Cooperation of Antiporter LAT2/Cd98hc with Uniporter TAT1 for Renal Reabsorption of Neutral Amino Acids

BASIC RESEARCH www.jasn.org

Cooperation of Antiporter LAT2/CD98hc with Uniporter TAT1 for Renal Reabsorption of Neutral Amino Acids

Clara Vilches ,1 Emilia Boiadjieva-Knöpfel,2,3,4 Susanna Bodoy ,5,6 Simone Camargo ,2,3,4 Miguel López de Heredia ,1,7 Esther Prat ,1,7,8 Aida Ormazabal ,7,9 Rafael Artuch ,7,9 Antonio Zorzano ,5,6,10 François Verrey ,2,3,4 Virginia Nunes ,1,7,8 and Manuel Palacín 5,6,7

1Molecular Genetics Laboratory, Genes Disease and Therapy Program, Institut d’Investigació Biomèdica de Bellvitge (IDIBELL), L’Hospitalet de Llobregat, Spain; 2Department of Physiology, 3Zurich Center for Integrative Human Physiology (ZIHP), and 4Swiss National Centre of Competence in Research (NCCR), Kidney Control of Homeostasis (Kidney.CH), University of Zurich, Zurich, Switzerland; 5Department of Biochemistry and Molecular Medicine, Biology Faculty, University of Barcelona, Barcelona, Spain; 6Molecular Medicine Unit, Amino acid transporters and disease group, Institute for Research in Biomedicine (IRB Barcelona), The Barcelona Institute of Science and Technology (BIST), Barcelona, Spain; 7Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERER) – U730, U731, U703, and 10Centro de Investigación Biomédica en Red de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM) – CB07/08/0017, Instituto de Salud Carlos III (ISCIII), Madrid, Spain; 8Genetics Section, Physiological Sciences Department, Health Sciences and Medicine Faculty, University of Barcelona, Barcelona, Spain; and 9Clinical Biochemistry Department, Institut de Recerca Sant Joan de Déu, Hospital Sant Joan de Déu, Esplugues de Llobregat, Spain

ABSTRACT Background Reabsorption of amino acids (AAs) across the renal proximal tubule is crucial for intracellular and whole organism AA homeostasis. Although the luminal transport step is well understood, with several diseases caused by dysregulation of this process, the basolateral transport step is not understood. In humans, only cationic aminoaciduria due to malfunction of the basolateral transporter y+LAT1/CD98hc (SLC7A7/SLC3A2), which mediates the export of cationic AAs, has been described. Thus, the physiologic roles of basolateral transporters of neutral AAs, such as the antiporter LAT2/CD98hc (SLC7A8/SLC3A2), a heterodimer that exports most neutral AAs, and the uniporter TAT1 (SLC16A10), which exports only aromatic AAs, remain unclear. Functional cooperation between TAT1 and LAT2/CD98hc has been suggested by in vitro studies but has not been evaluated in vivo. Methods To study the functional relationship of TAT1 and LAT2/CD98hc in vivo, we generated a double- knockout mouse model lacking TAT1 and LAT2, the catalytic subunit of LAT2/CD98hc (dKO LAT2-TAT1 mice). Results Compared with mice lacking only TAT1 or LAT2, dKO LAT2-TAT1 mice lost larger amounts of aromatic and other neutral AAs in their urine due to a tubular reabsorption defect. Notably, dKO mice also displayed decreased tubular reabsorption of cationic AAs and increased expression of y+LAT1/CD98hc. Conclusions The LAT2/CD98hc and TAT1 transporters functionally cooperate in vivo, and y+LAT1/CD98hc may compensate for the loss of LAT2/CD98hc and TAT1, functioning as a neutral AA exporter at the expense of some urinary loss of cationic AAs. Cooperative and compensatory mechanisms of AA transporters may explain the lack of basolateral neutral aminoacidurias in humans.

J Am Soc Nephrol 29: 1624–1635, 2018. doi: https://doi.org/10.1681/ASN.2017111205

Received November 21, 2017. Accepted February 24, 2018. Biomedicine-Barcelona, Baldiri i Reixac 10, 08028 Barcelona, Spain, or Dr. Virginia Nunes, Institut d’Investigació Biomèdica de C.V., E.B.-K., and S.B. contributed equally to this work. Bellvitge (IDIBELL), Gran Via de L’Hospitalet 199, L’Hospitalet de Llobregat, 08908 Barcelona, Spain, E-mail: manuel.palacin@ F.V., V.N., and M.P. shared senior authorship. irbbarcelona.org or [email protected] Published online ahead of print. Publication date available at www.jasn.org. Copyright © 2018 by the American Society of Nephrology Correspondence: Dr. Manuel Palacín, Institute for Research in

