Enantiomeric profiles of D- and L-amino acids in plasma and urine of the IRI model

Plasma and urine were collected from the mice after ischemia operation for 4, 8, 20, and 40 h or sham operation (0 h). The concentrations of twenty amino acids were measured by 2D-HPLC and plotted as mean ± SD. (A) Ratio of D-/L-amino acids from the plasma of the IRI model. The graph of Met is shown separately. n = 3. (B) – (D) Plasma amino acid profiles of serine, alanine, and proline from (A) were normalized with 0 h and plotted as ratios of D-/L-enantiomers (B), relative concentrations of D-isomers (C), and relative concentrations of L-isomers (D). (E) Ratio of D-/L-amino acids from the urine of the IRI model. n = 3 – 7. (F) – (H) Urinary amino acid profiles of serine, alanine, and proline from (E) were normalized with those at 0 h and plotted as ratios of D-/L-enantiomers (F), relative concentrations of D-isomers (G), and relative concentrations of L-isomers (H). C/CC: cysteine or cystine.

Molecular Transport in renal BBMVs proteome of the IRI model

(A) Ingenuity Pathway Analysis (IPA) shows the heatmaps of Molecular Transport in the membrane proteome from BBMVs of the IRI model (4h IRI/sham (4h) or 8h IRI/sham (8h)). Transport functions are categorized by types of substrates. Transport of all micromolecules (top) is derived from the combination of inorganic (middle) and organic compounds (bottom). Area and colors represent -log10(p-value) and annotated functions (z-score), respectively. Numbers inside the columns indicate the numbers of corresponding proteins. (B) Heatmaps of membrane transport proteins that mediate the transport of organic compounds. The proteins with ratios of more than 1.1-fold change are shown. The category “others” includes the transport of peptides, organic cations, organic anions, and drugs. Colors indicate log2 fold of 4h IRI/sham (4h) or 8h IRI/sham (8h).

ASCT2 is one of D-serine transporters at the apical membrane of renal proximal tubular epithelia.

(A) – (C) Localization of Asct2 in mouse kidney by immunofluorescence staining. Mouse kidney slides were co-stained with anti-Asct2(NT) antibody (Asct2; green) and protein markers for renal proximal tubule segments: anti-Sglt2 antibody (A: Sglt2, apical membrane marker of S1 + S2 segments), anti-Agt1 antibody (B: Agt1, apical membrane marker of S3 segment), and anti-Na+/K+-ATPase antibody (C: Na+/K+-ATPase, basolateral membrane marker). Merge images are shown in the right panel. Arrowheads indicate co-localization of the proteins. Scale bar = 20 μm. (D) Time course of 100 μM D-[3H]serine transport in wild-type (WT) and ASCT2-knockout (ASCT2-KO) HAP1 cells measured in PBS pH 7.4. Dot plot = mean ± SEM; n = 3. E) Left: Western blot using anti-ASCT2 antibody verified the suppression of ASCT2 in ASCT2-siRNA transfected HEK293 cells. Right: Transport of 100 μM D-[3H]serine (in PBS pH 7.4) was measured in ASCT2-knockdown (ASCT2-siRNA) in comparison to the Mock cells (control). Dot plot = mean ± SEM; n = 3.

Identification of SMCT1 and SMCT2 as candidates of D-serine transporters

(A) Volcano plots of 398 membrane transport proteins identified from the BBMV proteome of the IRI model. The Median of log2 fold of 4h IRI/sham (left) or 8h IRI/sham (right) were plotted against -log10 of p-value. Three proteins with log2 fold more than 4.0 (see values in Table S1) are omitted for the better view. The value of Asct2 (red dot) was set as a cut-off value to select D-serine transporter candidates (both increased (red area) and decreased (blue area) expressions). (B) Cell-growth measurement (XTT assay) of ASCT2-siRNA or without siRNA (control) transfected HEK293 cells treated with D-serine. Data represent percent cell growth compared to the non-treated cells. The graphs were fitted to inhibition kinetics (Dose-response – Inhibition). Dot plot = mean ± SEM; n = 5. (C) – (D) Candidates of D-serine transporters were screened by cell-growth measurement. HEK293 cells were transfected with various cDNA clones, as indicated. After transfection, the cells were treated with either 15 mM (C) or 25 mM (D) D-serine for two days and cell growth was examined. The growth effect by D-serine treatment was normalized with that of no treatment and then calculated as log2 fold change of Mock at the same D-serine concentration. Asc-1 is used as the positive control. Bar graph = mean ± SEM; n = 4; *p < 0.05; **p < 0.01.

Candidate transporters from proteomics of the BBMVs from the IRI model. The list is ordered according to the fold change.

