr on cry omicro scop y . J 534 Mol Biol 333 , 721- 745, doi:10.1016/ j.jm b.2003.07.013 (2003) . 535 536 Figure Legends 537 Figure 1. Cryo-EM structures of human TauT. 538 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint a, Cryo-EM density and the overall structure of human TauT (hTauT) in the apo state. 539 The Cl− ion is represented as green dot. 540 b, Topology of the secondary structure of hTauT. TM1-TM5 are colored in blue and 541 TM6-10 with inverted symmetry are colored in green. Extracellular and intracellular 542 helices are colored in yellow and pink, respectively. Dashed lines indicate the 543 disordered regions in the cryo-EM structure. 544 c-e, Structures of hTauT in complex with taurine (c), β -Alanine (d) and GABA (e) in 545 the presence of NaCl. Densities of ligands are shown as mesh. 546 f, Structures of hTauT in complex with guanidinoethyl sulfonate (GES) in the 547 presence of KCl. Densities of two GES molecules are shown as mesh. 548 Figure 2. Taurine and ion binding site of hTauT. 549 a, The transport kinetic of taurine uptake of hTauT. The Michaelis constant (Km) 550 value and Vmax for [3H]taurine uptake is 7.62 ± 1.96/i4μ M and 24.28 ± 2.04 551 pmol/oocyte/min. Data are shown as mean ± SD (n=3 biologically independent 552 experiments). 553 b, Structure of taurine-bound hTauT superimposes with hTauTAPO. TM1 of hTauTTAU 554 and hTauTAPO are colored in pink and light peach, repectively. Taurine is shown as 555 sticks. Na+ and Cl− ions are shown as dot. 556 c, Cut-open view of electrostatic surface of hTauTTAU with taurine binds in the central 557 pocket. 558 d, Model of the taurine-binding pocket. Cryo-EM densities of taurine, Na+, Cl− and 559 water are shown as mesh. Taurine is shown as sticks. Na+, Cl− and water are shown as 560 dot. 561 e, Close-up views of taurine at substrate binding pocket. Taurine and surrounding 562 residues are shown as stick and Cl− and Na+ ions are shown as dot. The coordination 563 of taurine is indicated by dashed lines. 564 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint f, Close-up views of Na+ at Na1 site. Key residues and taurine coordinate Na ion are 565 shown as stick and Cl− and Na+ ions are shown as dot. The coordination of Na is 566 indicated by dashed lines. 567 g, Close-up views of the chloride binding site. Key residues coordinate Cl− ion are 568 shown as stick and Cl− and Na+ ions are shown as dot. The coordination of Cl is 569 indicated by dashed lines. 570 h, Effects of alanine substitution of key residues in the taurine-bound pocket on 571 transport activity of hTauT. Data are expressed relative to wild-type hTauT and the 572 bars represent mean ± SD (n≥ 3 biologically independent experiments). 573 Figure 3. Binding mode of β -alanine and structural comparison with 574 inward-facing conformation. 575 a, Effect of β -alanine on taurine uptake by hTaut in oocytes. The IC50 value for 576 β -alanine is 31.65 ± 10.76/i4 µM. Data are shown as mean ± SD (n=3 biologically 577 independent experiments). 578 b, Structure of β -alanine-bound hTauT superimposes with hTauTAPO. TM1 of 579 hTauTβ A and hTauTAPO are colored in blue and light peach, respectively. β -alanine is 580 shown as sticks. Na+ and Cl− ions are shown as dot. 581 c, Cut-open view of electrostatic surface of hTauTβ A with β -alanine binds in the 582 central pocket. 583 d, Close-up view of the β -alanine-binding site. Densities of β -alanine, Na+, Cl− and 584 water are shown as mesh. β -alanine is shown as sticks. Na+, Cl− and water are shown 585 as dot. The coordination of β -alanine is indicated by dashed lines. 586 e, Superposition of hTauTβ A with hTauTTAU. TM1 and TM6 of hTauTβ A and 587 hTauTTAU are colored in blue and pink, respectively. β -alanine and taurine are shown 588 as sticks. 