1624 ISSN : 1046-6673/2906-1624 JAmSocNephrol29: 1624–1635, 2018 www.jasn.org BASIC RESEARCH

Renal reabsorption accounts for .98% of recovery of most Significance Statement circulating amino acids (AAs) filtered in the glomerulus. To this end, AAs are actively transported across epithelial cells of Renal amino acid reabsorption is crucial for the maintenance of the renal proximal tubule by AA transporters located in their whole body homeostasis and its impairment leads to several dis- apical and basolateral membranes1 (Figure 1A). Primary in- eases such as cystinuria and Hartnup disorder. Whereas these and otherwell describedaminoacidurias are causedbydefectsof luminal herited aminoacidurias caused by loss-of-function mutations transport proteins, only one aminoaciduria, specifically of cationic demonstrate the role of some transporters in renal reabsorp- amino acids, is associated with the dysfunction of a basolateral tion.2 Thus, mutations in the apical transporters rBAT/b0,+AT transporter. This work demonstrates in vivo the functional co- 3,4 5,6 operation of two basolateral neutral amino acid transporters, LAT2 (SLC3A1/SLC7A9), B0AT1 (SLC6A19), EAAC1 + (SLC1A1),7 and PAT2 (SLC36A2) alone or in combination and TAT1, and shows that another basolateral transporter, y LAT1, can largely compensate for their defect. These findings reveal 8 with mutations in IMINO (SLC6A20) cause cystinuria, synergistic and compensatory reabsorption mechanisms in renal Hartnup disorder, dicarboxylic aminoaciduria, and iminogly- epithelial cells that can explain why no neutral aminoaciduria due cinuria, respectively. Knockout mouse models of these trans- to the defect of basolateral transporters has been identified in porters mimic these human aminoacidurias.9–12 humans. The molecular bases of renal reabsorption of AAs at the basolateral membrane are less well understood. The study of the only METHODS known human aminoaciduria involving a basolateral transporter + demonstrated the role of y LAT1/CD98hc (SLC7A7/SLC3A2)in A full, detailed description of all materials and methods used is lysinuric protein intolerance and in renal reabsorption of cationic provided in the Supplemental Material. AA.13,14 Ablation of y+LAT1 also resulted in cationic aminoacid- 15 uria and large neonatal lethality in mice. Mouse Model Generation, Genotyping, and For neutral AAs, three knockout mouse models of basolateral Experimental Diets transporters have been reported, two of which showed a mild Single loss-of-function mouse models for LAT2 (null knockout)24–26 phenotype in AA renal reabsorption, whereas the third (LAT4 and TAT1 (premature STOP codon at position Y88)17 were crossed 16 [Slc43a2] knockout) was postnatally lethal. Ablation of the aro- to obtain double heterozygous mice and backcrossed to get the matic AA uniporter TAT1 (Slc16a10) caused, next to a substantial F2 generation, including dKO LAT2-TAT1 (dKO) mice. Geno- aromatic aminoaciduria, also a moderate aminoaciduria of other type was confirmed by PCR and/or Sanger sequencing. For 17 neutral AAs that became exacerbated under a protein-rich diet. exacerbation of renal phenotype, a protein-rich diet was used It is interesting to note that the plasma concentration of aromatic (40% casein). As control, mice were fed with a 20% casein diet. AAs was strongly increased despite their urinary loss, presumably C57BL/6J mice were maintained in a 12-hour light/dark cycle, due to the absence of TAT1 from hepatocytes that normally func- with free access to food and water. Experimental diets were 17 tion as a metabolic sink for aromatic AAs. The knockout of maintained for 8 days, with free access to water. Males and LAT2, the catalytic subunit of LAT2/CD98hc (Slc7a8/Slc3a2)het- females were used for initial characterization of the dKO mouse erodimer, showed a very mild aminoaciduria.18 Redundancy and model; only male mice were included in renal function studies. compensatory mechanisms most probably underlie these mild phenotypes of hyperexcretion of AAs in urine. Mouse Sample Collection and Analyses TAT1 and LAT2/CD98hc have been shown to functionally co- Mice were individually housed in metabolic cages for 4 days during 19 operate in a cellular model. TAT1 is a uniporter that mediates which experimental diet was maintained. Twenty-four-hour urine 20,21 downhill transport of L-aromatic AAs (Tyr, Phe, and Trp) and samples, blood plasma, and organs of interest were harvested. AAs LAT2/CD98hc is an obligatory exchanger of any L-neutral AA be- and creatinine were determined in urine and plasma. Kidney total 22,23 side proline. ThecoexpressionofTAT1wasshowntoenable RNA was analyzed by RT-qPCR by using UPL probes (Roche Life- fl glutamine to ef ux via LAT2/CD98hc in exchange with aromatic Science, Switzerland) in a microfluidic chip (Fluidigm, CA). LAT2, fl 19 AAs ef uxed via TAT1 in Xenopus oocytes. TAT1 and LAT2 are TAT1, and y+LAT1 proteins were detected from total membranes expressed in the basolateral membrane of the same epithelial cells in extracted from kidneys as detailed in the Supplemental Material. the kidney proximal tubule.19 Thus, these two transporters might cooperate in renal reabsorption using common substrates (Figure 1A). To study their functional relationship in vivo,adouble-knock- RESULTS out mouse model (dKO LAT2-TAT1) has been generated. Here, we show that dKO LAT2-TAT1 mice presented a synergistic defect of General Features of the Double-Knockout LAT2-TAT1 the tubular reabsorption (TR) of neutral AAs (aromatic and non- Mice aromatic), indicating their cooperation. However, the substantial Single ablation mouse models for LAT2 (null knockout) (KO residual TR of neutral AAs suggested compensation by other trans- LAT2)26,25 and TAT1 (premature STOP codon at position Y88; porters. Cationic aminoaciduria and upregulation of the cationic TAT1 Y88*; KO TAT1)17,24 in pure genetic background and neutral AA antiporter y+LAT1/CD98hc suggest its involvement C57BL/6J were crossed to obtain the full range of possible in compensatory reabsorption of neutral AAs (Figure 1B). genotypes including the double TAT1 and LAT2 knockout

J Am Soc Nephrol 29: 1624–1635, 2018 LAT2 and TAT1 in Renal Reabsorption 1625 BASIC RESEARCH www.jasn.org

Figure 1. Proposed model for renal reabsorption of neutral and cationic AAs in mice, showing cooperation of LAT2/CD98hc and TAT1 transporters. Schematic representation of an epithelial cell from the proximal tubule in (A) wild-type and (B) dKO LAT2-TAT1 mice. Arrow head size indicates favored transport direction, and different shape sizes for AAs are used to indicate corresponding concentrations in tubule lumen, cell, and blood. Data presented in this work support a role of TAT1 and LAT2/CD98hc in the basolateral efflux of aromatic and nonaromatic neutral AAs. Ablation of LAT2 and TAT1 resulted in hyperexcretion of neutral and cationic AAs and upregulation of y+LAT1/CD98hc heterodimers. The increased intracellular content of neutral and aromatic AAs is hypothesized to compete with cationic AAs for transport via y+LAT1/ 4Fh2 that therefore may contribute to the basolateral efflux of aromatic and nonaromatic neutral AAs. It is also proposed that the increased content of neutral and aromatic AAs might block apical B0AT1 and PAT2 transporters. It is speculated that the missing transporter(s) for Pro basolateral efflux might share transport with neutral AAs. Basolateral LAT4 and reversion of basolateral SNAT3 are candidates for efflux of the