Characterization of SMCT1 and SMCT2 as D-serine transporters using SMCT1- and SMCT2-stably expressing cells

(A) Inhibition effect of ibuprofen on D-serine-induced cell growth. FlpInTR-Mock (Mock), FlpInTR-SMCT1 (SMCT1), or FlpInTR-SMCT2 (SMCT2) cells were treated with D-serine for two days in the presence or absence of 0.5 mM ibuprofen. Cell growth was measured by XTT assay. For comparison, the maximum growth inhibition by 25 mM D-serine treatment was set as 100 % inhibition, and no D-serine treatment was set as 0 % inhibition. The graphs were fitted to inhibition kinetics (Dose-response – Inhibition). Dot plot = mean ± SEM; n = 5. (B) Time course of 100 μM D-[3H]serine uptake in FlpInTR-SMCT1 (SMCT1), FlpInTR-SMCT2 (SMCT2) and Mock cells. D-[3H]Serine transport was measured in PBS pH 7.4. Dot plot = mean ± SEM; n = 4. (C) Time course of 100 μM D-[3H]serine uptake was measured similarly to B), but the cells were subjected to ASCT2 knockdown (siRNA transfection) two days before the assay. Dot plot = mean ± SEM; n = 4. (D) Transport of 20 μM D-[3H]serine by ASCT2-siRNA-transfected FlpInTR-SMCT1 or FlpInTR-SMCT2 stable cell lines were measured in the absence (-) or presence of 5 mM indicated inhibitors. The uptake was measured for 10 min in PBS pH 7.4. Graphs represented the uptake data subtracted from those of Mock cells. Bar graph = mean ± SEM; n = 3. (E) Concentration dependence of D-[3H]serine transport in ASCT2-siRNA-transfected FlpInTR-SMCT1 cells. Uptake of D-[3H]serine (0.5 – 10 mM) was measured for 10 min in PBS pH 7.4. Raw data was shown in Figure supplement 5D. The uptake data in FlpInTR-SMCT1 were subtracted from those of Mock cells and fitted to Michaelis-Menten plot with the apparent Km of 3.39 ± 0.79 mM and Vmax of 18.23 ± 1.73 pmol/μg protein/min. Dot plot = mean ± SEM; n = 3 – 4.

Characterization of SMCT1 as D-serine transporter using SMCT1-reconstituted proteoliposomes (SMCT1-PL)

(A) Left: Stain-free SDS-PAGE gel shows SMCT1 (purified SMCT1) purified from pCMV14-SMCT1-transfected Expi293F cells, reconstituted empty liposomes (Liposome) and SMCT1-reconstituted proteoliposomes (SMCT1-PL). (B) Time course of D-[3H]serine transport in SMCT1-PL. Uptake of 200 μM D-[3H]serine was measured in Na+-containing buffer (Na+) compared to Na+-free buffer (K+). Dot plot = mean ± SEM; n = 3. (C) Ibuprofen effect on the uptake of [3H]lactate, [3H]propionate, L-[3H]serine, and D-[3H]serine in SMCT1-PL. Uptakes of 50 μM radiolabeled substrates were measured for 5 min in Na+-containing buffer (Na+) or Na+-free buffer (K+) in the presence or absence of 1 mM ibuprofen. Bar graph = mean ± SEM; n = 3. (D) Amino acid selectivity of SMCT1-PL. Transport of 50 μM radiolabeled amino acids was measured in SMCT1-PL for 5 min. The substrate uptake in Na+-containing buffer was subtracted from those in Na+-free buffer and calculated as % lactate uptake. Bar graph = mean ± SEM; n = 4.

Characterization of D-serine transporters in BBMVs of normal mice and the IRI model

(A) Transport of 10 μM D-[3H]serine in renal BBMVs isolated from normal mice. The uptake was measured in Na+-free buffer (K+), Na+-containing buffer (Na+), or the presence of 1 mM ibuprofen (Na+ + Ibuprofen). Dot plot = mean ± SEM; n = 3. (B) Transport of 10 μM D- [3H]serine in renal BBMVs isolated from the normal mice was performed similarly to (A) but the BBMVs were preloaded with 4 mM L-Gln prior to the measurement. Dot plot = mean ± SEM; n = 4. (C). Transport of 10 μM D-[3H]serine in renal BBMVs isolated from IRI model. Prior to uptake measurement, the BBMVs were preloaded with 4 mM L-Gln or buffer (no preload). The uptake was measured for 1 min in Na+-free buffer (K+), Na+-containing buffer (Na+), or the presence of 1 mM ibuprofen (Na+ + Ibuprofen). Bar graph = mean ± SEM; n = 3 – 4.

Proposed model of D-serine transport systems in renal proximal tubules

The model summarizes the contributions of ASCT2, SMCT1, and SMCT2 to D-serine transport in renal proximal tubules. Left: the physiological conditions (Normal kidney). SMCT2 at S1 + S2 segments and SMCT1 at segment S3 have high expression levels, whereas the expression of ASCT2 is low. The higher D-serine affinity of SMCT1 compared to SMCT2 supports that the D-serine reabsorption tends to be exceeded at the S3 segment. Right: the pathological conditions (IRI model). expressions of SMCTs decrease while that of ASCT2 increases. ASCT2 expresses ubiquitously and exhibits high D-serine affinity. Blue arrows anticipate total D-serine fluxes in each proximal tubular segment.