y+LAT1 and y+LAT2 contribution to arginine uptake in different human cell models: Implications in the pathophysiology of Lysinuric Protein Intolerance

Abstract y+LAT1 (encoded by SLC7A7), together with y+LAT2 (encoded by SLC7A6), is the alternative light subunits composing the heterodimeric transport system y+L for cationic and neutral amino acids. SLC7A7 mutations cause lysinuric protein intolerance (LPI), an inherited multisystem disease characterized by low plasma levels of arginine and lysine, protein‐rich food intolerance, failure to thrive, hepatosplenomegaly, osteoporosis, lung involvement, kidney failure, haematologic and immunological disorders. The reason for the heterogeneity of LPI symptoms is thus far only poorly understood. Here, we aimed to quantitatively compare the expression of SLC7A7 and SLC7A6 among different human cell types and evaluate y+LAT1 and y+LAT2 contribution to arginine transport. We demonstrate that system y+L‐mediated arginine transport is mainly accounted for by y+LAT1 in monocyte‐derived macrophages (MDM) and y+LAT2 in fibroblasts. The kinetic analysis of arginine transport indicates that y+LAT1 and y+LAT2 share a comparable affinity for the substrate. Differences have been highlighted in the expression of SLC7A6 and SLC7A7 mRNA among different cell models: while SLC7A6 is almost equally expressed, SLC7A7 is particularly abundant in MDM, intestinal Caco‐2 cells and human renal proximal tubular epithelial cells (HRPTEpC). The characterization of arginine uptake demonstrates that system y+L is operative in renal cells and in Caco‐2 where, at the basolateral side, it mediates arginine efflux in exchange with leucine plus sodium. These findings explain the defective absorption/reabsorption of arginine in LPI. Moreover, y+LAT1 is the prevailing transporter in MDM sustaining a pivotal role in the pathogenesis of immunological complications associated with the disease.


| INTRODUC TI ON
Lysinuric protein intolerance (LPI; MIM 222700) is a recessively inherited autosomal disease caused by mutations of SLC7A7 (solute carrier family 7A member 7; MIM #603593), coding for y+LAT1 protein, that is part of the cationic amino acid (CAA; arginine, lysine and ornithine) transport system y + L. 1,2 Patients manifest with low plasma levels of arginine and lysine and typically display protein-rich food intolerance and secondary urea cycle disorders; other symptoms are heterogeneous, and include failure to thrive, recurrent vomiting, hepatosplenomegaly, osteoporosis, lung involvement, kidney failure, haematologic abnormalities and immunological disorders. 3 y+LAT1, together with y+LAT2 (encoded by SLC7A6), is one of the alternative light subunits composing the heterodimeric transport system y + L; the heavy chain is, instead, the glycoprotein 4F2hc (encoded by SLC3A2), that is required for the proper expression of the transporter on the plasma membrane. 4 This transport system selectively couples the sodium-independent efflux of CAA to the influx of neutral amino acids (leucine, glutamine) and sodium, working as an antiport. 5 The activity of system y + L has been initially described in erythrocytes, then it has been identified in placenta, platelets, skin fibroblasts, hepatocytes, small intestine and kidney; in polarized epithelia, it is mainly located onto the basolateral cell membrane, where it mediates the transport of cationic amino acids from the renal and intestinal epithelia towards the bloodstream. 6 Previous studies performed by our group have shown that monocytes/macrophages express high levels of SLC7A7 and, consistently, that arginine transport by system y + L is definitely impaired in mononuclear cells isolated from LPI patients. 7 On the contrary, other LPI cells, such as fibroblasts and erythrocytes, 8,9 do not display any defect in arginine transport, suggesting that the alternative transporter y+LAT2 can compensate the defect of y+LAT1 in these models.
In light of these findings and of clinical evidence, 10 the most accredited pathogenetic hypothesis for LPI symptoms is that circulating cells as monocytes/macrophages, that actually manifest the transport defect, are responsible for the multiorgan complications of the disease and, in particular, for pulmonary and immunological ones. However, since little is known about the relative contribution of SLC7A7/y+LAT1 and SLC7A6/y+LAT2 to arginine transport in other tissues, in this study, we aimed to compare their actual expression and activity among different cell types, paying particular interest to intestinal and renal cells, targets of LPI disease.