1626 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1624–1635, 2018 www.jasn.org BASIC RESEARCH

Figure 2. Confirmation of ablation of LAT2 and TAT1 protein expression in kidney of single and double LAT2 and TAT1 knockout mice. (A) Total membranes from kidney were analyzed by western blot for LAT2 and TAT1 in mice of all genotypes studied (WT and dKO denote wild-type mice and dKO LAT2-TAT1 homozygotes, respectively). Images correspond to a representative sample for each genotype. Arrow heads point at bands of interest (LAT2/CD98hc and TAT1), as indicated. Because gels were run in nonreducing conditions, LAT2 antibody detected LAT2/CD98hc heterodimer. Samples (50 mg of protein) were loaded into all lanes and run in an 8% acrylamide gel. (B) Immunofluorescence signals of LAT2 and TAT1 (green) and nuclei (DAPI in blue) of kidney cortex sections of mice with the same genotypes as in (A). Antibodies against mouse LAT236 and TAT119 were used as indicated in the Supplemental Material. Bar=50 mm. mouse model (dKO LAT2-TAT1). The observed frequencies mice (Figure 2A), as it has been reported for KO TAT1 mice.17 As did not follow the expected Mendelian distribution of geno- expected, the dKO LAT2-TAT1 mice presented no TAT1 and types, with less than half of the expected frequency for dKO LAT2 protein bands in renal total membranes. Immunofluores- LAT2-TAT1 (Supplemental Table 1). cence showed basolateral expression of TAT1 and LAT2 During initial characterization, mice were kept under standard compatible with expression in proximal tubule epithelial diet (14% protein content of vegetal origin). In these conditions, cellsthatislackingindKOLAT2-TAT1mice(Figure2B). dKO LAT2-TAT1 homozygote mice showed lower body weight compared with wild-type mice, this difference being more robust AA Hyperexcretion in Urine of Single and Double LAT2 in males than in females (Supplemental Figure 1). Body weights of and TAT1 Knockout Mice LAT2andTAT1singleKOmalemicewereinbetweenwild-type Urinary AA excretionwas measured in 3-month-old male mice and dKO LAT2-TAT1 weight curves (Supplemental Figure 1A). under normal and protein-rich diet (20% or 40% protein of Lower body weight has been previously reported for an LAT2 animal origin). During this short diet (11 days), the weight of knockout mouse model18 but not for KO TAT1 mice,17 possibly the KOTAT1 and dKO LAT2-TAT1 remained lower than that of due to the higher protein content of the diet (20% casein). The the other genotypes (Supplemental Figure 4) and, as previ- lower body weight of dKO LAT2-TAT1 mice is the consequence ously reported for KO TAT1 mice, water intake and urine flow of a general lower weight of all organs, with the exception of the were increased in all genotypes under protein-rich diet brain, which showed a small but statistically significant higher (Supplemental Figure 5), presumably due to increased relative weight (Supplemental Figure 2). Reduced food intake urea excretion.17 was not the basis of the lower weight of dKO LAT2-TAT1 mice dKO LAT2-TAT1 mice presented urinary hyperexcretion of (Supplemental Figure 3). almost all neutral AAs that was more dramatic under protein- rich diet and was highest for tyrosine (common substrate for Ablation of TAT1 and LAT2 Transporters in the Kidneys TAT1 and LAT2). Surprisingly, the cationic AA lysine and pro- of the Studied Mouse Models line were also hyperexcreted, although they are not substrates Protein bands corresponding to both transporters were not de- of these transporters (Figure 3). AA hyperexcretion was much tectable in renal total membranes of the corresponding single KO lower in the single ablation mouse models (Supplemental

indicated AA substrates during renal reabsorption. ?, unknown transporter; AA+, cationic amino acids; AA0, neutral amino acids; ARO, aromatic amino acids; Collec, collectrin.

J Am Soc Nephrol 29: 1624–1635, 2018 LAT2 and TAT1 in Renal Reabsorption 1627 BASIC RESEARCH www.jasn.org

Figure 3. Double LAT2 and TAT1 knockout mice show urinary hyperexcretion of aromatic, neutral and cationic amino acids. Excretion expressed as nmols of the indicated AA per gram of body weight in 24-hour urine samples of wild-type (WT) and dKO LAT2-TAT1 homozygote mice. Data (mean6SEM) from male mice at 3–4 months of age and after 11 days with the indicated experimental diet. Number of animals analyzed: for 20% protein diet, five WT and five dKO LAT2-TAT1 mice; for 40% protein diet, five WT and 11 dKO LAT2-TAT1 mice. Significant statistical differences (t test) between dKO LAT2-TAT1 and WT values are indicated: *P#0.05; **P#0.01; ***P#0.001. AAs are indicated with the three-letter code.