589 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint f, Structural comparison of hTauTAPO in the inward-facing conformation with 590 hTauTβ A and hTauTTAU in the occluded conformation. The arrow indicates the 591 movement of TM1a from inward-facing to occluded conformation. The black lines 592 indicate the same binding mode of taurine and β -alanine. Na+ and Cl− ions are shown 593 as dot. 594 g, Detailed conformational change of gating residues. Residues preclude ligand 595 release are shown as sticks. Distance between the Cα of F58 of hTauTTAU and 596 hTauTAPO is indicated by dashed line. 597 h-i, Hydrophobic interactions (h) and salt bridges (i) formed the intracellular gate. 598 Salt bridges are indicated by dashed lines. 599 Figure 4. Recognition of GABA in the central pocket. 600 a, Effect of GABA on taurine uptake by wild-type and E406A mutant hTauT in 601 oocytes. The IC50 value for GABA of wild-type and E406A mutant hTauT are 66.42 ± 602 53.56/i4 µM and 134.28 ± 25.92/i4 µM. Data are shown as mean ± SD (n=3 biologically 603 independent experiments). 604 b, Structure of GABA-bound hTauT superimposes with hTauTAPO. TM1 of 605 hTauTGABA and hTauTAPO are colored in orange and light peach, respectively. GABA 606 is shown as sticks. Na+ and Cl− ions are shown as dot. 607 c, Cut-open view of electrostatic surface of hTauTGABA with GABA binds in the 608 central pocket. 609 d, Model of the GABA-binding pocket. Cryo-EM densities of GABA, Na+, Cl− and 610 water are shown as mesh. GABA is shown as sticks. Na+, Cl− and water are shown as 611 dot. 612 e, Close-up views of GABA at substrate binding pocket. GABA and surrounding 613 residues are shown as stick and Cl− and Na+ ions are shown as dot. The coordination 614 of GABA is indicated by dashed lines. 615 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint f, Close-up views of Na+ at Na2 site. Key residues coordinate Na+ ion are shown as 616 stick and Na+ ion is shown as dot. The coordination of Na is indicated by dashed 617 lines. 618 Figure 5. Structural basis for inhibition of GES. 619 a, Effect of GES on taurine uptake by wild-type and E406A mutant hTauT in oocytes. 620 The IC50 value for GES of wild-type and E406A mutant hTauT are 5.27 ± 1.46/i4 µM 621 and 19.68 ± 2.42/i4 µM. Data are shown as mean ± SD (n=3 biologically independent 622 experiments). 623 b, Structure of GES-bound hTauT superimposes with hTauTAPO. TM1 of hTauTGES 624 and hTauTAPO are colored in green and light peach, repectively. GES is shown as 625 sticks. Cl− ion is shown as dot. 626 c, Cut-open view of electrostatic surface of hTauTGES with two GES molecules bind 627 in the central pocket. 628 d, Comparison of ligand binding pocket of hTauTGES and hTauTTAU. 629 e, Model of the GES-binding pocket. Cryo-EM densities of GES, Cl− and water are 630 shown as mesh. GES is shown as sticks. Cl− and water are shown as dot. 631 f, Close-up views of GES1 at substrate binding pocket. GES1, GES2 and surrounding 632 residues are shown as stick. Water molecule is shown as dot. Hydrogen bonds form 633 between residues, water and GES are indicated by dashed lines. 634 g, Close-up views of GES2 at substrate binding pocket. GES1, GES2 and surrounding 635 residues are shown as stick. Water molecules are shown as dot. Hydrogen bonds form 636 between residues, water and GES are indicated by dashed lines. 637 Extended data figures and tables 638 Extended Data Fig 1. Sequence alignment and purification of hTauT. 639 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint a, Sequence alignment of Homo sapiens TauT, Mus musculus TauT, Bos taurus TauT, 640 Homo sapiens SLC6A1 and Homo sapiens SLC6A8. Secondary structures of TauT 641 are shown above the alignment with TM1-5, TM6-10, TM11-12, extracellular helices 642 and cytosolic helices colored in blue, green, orange, yellow and pink, respectively. 643 Unmodeled residues are shown as dashed lines. Non-conserved residues are 644 highlighted in gray. Alignment was made using PROMALS3D. 645 b, Time-dependent accumulation of [3H]taurine in oocytes. Accumulation of oocytes 646 injected with (blue) and without (red) cRNA of wild-type human TauT. Data are 647 shown as mean ± SD (n=3 biologically independent experiments). 648 c, The dose-dependent inhibition of hTauT by taurine. The IC50 value for taurine of 649 wild-type hTauT is 2.63 ±1.04 µM. Data are shown as mean ± SD (n=3 biologically 650 independent experiments). 