| Cell cultures
Human monocytes were isolated from buffy coats of normal, healthy donors, provided by the Unit of Immunohematology and Transfusion of the Azienda Ospedaliero-Universitaria of Parma (local ethics committee approval # 43899, 03/12/2015), as previously described. 11 Monocyte-derived macrophages (MDM) were obtained by incubating monocytes for 6 d in RPMI1640 supplemented with 10% endotoxin-free foetal bovine serum (FBS) and 50 ng/mL of recombinant human GM-CSF. Normal human fibroblasts, obtained from a 15year-old healthy donor, 12 were cultured in high glucose Dulbecco's modified Eagle's medium (DMEM) with 4 mmol/L glutamine. Human renal proximal tubular epithelial cells (HRPTEpC), purchased from Sigma-Aldrich (Italy), were grown in REGM medium, as indicated by the manufacturer; only cells between first and third passages were used. Caco-2 intestinal epithelial cells were obtained from American Type Culture Collection (ATCC) and routinely grown in DMEM.
About 200 μL and 700 µL culture media were added to the apical and basolateral compartments of the insert, respectively. Cell cultures were employed 7 days after seeding when monolayers exhib-

| RT-qPCR analysis
Total RNA was isolated with GeneJET RNA Purification Kit and reverse transcribed with RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific); the amount of the genes of interest (SLC7A6/y+LAT2 and SLC7A7/y+LAT1) was, then, determined using TaqMan® Gene Expression Assays by Thermo Fisher Scientific (Cat# Hs00187757_m1, Hs00909952_m1, and Hs03855120_g1, respectively), and expressed as numbers of mRNA molecules upon normalization to that of the housekeeping gene (RPL15, Ribosomal Protein Like 15). 13

| L-Arginine uptake
In mammalian cells, four distinct transport mechanisms, namely systems y + , y + L, b 0,+ and B 0,+ , mediate cationic amino acid uptake. 5 System y + operates a sodium-independent and membrane potential-dependent unidirectional transport specific for cationic amino acids, while the other three transporters accept both cationic and neutral amino acids. 14 ATB 0,+ is a highly concentrative system, being energized by the transmembrane gradients of Na + and Cl − , as well as by membrane potential. System b 0,+ is Na + -independent and me- L-arginine uptake was normalized for protein content, determined directly in each well by using a modified Lowry method. 15 Uptake is expressed as nmol/mg of protein/min. The activity of system b 0,+ was calculated as the difference between arginine transport in the absence of inhibitors (total) and the transport measured in the presence of leucine in the absence of sodium. System y + L was calculated as the difference between arginine transport in the absence of inhibitors (total) and the transport measured in the presence of leucine in the presence of sodium, after the subtraction of b 0,+ quote, when present. System y + activity was determined as the quote of transport further inhibited by lysine in the presence of sodium.

The apparent kinetic parameters K m (Michaelis constant) and
V max (maximum transport rate) of arginine uptake were calculated by non-linear regression fitting according to the Michaelis-Menten equations: for a single saturable component, where v is the initial influx, V max is the maximal influx and K m is the Michaelis constant.

| Statistics
Prism ® 5.0 GraphPad software has been employed for the statistical analysis; statistical significance was calculated with Student's t test for unpaired data, unless stated otherwise. Differences were considered significant when P < .05.