Table 2). None of the tested models showed hyperexcretion of LAT2) upon TAT1, but not LAT2, ablation, whereas upon ab- the anionic AA glutamate, which is not a substrate of TAT1 and lation of both transporters, excretion increased further (Figure LAT2 (Figure 3, Supplemental Table 2). 5A). The high absolute amount of Tyr excretion in TAT1 KO Plasma AA concentration measurements revealed a strong and dKO mice is also related to its high plasma level due to increase only for tyrosine in dKO LAT2-TAT1, as previously defective hepatic uptake and thus metabolism in the absence reported in KO TAT1 mice, that may be attributed to decreased of TAT1 (see the introduction and Mariotta et al.17). In the case hepatic uptake and thus catabolism (Figure 4 for wild-type and of Phe, the other tested aromatic AA, a similar effect was ob- dKO LAT2-TAT1 mice; all genotypes in Supplemental Table served under protein-rich diet only (Supplemental Table 4). 3).17 In contrast, most of the other neutral AAs and cationic For neutral nonaromatic AAs that are substrates of LAT2, but AAs presented a decreased plasma concentration in dKO not of TAT1 (as serine), RC was slightly, but signiﬁcantly, in- LAT2-TAT1 mice that was more marked under protein-rich creased under protein-rich diet in both single KO mouse mod- diet (Figure 4). To estimate the amount of a given AA excreted els compared with wild-type mice (Figure 5B). In contrast, in in urine relative to the concentration of this AA in the plasma dKO LAT2-TAT1 mice, RC of neutral nonaromatic AAs was we calculated the renal clearance of the different AAs (RC; the further strongly increased, speciﬁcally 2–7-fold under 20% hypothetic volume of plasma from which a substance is com- protein diet and 10–40-fold under 40% diet compared with pletely removed per minute by the kidney; Supplemental Ma- wild-type mice (Supplemental Table 4). This exacerbated ex- terial and Supplemental Table 4). The higher RC values cretion in double KO mice compared with single KO mice was revealed an increased urinary loss, irrespective of the plasma observed for all neutral AA substrates of LAT2 and not of TAT1 concentration, of the aromatic AATyr (substrates of TAT1 and (Supplemental Table 4; see also Figure 5B for serine values).

1628 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1624–1635, 2018 www.jasn.org BASIC RESEARCH

Figure 4. Double LAT2 and TAT1 knockout mice show increased tyrosine but decreased concentration of neutral and cationic amino acids. Plasma concentration (mM) of the indicated AAs (three-letter code) in wild-type (WT) and dKO LAT2-TAT1 homozygote mice. Data (mean6SEM) from male mice at 3–4 months of age and after 11 days with the indicated experimental diet. Number of animals analyzed as in Figure 2. Signiﬁcant statistical differences (t test) between dKO LAT2-TAT1 and WT values are indicated: *P#0.05; **P#0.01; ***P#0.001; ****P,0.001.

These results thus indicate a synergistic interaction of TAT1 confirms that the combined lack of both LAT2 and TAT1 and LAT2/CD98 for the renal reabsorption of nonaromatic leads to a clear defect of neutral and cationic AA reabsorp- neutral AAs. Surprisingly, a similar effect was also observed tion. In view of the substantial remaining neutral AATR, the under protein-rich diet for the cationic AA lysine and the intriguing urinary loss of cationic AAs suggests the hypoth- imino acid proline (Figure 5, C and D). esis that (a) transporter(s) able to accommodate both neutral and cationic AAs might compensate to some extent for AA TR in Single and Double LAT2 and TAT1 Knockout the ablation of LAT2 and TAT1 on the expense of a urinary Mice loss of cationic AAs. AA tubular reabsorption (TR) (Figure 6, Supplemental Table 5) was estimated (Supplemental Material) after determination Expression of Neutral AA Transporters in the Kidneys of the GFR (Supplemental Material), on the basis of the 24- of Single and Double LAT2 and TAT1 Knockout Mice hour urinary flow (Supplemental Figure 5B) and creatinine To detect transporters potentially compensating the defective plasma and urinary concentrations (Supplemental Figure 6, reabsorption activities in our mouse models, the mRNA ex- B–D). The eGFR was similar in all mouse models regardless of pression of neutral AA transporters was determined in kidney. the diet (Supplemental Figure 6A). Therefore, the genotype- The expression of 31 isoforms of 22 AA transporter subunits dependent differences in urinary AA excretion (clearance) dis- was determined with nanofluidic chips. Results are summa- cussed above can be fully attributed to differences in TR and rized in Figure 7 as a heatmap-like table showing in green and thus fractional excretion of AAs (Figures 5 and 6). This red, respectively, the up- and downregulations of all gene

J Am Soc Nephrol 29: 1624–1635, 2018 LAT2 and TAT1 in Renal Reabsorption 1629 BASIC RESEARCH www.jasn.org

Figure 5. Higher increase of RC of tyrosine, serine, lysine and proline in double knockout mice compared to single knockouts. RC (ml/24 hrzg body weight) of mouse homozygotes of all of the indicated genotypes (WT denotes wild type and dKO denotes dKO LAT2- TAT1). Data (mean6SEM)arefrommalemiceat3–4 months of age after 11 days with the indicated experimental diet. Number of animals analyzed: for 20% protein diet, five animals in each group; for 40% protein diet, five WT, 11 dKO, six KO LAT2, and eight KO TAT1. Statistical significance (t test) is indicated as follows: * is used to compare versus WT, # to compare versus dKO, and & to compare versus KO LAT2. Number of symbols indicates P values as follows: one symbol for P#0.05; two for P#0.01; three for P#0.001; and four for P,0.001.

products for each genotype and condition, using samples of SNAT2, ASCT2, and b0,+AT mRNAs were downregulated in wild-type mice under 20% protein-diet as reference. As ex- wild-type mice under protein-rich diet. pected, LAT2 mRNA was absent in the mouse models with The robust upregulation of y+LAT1 mRNA variant 1 in LAT2 ablation and, interestingly, mice with the Y88* TAT1 LAT2-deﬁcient mouse models was further investigated and mutation showed a strong decrease in TAT1 mRNA suggesting no effect on total y+LAT1 mRNA expression (i.e., considering non–sense-mediated decay. In contrast, mRNA expression of all variants together) was observed (Figure 7, Supplemental most other transporters was not affected in single KO mouse Figure 7A). This appears to be due to the very low expression models. of the variant 1 in comparison with the variants 2 and 3 (data Intriguingly, however, y+LAT1 mRNA variant 1 was con- not shown). At the protein level, a trend in increased expres- sistently overexpressed when LAT2 was ablated, irrespective of sion of y+LAT1 was observed under normal protein diet (Fig- the dietary protein content and unlike variants 2 and 3 (Figure ure 8A). In contrast, under protein-rich diet y+LAT1 was 7, Supplemental Figure 7A). In addition, SNAT3 mRNA var- clearly overexpressed in kidneys of dKO and KO LAT2 animals, iants 1 and 3 were upregulated (3–4 fold over wild-type levels) to a lesser extent in KO TAT1 mice, and not in wild-type mice in the kidneys of KO TAT1 mice but only under protein-rich (Figure 8B). These results indicate a complex regulation of y+ diet (Figure 7, Supplemental Figure 7B). On the contrary, LAT1 and suggest that upregulation of y+LAT1 in these models SNAT2 mRNA and SNAT3 mRNA (variant 3) were downre- could work as a compensatory mechanism to the loss-of-function gulated in KO LAT2 and dKO LAT2-TAT1 mice. Similarly, of LAT2 and TAT1 when mice are fed a protein-rich diet.