651 d, Size-exclusion chromatography profile and representative SDS-PAGE analysis of 652 hTauT in detergent. The asterisk indicates the fractions used for cryo-EM sample 653 preparation. 654 Extended Data Fig 2. Data processing of different TauT datasets 655 a, Representative raw electron micrograph of TauT. Scale bar, 50/i4 nm. 656 b, Two-dimensional class averages of hTauT output from cryoSPARC. 657 c-d, Flow chart showing data processing of hTauTAPO and ligand-bound hTauT. 658 Extended Data Fig 3. Cryo-EM analysis of TauT in different states. 659 a-e, Resolution estimation based on the criterion of the FSC 0.143 cut-off, angular 660 distribution of the final reconstruction and local resolution distribution of hTauTAPO 661 (a), hTauTTAU (b), hTauTβ A (c), hTauTGABA (d) and hTauTGES (e), respectively. 662 Extended Data Fig 4. Cryo-EM density of hTauT in the different states. 663 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint a, Cryo-EM density corresponding to the TM1-12, N-terminal and C-terminal of 664 TauTTAU. 665 b-c, Cryo-EM density corresponding to the TM1 of TauTAPO and TauTGES. 666 Extended Data Fig 5. Structural details in the inward-facing conformation. 667 a, Cut-open view of electrostatic surface of hTauTAPO shows the substrate 668 translocation pathway. 669 b, Close-up view of the residues in extracellular gate. Hydrogen bonds and salt 670 bridges are indicated by dashed lines. 671 c, Cryo-EM density corresponding to conserved chloride binding site. Residues are 672 shown as sticks. Cl− ion is shown as dot. Coordination of Cl− is indicated by dashed 673 lines. 674 Extended Data Fig 6. Structural analysis of the Na1, Na2 and chloride binding 675 sites. 676 a-c, Cryo-EM density corresponding to Na1 binding site of hTauTTAU (a), hTauTβ A (b) 677 and hTauTGABA (c). Residues and ligands are shown as sticks. Na+ ion is shown as 678 dot. 679 d-f, Cryo-EM density corresponding to Cl− binding site of hTauTTAU (d), hTauTβ A (e) 680 and hTauTGABA (f). Residues and ligands are shown as sticks. Cl− ion is shown as dot. 681 g-i, Cryo-EM density corresponding to Na2 binding site of hTauTTAU (g), hTauTβ A (h) 682 and hTauTGABA (i). Residues and ligands are shown as sticks. Taurine-like and 683 β -Alanine-like densities are colored in yellow. Na+ densities are colored in purple. 684 j, Superposition of Na1 binding site of hTauTAPO with hTauTTAU. Na+ ion is shown as 685 dot and surrounding residues are shown as sticks. Shift of Cα of F58 is indicated by 686 the arrow. hTauTAPO and hTauTTAU are colored in light peach and pink, respectively. 687 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint k, Superposition of Na2 binding site of hTauTAPO with hTauTGABA. Na+ ion is shown 688 as dot and surrounding residues are shown as sticks. Shift of Cα of G56 and V59 are 689 indicated by the arrow. hTauTAPO and hTauTGABA are colored in light peach and 690 orange, respectively. 691 l, Superposition of Cl binding site of hTauTAPO with hTauTTAU. Cl− ion is shown as 692 dot and surrounding residues are shown as sticks. hTauTAPO and hTauTTAU are 693 colored in light peach and pink, respectively. 694 Extended Data Fig 7. Structure comparison of TauT with related transporters. 695 a, Superposition of hTauTGABA (orange) with GABA-bound GAT1 (grey). Close-up 696 view shows the binding pocket of GABA. GABA and surrounding residues are shown 697 as sticks. Na+ and water molecules are shown as dot. 698 b, Superposition of hTauTGABA (orange) with nipecotic acid-bound GAT1 (grey). 699 Close-up view shows the binding pocket of GABA and nipecotic acid. GABA, 700 nipecotic acid and surrounding residues are shown as sticks. Na+ and water molecules 701 are shown as dot. 702 c, Superposition of hTauTGES (green) with tiagabined-bound GAT1 (grey). Close-up 703 view shows the binding pocket of GES and tiagabine. GES, tiagabine and surrounding 704 residues are shown as sticks. 705 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 5, 2024. ; https://doi.org/10.1101/2024.12.04.626611doi: bioRxiv preprint