| Biochemical features of y+LAT1 and y+LAT2 transporters
In a previous contribution, we showed that the activity of system y + L was impaired in LPI monocytes and alveolar macrophages, but readily detectable in fibroblast-like mesenchymal cells obtained from an LPI patient, as well as in fibroblasts from healthy donors. 7 Based on the analysis of gene expression, we there suggested that a normal expression of SLC7A6/y+LAT2 could compensate the mutation of SLC7A7/y+LAT1 in LPI fibroblasts conferring them normal arginine uptake, whereas the predominant expression of y+LAT1 in monocytes and macrophages could explain the defective transport observed in LPI cells.
In order to verify this hypothesis, we here first tried to discriminate the relative contribution of the two transporters in fibro- moreover, since leucine exerted its inhibitory effect in SLC7A7-, but not in SLC7A6-silenced cells (panel F), we can conclude that system y + L-mediated arginine transport in these cells is likely accounted for by y+LAT2 protein.
Starting from these findings, we next addressed the biochemical features of the transporters by performing the kinetic analysis of arginine transport in silenced cells; more precisely, we measured the kinetic constants of system y + L (calculated as the difference between total transport and the transport measured in the present of leucine) in MDM silenced for SLC7A6 (hence measuring the activity of the sole y+LAT1) and in fibroblasts silenced for SLC7A7 (hence measuring the activity of the sole y+LAT2). Results, presented in Figure 3

| Tissue-specific pattern of expression of SLC7A6/y+LAT2 and SLC7A7/y+LAT1
Next, we addressed the differential expression of SLC7A6/y+LAT2 and SLC7A7/y+LAT1 in different tissues through a quantitative RT-qPCR analysis of the two genes in various cell models. As shown in

| Arginine transport in renal and intestinal epithelial cells
Data of gene expression are consistent with the notion that beside immune cells, intestine and kidney are the main targets of LPI defect; however, little is thus far known about system y + L activity in human renal and intestinal cells. Figure 5 shows the characterization of arginine transport in primary renal proximal tubular epithelial cells (HRPTEpC) and in Caco-2 cells, the representative in vitro model of the intestinal barrier. The relative contribution of each transporter has been assessed through a methodology previously used in other cell models. 16 Arginine uptake was comparable in the absence and in the presence of sodium in both renal (panel A) and intestinal (panel C) cells, thus excluding a significant contribution of sodium-dependent transport systems, such as B 0,+ . In the absence of sodium, leucine significantly inhibited arginine uptake, demonstrating that system b 0,+ is operative in these cells. Moreover, since the inhibitory effect of leucine was much more evident in the presence of the cation, we can conclude that also system y + L, typically inhibited by In these latter, the transcellular flux of arginine has been also monitored in polarized monolayers; to this end, cells grown on inserts were incubated, at the apical side, in the presence of labelled F I G U R E 2 Effect of SLC7A6 and SLC7A7 silencing on arginine transport in monocyte-derived macrophages (MDM) and fibroblasts. MDM and fibroblasts were transfected with scrambled or with SLC7A6 or SLC7A7 siRNA, as described in Methods. Panels A and D: the expression of both SLC7A6 and SLC7A7 in silenced cells was referred to that of scrambled siRNA (=1). Bars are mean ± SEM of three independent experiments, each performed in duplicate. **P < .01, ***P < .001 vs scrambled siRNA calculated with one sample t test. Panels B and E: arginine transport was measured through a 1-min incubation in EBSS containing [ 3 H]arginine (0.05 mmol/L; 2 μCi/mL) in the absence or presence of 2 mmol/L leucine (+ Leu) or 2 mmol/L leucine + 2 mmol/L lysine (+Leu +Lys) (see Methods). Data are mean ± SD of three independent determinations in a representative experiment, that, repeated twice, gave comparable results. *P < .05, ***P < .001 vs scrambled siRNA; $$ P < .01, $$$ P < .001 vs total uptake; ### P < .001 vs +Leu. Panels C and F: data shown in panels B and E were employed to calculate the relative contribution of system y + L as the difference between total transport and transport measured in the presence of leucine. * P < .05, ** P < .01, *** P < .001 vs scrambled siRNA