1630 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1624–1635, 2018 www.jasn.org BASIC RESEARCH

LAT2 participate in renal reabsorption of aromatic AAs, (2) TAT1 and LAT2 have a synergistic functional role in the reabsorption of nonaromatic neutral AAs, and (3) other basolateral transporters compensate for the lack of TAT1 and LAT2 in the renal reabsorption of aromatic and nonaromatic neutral AAs (Figure 1B). Defective TR of some neutral AAs in the single LAT2 KO mice stressed with protein-rich diet (Figure 6) is the first demonstration of the role of LAT2/CD98hc in AA renal reabsorption. Urine hyperexcretion of neutral AAs in LAT2 ablated mice under normal diet, previously reported18 and confirmed here (Figure 3), is only caused by the increased concentration of these AAs in plasma with no effect on TR. The stronger defect in TR of Tyr (substrate for TAT1 and LAT2) in TAT1 KO mice compared with LAT2 KO mice supports a higher contribution of TAT1 in the renal reabsorption of this AA. The synergistic effect of simultaneous TAT1 and LAT2 ablation on TR of aromatic Figure 6. TR of amino acids is more affected in double LAT2 and TAT1 knockout than AAs (Tyr and Phe) strongly suggests that in single knockout mice. The percentage (mean) of TR, shown inside each cell of the LAT2/CD98hc compensates the ablation table, was estimated for the indicated AAs (three-letter code) in homozygous mice for of TAT1. the indicated genotypes for (A) 20% protein content in the diet and (B) 40% protein Functional coupling of uniporter TAT1 content in the diet. WT denotes wild type and dKO denotes dKO LAT2-TAT1 ho- and antiporter LAT2/CD98hc for the baso- – mozygotes. Male mice at 3 4 months of age after 11 days with the indicated exper- lateral efflux of a non-TAT1 substrate was fi imental diet were studied. The number of mice is ve for all groups under 20% protein previously reported in Xenopus oocytes ex- diet, and under 40% protein diet as follows: WT, four; KO LAT2, five;KOTAT1,five; pressing these two human transporters.19 dKO, six. For each particular AA, a three-color scale is used on the basis of both statistically significant differences from the group in the preceding column (after t test Both transporters are expressed in the ba- analysis of data) and % of TR, with green as higher % and red as lower TR. * and & solateral membrane of the murine renal indicate statistical differences compared with WT and KO LAT2, respectively. A proximal tubule (Figure 2A), in agreement complete version of these data, with mean6SEM of TR and statistical significance, is with a previous report,19 and their func- shown in Supplemental Table 5. tional coupling could thus affect TR of neutral AAs. Supporting this hypothesis, In an attempt to corroborate our hypothesis, we generated two namely that the efflux of nonaromatic neutral AAs via LAT2 new double-knockout mouse models by crossing a newly gener- may, to some extent, be driven by aromatic AAs recycling via ated tamoxifen-inducible (CRE recombinase under ubiquitin C TAT1, their TR was synergistically decreased in the double promoter) y+LAT1 knockout mouse (S. Bodoy and M. Palacín, compared with the single LAT2 and TAT1 knockout mice (Fig- unpublished observations) with the KO LAT2 and the KO TAT1 ure 6). However, the fact that the lack of TAT1 alone had much mice (data not shown). Unfortunately, the high mortality rate less effect on nonaromatic neutral AA reabsorption than the (approximately 60% within 2 weeks after induction with tamox- double KO shows that the functional coupling of the LAT2 ifen) and the deficient health state of surviving animals (dramatic exchanger is not restricted to TAT1, but probably extends to decrease in body weight and GFR) precluded an accurate as- (an)other basolateral transporter(s), for instance to LAT4 sessment of AA reabsorption in these new double-knockout (Slc43a2) which is also a uniporter transporting some essential mice (y+LAT1-LAT2 and y+LAT1-TAT1). AAs.16,27 Additionally, the fact that the efflux of these nonaromatic neutral AAs that are not substrates of TAT120 (and mostly also not of LAT4) was much less reduced in LAT2 single DISCUSSION KO than in double KO mice indicates that (an)other transporter(s) partially compensate(s) for the lack of efflux via AATR in the single and the newly generated double LAT2 and LAT2/CD98hc (Figure 1B). Indeed, .60% of the TR of neu- TAT1 knockout mice studied here support that: (1)TAT1and tral and aromatic AAs remains in dKO LAT2-TAT1 mice even

J Am Soc Nephrol 29: 1624–1635, 2018 LAT2 and TAT1 in Renal Reabsorption 1631 BASIC RESEARCH www.jasn.org

Figure 7. Upregulation of Slc7a7 mRNA upon ablation of LAT2 in kidney. AA transport system and protein and gene names are indicated. The expression of particular mRNA variants (1, 2, or 3) or the complete set of variants detected together (C) was analyzed in homozygous male mice for the indicated genotypes. When not listed, variants have not been described for that transporter gene, with the exception of CD98hc mRNA variants that have not been analyzed. Green/red table cells indicate gene up-/downregulation, respectively, in comparison with the control condition (wild-type mice under 20% protein diet). Gray table cells indicate no signiﬁcant changes in gene expression. Expression results of the ablated transporters, TAT1 and LAT2, are shown separately at the bottom of the

1632 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1624–1635, 2018 www.jasn.org BASIC RESEARCH

the influx of neutral AAs. Indeed, loss-of- function mutations in y+LAT1 (SLC7A7) cause lysinuric protein intolerance charac- terized by hyperexcretion of cationic AAs.31 Interestingly, under the stress of protein-rich diet, a small but significant defect in the TR of cationic AAs was observed in the single LAT2 and TAT1 KO mice that was further increased in the dKO LAT2- TAT1 mice (Figure 6), suggesting that intracellular accumulated neutral AAs compete for efflux with cationic AAs. The only renal basolateral transporter known to share neutral and cationic AA substrates is y+LAT1/CD98hc.8 Moreover, y+LAT1 protein was upregulated in the single and double + Figure 8. y LAT1 protein expression in kidney is upregulated upon ablation of LAT2 LAT2 and TAT1 KO mice under protein-rich fi and TAT1 in mice under high protein diet. Expression was quanti ed in kidneys from diet (Figures 7 and 8). Reversion of y+LAT1/ homozygous male mice for the indicated genotypes and under diet with (A) 20% or (B) CD98hc transport activity (i.e.,efflux of neu- 40% protein content. y+LAT1 protein expression relative to actin as reference protein tral AAs plus sodium instead of cationic AA (upper panels) and a representative image of the immunoblots (lower panels) are fl shown for all genotypes in (A) and (B). Kidney total membranes (50 mgprotein)were ef ux) could be accomplished if the ratio of run in 10% acrylamide gels in nonreducing conditions. In these conditions, y+LAT1 intracellular neutral/cationic AAs increases to band corresponds to y+LAT1/CD98hc heterodimer. Data (mean6SEM) are from 5–6 compensate the chemical gradient of sodium + mice per group at 3–4 months of age and after 11 days with the indicated experi- (Figure 1B). Indeed, y LAT1/CD98hc has mental diet. Statistical significance (t tests) is shown with * to indicate differences been proposed as the main Arg influx medi- compared with wild-type and # to indicate differences versus same genotype in 20% ator in human alveolar macrophages, which protein diet (one symbol P#0.05; two symbols P#0.01; three symbols P#0.001; and show little expression of cationic AA uniport- four symbols P,0.001). WT, wild-type. ers, e.g.,CAT1(SLC7A1)andCAT2b (SLC7A2).32 Unfortunately, double ablation in the more stressing condition of protein-rich diet (Figure 6). At of y+LAT1 (induced after weaning) and LAT2 or TAT1 led to first sight, it is intriguing that under these high-protein diet con- high mortality and surviving animals showed defective GFR ditions no increase in plasma AA concentration was observed, that (data not shown), hindering their use to assess the expected com- would explain the strongly increased aminoaciduria. This discrep- pensation in renal reabsorption of neutral AAs. ancy is, however, due to the fact that urine was collected over 24 A noticeable surprise in this study is the defect of the TR of hours, whereas blood was collected only once during the postab- proline (Pro) in the single and double LAT2 and TAT1 mice under sorptive phase, namely a few hours after the switch to the inactive protein-rich diet (Figure 6). Pro is not a substrate for LAT2 or light phase. Thus, because plasma AA levels are transiently higher TAT1,20,22,23 which suggests that the intracellular accumulation of during the nocturnal absorptive phase (data not shown), it is neutral AAs blocks reabsorption of Pro at the apical membrane or proposed that it is the stronger transient blood AA increase (ex- efflux at the basolateral membrane (Figure 1B). The study of cursion) taking place under high-protein diet that causes a mas- iminoglycinuria has defined the participation of two main trans- 2 sively increased AA spillover and thus a high aminoaciduria. porters on the apical reabsorption of Pro, the Na+ and Cl co- Our data suggest that the antiporter y+LAT1/CD98hc com- transporter of Pro Imino (SLC6A20), and the H+ cotransporter of pensates, to some extent, the absence of LAT2 and TAT1 for TR Pro and glycine (Gly) PAT2 (SLC36A2).8,33 The intracellular ac- of neutral AAs (Figure 1B). This antiporter mediates the elec- cumulationofGlyinTAT1KOanddKOLAT2-TAT1micemight troneutral exchange of cationic AAs with neutral AAs plus blockorreversetheuptakeofProvia PAT2 (Figure 1B). To our sodium with 1:1:1 stoichiometry.28–30 Thus, the inward chem- knowledge, it is not known which basolateral transporters medi- ical gradient of sodium favors the efflux of cationic AAs and ate the efflux of Pro in renal reabsorption (Figure 1).2,8 The fact

table, conﬁrming the loss-of-function of each model. ACTB was used as a reference gene (see the Methods section). Data are from 5–8male mice per experimental group at 3–4 months of age and after 11 days with the indicated experimental diet. Statistical signiﬁcance of expression ratios is indicated according to Boostratio,37 as follows: ~ for P#0.1 (trend); *P#0.05; **P#0.01; ***P#0.001; and ****P,0.001. No expression was detected for: Slc38a1 isoform 3, Slc43a1 isoform 3, Slc6a15 common design for all variants, and Slc7a9 isoform 2. WT denotes wild type and dKO denotes dKO LAT2-TAT1 homozygotes.