| D ISCUSS I ON
In this study, we address the relative contribution of y+LAT1 and y+LAT2 to the uptake of cationic amino acids in different human cell models, so as to help to shed light onto the organ specificity of LPI manifestations. By interacting with the heavy chain 4F2hc, these two proteins form alternative heterodimeric transporters belonging to system y + L, that is responsible for the exchange of dibasic amino acids with neutral amino acids and Na + .
On a molecular basis, y+LAT1 and y+LAT2 share an high grade of sequence homology 6 ; accordingly, we show here that their affinity for the common substrate arginine is very similar (0.182 mmol/L for y+LAT1 and 0.145 mmol/L for y+LAT2).
Differences in V max values are likely consistent with the diverse expression of the transporters in the two cells models. To the same extent, differences in the expression of SLC7A6 and SLC7A7 mRNAs among various in vitro cell models suggest a different contribution of y+LAT2 and y+LAT1 transporters to arginine uptake in the different districts. In particular, while SLC7A6 appears quite F I G U R E 3 Kinetic constants of arginine transport through y+LAT1 in MDM, and through y+LAT2 in fibroblasts. MDM and fibroblasts were transfected for 72 h with SLC7A6 and SLC7A7 siRNA, respectively. After gene silencing, cells were incubated for 1 min in the presence of the indicated concentrations (from 0.04 to 1 mmol/L) of [ 3 H]arginine (from 2 to 10 μCi/mL), in the absence and in the presence of 5 mmol/L leucine. Non-linear fitting of the data, obtained by calculating y + L activity as the difference between total uptake and the uptake obtained in the presence of leucine, was performed employing Equation (1)  indicates that this transporter, codified by ATB0,+/SLC6A14, localizes in murine colon, lung and eye, 18 while, in humans its expression is detected at low levels in normal cells, but significantly induced under pathological conditions, such as tumours and inflammation. 19 In our hands, system B 0,+ is not operative in Caco-2 human intestinal cells and in primary renal cells, a finding congruent with lack of expression of SLC6A14 in these models, as well as in MatTek's 3D tissue model of human small intestine (EpiIntestinal) (data not shown). Consistently, Ahmadi et al recently reported the absence of expression of this gene in Caco-2 cells. 20 In contrast, system B 0,+ is readily detectable in respiratory epithelium where, beside arginine, it accepts carnitine. 13,16 As far as the immune system is concerned, experimental and clinical evidence sustains a key role for circulating cells in the onset of LPI pulmonary and immunological complications. 21 To this concern, we demonstrated in years that system y + L-mediated arginine transport is severely compromised in LPI monocytes 7 and that phagocytosis is defective in monocytes-derived macrophages (MDM) obtained from LPI patients, 22 suggesting a role for these cells in LPI-associated immune dysfunctions. Now, we demonstrate here that, although expressing the mRNA for both SLC7A6 and SLC7A7, MDM largely depend on y+LAT1 for arginine transport.
Gene silencing of both transporters causes, indeed, a significant decrease of system y+L activity; however, it is only SLC7A7 silencing that almost completely suppresses system y+L-mediated arginine transport, while the knockdown of SLC7A6/y+LAT2 only modestly impact on the uptake of the amino acid, confirming that y+LAT1 is the predominant transporter in monocytes/macrophages.
It is important to remind, however, that the mechanisms linking the genetic defect in innate immune cells to the clinical manifestations of the disease still remain to be elucidated. The original hypothesis was that the impairment of arginine efflux due to y+LAT1 defect in LPI immune cells leads to an intracellular accumulation of arginine responsible for an increased production of the inflammatory mediator nitric oxide. 23 More recently, however, we demonstrated that the silencing of SLC7A7 gene in THP-1 monocytes associates with the induction of an inflammatory phenotype, no matter arginine intracellular concentration, 24 thus proposing to enrol LPI in the group of auto-inflammatory diseases. This finding is consistent with the occurrence in LPI patients of hemophagocytic lymphohistiocytosis (HLH), a disorder caused by an uncontrolled and self-sustained cytokine-driven immune activation of T lymphocytes and macrophages. 25 We show here, in accordance with data by Werner et al, 26 that CD3 + lymphocytes almost exclusively express SLC7A6; as a consequence, no alteration of arginine transport is expected to occur in LPI T lymphocytes. According to this finding, the pathogenetic mechanism for LPI-associated HLH is expected to involve pathological monocyte/macrophages, rather than T cells.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.