J Am Soc Nephrol 29: 1624–1635, 2018 LAT2 and TAT1 in Renal Reabsorption 1633 BASIC RESEARCH www.jasn.org that TR of Pro is affected in LAT2 KO, where Gly TR is not 3. Calonge MJ, Gasparini P, Chillarón J, Chillón M, Gallucci M, Rousaud F, affected, opens the possibility that the missing basolateral Pro et al.: Cystinuria caused by mutations in rBAT, a gene involved in the – transporter might share transport with neutral AAs. transport of cystine. Nat Genet 6: 420 425, 1994 4. Feliubadaló L, Font M, Purroy J, Rousaud F, Estivill X, Nunes V, et al.; This work supports a functional cooperation between TAT1 International Cystinuria Consortium: Non-type I cystinuria caused by and LAT2/CD98hc for the renal reabsorption of neutral AAs and mutations in SLC7A9, encoding a subunit (bo,+AT) of rBAT. Nat Genet additionally suggests a compensation by y+LAT1/CD98 in case of 23: 52–57, 1999 their defect. The large percentage of TR remaining after knocking 5. Kleta R, Romeo E, Ristic Z, Ohura T, Stuart C, Arcos-Burgos M, et al.: out LAT2 and TAT1 is at odds with the larger effect of loss-of- Mutations in SLC6A19, encoding B0AT1, cause Hartnup disorder. Nat Genet 36: 999–1002, 2004 function of the apical neutral AA transporters B0AT1, PAT2, and 6. Seow HF, Bröer S, Bröer A, Bailey CG, Potter SJ, Cavanaugh JA, et al.: Imino.8,34 Compensations by other transporters for the basolat- Hartnup disorder is caused by mutations in the gene encoding the eral efflux of neutral AAs most probably explain why no neutral neutral amino acid transporter SLC6A19. Nat Genet 36: 1003–1007, aminoaciduria due to the defect of a basolateral transporter has 2004 been uncovered. SNAT3, the Na+ cotransporter and H+ antiporter 7. Bailey CG, Ryan RM, Thoeng AD, Ng C, King K, Vanslambrouck JM, et al.: Loss-of-function mutations in the glutamate transporter SLC1A1 cause of the neutral AAs Gln, Asn, and His, and LAT4, the uniporter for human dicarboxylic aminoaciduria. J Clin Invest 121: 446–453, 2011 neutral branched-chain AAs, Met and Phe, expressed in the prox- 8. Bröer S, Bailey CG, Kowalczuk S, Ng C, Vanslambrouck JM, Rodgers H, imal tubule,16,27,35 mightalsobeatthebasisofthiscompensation et al.: Iminoglycinuria and hyperglycinuria are discrete human pheno- (Figure 1B). Kidney proximal tubule–specificablationofthese types resulting from complex mutations in proline and glycine trans- – basolateral AA transporters will be necessary to fully understand porters. JClinInvest118: 3881 3892, 2008 9. Peghini P, Janzen J, Stoffel W: Glutamate transporter EAAC-1-deficient the molecular basis of neutral AA reabsorption. mice develop dicarboxylic aminoaciduria and behavioral abnormalities but no neurodegeneration. EMBO J 16: 3822–3832, 1997 10. Feliubadaló L, Arbonés ML, Mañas S, Chillarón J, Visa J, Rodés M, et al.: Slc7a9-deficient mice develop cystinuria non-I and cystine urolithiasis. ACKNOWLEDGMENTS Hum Mol Genet 12: 2097–2108, 2003 11. Peters T, Thaete C, Wolf S, Popp A, Sedlmeier R, Grosse J, et al.: A – Wewant to thank Laura Gónzalez for assistance with handling mice. mouse model for cystinuria type I. Hum Mol Genet 12: 2109 2120, 2003 The Department of Clinical Biochemistry of Sant Joan de Déu 12. Jiang Y, Rose AJ, Sijmonsma TP, Bröer A, Pfenninger A, Herzig S, et al.: Hospital is part of the Centro de Investigación Biomédica en Red de Mice lacking neutral amino acid transporter B(0)AT1 (Slc6a19) have Enfermedades Raras-Instituto de Salud Carlos III (CIBERER-ISCIII) elevated levels of FGF21 and GLP-1 and improved glycaemic control. and ‘Centre Daniel Bravo de Diagnòstic i Recerca en Malalties Mol Metab 4: 406–417, 2015 Minoritàries’. 13. Borsani G, Bassi MT, Sperandeo MP, De Grandi A, Buoninconti A, Financial support of the research: Spanish Ministry of Science and Riboni M, et al.: SLC7A7, encoding a putative permease-related protein, is mutated in patients with lysinuric protein intolerance. Nat Genet Innovation SAF2015-64869-R-FEDER (to M.P.); Spanish Health 21: 297–301, 1999 Institute Carlos III Grant Fondo de Investigación en Salud (FIS) PI13/ 14. Torrents D, Mykkänen J, Pineda M, Feliubadaló L, Estévez R, de Cid R, 00121-R-FEDER and PI16/00267-R-FEDER (to V.N.); Generalitat de et al.: Identification of SLC7A7, encoding y+LAT-1, as the lysinuric Catalunya Grants SGR2009-1490 (to V.N.), SGR2009-1355 (to M.P.); protein intolerance gene. Nat Genet 21: 293–296, 1999 Swiss National Science Foundation grant 310030_166430 (to F.V.). 15. Sperandeo MP, Annunziata P, Bozzato A, Piccolo P, Maiuri L, D’Armiento M, et al.: Slc7a7 disruption causes fetal growth retardation Double knock-out (dKO) generation: C.V., E.B.-K., E.P., by downregulating Igf1 in the mouse model of lysinuric protein in- andS.B.Mousemanagement:E.P.andC.V.QuantitativeReverse tolerance. Am J Physiol Cell Physiol 293: C191–C198, 2007 transcription-polymerasechainreaction(RT-qPCR)designand 16. Guetg A, Mariotta L, Bock L, Herzog B, Fingerhut R, Camargo SMR, analysis: C.V. and M.L.d.H. Western blot (WB): E.B.-K., M.L.d.H., et al.: Essential amino acid transporter Lat4 (Slc43a2) is required for S.B., and C.V. Immunofluorescence: E.B.-K. Amino acid and cre- mouse development. J Physiol 593: 1273–1289, 2015 atinine analysis: A.O. and R.A. Renal function data analysis: C.V. and 17. Mariotta L, Ramadan T, Singer D, Guetg A, Herzog B, Stoeger C, et al.: T- type amino acid transporter TAT1 (Slc16a10) is essential for extracellular S.B. Study design: M.P., F.V.,V.N., and A.Z. Manuscript draft writing: aromatic amino acid homeostasis control. JPhysiol590: 6413–6424, 2012 M.P., C.V., V.N., and F.V. Manuscript review: all authors. 18. Braun D, Wirth EK, Wohlgemuth F, Reix N, Klein MO, Grüters A, et al.: Aminoaciduria, but normal thyroid hormone levels and signalling, in mice lacking the amino acid and thyroid hormone transporter Slc7a8. DISCLOSURES Biochem J 439: 249–255, 2011 None. 19. Ramadan T, Camargo SMR, Herzog B, Bordin M, Pos KM, Verrey F: Recycling of aromatic amino acids via TAT1 allows efflux of neutral amino acids via LAT2-4F2hc exchanger. Pflugers Arch 454: 507–516, 2007 REFERENCES 20. Kim DK, Kanai Y, Chairoungdua A, Matsuo H, Cha SH, Endou H: Ex- pression cloning of a Na+-independent aromatic amino acid transporter with structural similarity to H+/monocarboxylate transporters. J 1. Makrides V, Camargo SM, Verrey F: Transport of amino acids in the Biol Chem 276: 17221–17228, 2001 kidney. Compr Physiol 4: 367–403, 2014 21. Ramadan T, Camargo SMR, Summa V, Hunziker P, Chesnov S, Pos KM, 2. Bröer S, Palacín M: The role of amino acid transporters in inherited and et al.: Basolateral aromatic amino acid transporter TAT1 (Slc16a10) acquired diseases. Biochem J 436: 193–211, 2011 functions as an efflux pathway. JCellPhysiol206: 771–779, 2006

1634 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1624–1635, 2018 www.jasn.org BASIC RESEARCH

22. Pineda M, Fernández E, Torrents D, Estévez R, López C, Camps M, et al.: 30. Kanai Y, Fukasawa Y, Cha SH, Segawa H, Chairoungdua A, Do Kyung K, Identification of a membrane protein, LAT-2, that Co-expresses with 4F2 et al.: Transport properties of a system y L neutral and basic amino acid heavy chain, an L-type amino acid transport activity with broad specificity for transporter. JBiolChem275: 20787–20793, 2000 small and large zwitterionic amino acids. JBiolChem274: 19738–19744, 1999 31. Ogier de Baulny H, Schiff M, Dionisi-Vici C: Lysinuric protein in- 23. Rossier G, Meier C, Bauch C, Summa V, Sordat B, Verrey F, et al.: tolerance (LPI): A multi organ disease by far more complex than a classic LAT2, a new basolateral 4F2hc/CD98-associated amino acid trans- urea cycle disorder. Mol Genet Metab 106: 12–17, 2012 porter of kidney and intestine. JBiolChem274: 34948–34954, 1999 32. Rotoli BM, Dall’asta V, Barilli A, D’Ippolito R, Tipa A, Olivieri D, et al.: Alveolar 24. Espino Guarch M, Font-Llitjós M, Murillo-Cuesta S, Errasti-Murugarren macrophages from normal subjects lack the NOS-related system y+ for ar- E, Celaya AM, Girotto G, et al.: Mutations in L-type amino acid trans- ginine transport. Am J Respir Cell Mol Biol 37: 105–112, 2007 porter-2 supportSLC7A8as a novel gene involved in age-related 33. Vanslambrouck JM, Bröer A, Thavyogarajah T, Holst J, Bailey CG, Bröer S, hearing loss. eLife 7: e31511, 2018 et al.: Renal imino acid and glycine transport system ontogeny and in- 25. Núñez B, Martínez de Mena R, Obregon MJ, Font-Llitjós M, Nunes V, Palacín volvement in developmental iminoglycinuria. Biochem J 428: 397–407, 2010 fi M, et al.: Cerebral cortex hyperthyroidism of newborn mct8-de cient mice 34. Böhmer C, Bröer A, Munzinger M, Kowalczuk S, Rasko JEJ, Lang F, transiently suppressed by lat2 inactivation. PLoS One 9: e96915, 2014 et al.: Characterization of mouse amino acid transporter B0AT1 26. Font-Llitjós M: Renal reabsoprtion of amino acids: SLC7A9 mutations (slc6a19). Biochem J 389: 745–751, 2005 analysis, the type B cystinuria gene and characterization of a knockout 35. Bröer S: Amino acid transport across mammalian intestinal and renal murin model Slc7a8, Barcelona, Spain, University of Barcelona, 2005. epithelia. Physiol Rev 88: 249–286, 2008 Available at http://hdl.handle.net/10803/977. Accessed April 25, 2013 36. Fernández E, Jiménez-Vidal M, Calvo M, Zorzano A, Tebar F, Palacín M, 27. Bodoy S, Martín L, Zorzano A, Palacín M, Estévez R, Bertran J: Identi- et al.: The structural and functional units of heteromeric amino acid trans- fication of LAT4, a novel amino acid transporter with system L activity. J porters. The heavy subunit rBAT dictates oligomerization of the heteromeric Biol Chem 280: 12002–12011, 2005 amino acid transporters. J Biol Chem 281: 26552–26561, 2006 28. Torrents D, Estévez R, Pineda M, Fernández E, Lloberas J, Shi YB, et al.: 37. Clèries R, Galvez J, Espino M, Ribes J, Nunes V, de Heredia ML: BootstRatio: Identification and characterization of a membrane protein (y+L amino A web-based statistical analysis of fold-change in qPCR and RT-qPCR data acid transporter-1) that associates with 4F2hc to encode the amino acid using resampling methods. Comput Biol Med 42: 438–445, 2012 transport activity y+L. A candidate gene for lysinuric protein intolerance. J Biol Chem 273: 32437–32445, 1998 29. Pfeiffer R, Rossier G, Spindler B, Meier C, Kühn L, Verrey F: Amino acid transport of y+L-type by heterodimers of 4F2hc/CD98 and members of the glycoprotein-associated amino acid transporter family. EMBO J 18: This article contains supplemental material online at http://jasn.asnjournals. 49–57, 1999 org/lookup/suppl/doi:10.1681/ASN.2017111205/-/DCSupplemental.

J Am Soc Nephrol 29: 1624–1635, 2018 LAT2 and TAT1 in Renal Reabsorption 1635