Small molecule inhibitors provide insights into the relevance of LAT1 and LAT2 in materno‐foetal amino acid transport

Abstract The placenta supplies the foetus with critical nutrients such as essential amino acids (AA, eg leucine) for development and growth. It also represents a cellular barrier which is formed by a polarized, differentiated syncytiotrophoblast (STB) monolayer. Active Na+‐independent leucine transport across the placenta is mainly attributed to the System L transporters LAT1/SLC7A5 and LAT2/SLC7A8. This study explored the influence of trophoblast differentiation on the activity of LAT1/LAT2 and the relevance of LAT1/LAT2 in leucine uptake and transfer in trophoblasts by applying specific small molecule inhibitors (JPH203/JG336/JX009). L‐leucine uptake (total dose = 167 μmol/L) was sensitive to LAT1‐specific inhibition by JPH203 (EC50 = 2.55 µmol/L). The inhibition efficiency of JPH203 was increased by an additional methoxy group in the JPH203‐derivate JG336 (EC50 = 1.99 µmol/L). Interestingly, JX009 showed efficient System L inhibition (EC50 = 2.35 µmol/L) and was the most potent inhibitor of leucine uptake in trophoblasts. The application of JPH203 and JX009 in Transwell®‐based leucine transfer revealed LAT1 as the major accumulative transporter at the apical membrane, but other System L transporters such as LAT2 as rate‐limiting for leucine efflux across the basal membrane. Therefore, differential specificity of the applied inhibitors allowed for estimation of the contribution of LAT1 and LAT2 in materno‐foetal AA transfer and their potential impact in pregnancy diseases associated with impaired foetal growth.

family, are driven by System A-dependent AA gradients. Both AA transporter families are highly expressed in polarized epithelial tissues such as blood-brain and placental barriers. 3,4 There are two important cell layers in the placental villi coordinating the nutrient transfer across the placental barrier, namely syncytiotrophoblasts (STB) and foetal capillary endothelial cells. Endothelial cells lining the foetal vessels allow relatively unrestricted paracellular diffusion of small molecules like glucose and AA through endothelial junctions. 5 In contrast, STB represent the limiting barrier for AA due to the formation of an epithelial syncytium composed of two polarized monolayers, the microvillous plasma membrane (MVM) facing the maternal blood supply and the basal membrane (BM) directed towards the foetal capillary ( Figure 1A).
The Na + -independent System L transporters expressed in the human placenta are heterodimeric exchangers consisting of the light chain L-type AA transporter LAT1 (SLC7A5) or LAT2 (SLC7A8) covalently attached to the heavy chain 4F2hc (SLC3A2). Moreover, the SLC43 family members LAT3 (SLC43A1) and LAT4 (SLC43A2), known to be involved in facilitated AA diffusion, 6,7 are expressed at the BM ( Figure 1A). Both LAT1 and LAT2 are predominantly localized to the MVM of human term placenta, LAT2 as well as LAT3 and LAT4 are also present at the BM and in endothelial cells lining the foetal capillaries. 7,8 In the last decade, increasing evidence suggests a tight link between the reduced activity of placental System L transporters and intrauterine growth restriction (IUGR), 9,10 and their up-regulation in placentae of large for gestational age (LGA) infants. 11 Such altered foetal development has a fundamental impact on lifelong health and wellbeing, and may contribute by foetal programming to an increased prevalence for cardiovascular disease and diabetes/adiposity later in life. [12][13][14][15] Notably, it has been reported that LAT1 or its associated glycoprotein 4F2hc is involved in placenta decidualization and fusogenic trophoblast differentiation. 16 This could imply that the diminished leucine uptake found in knock-down cell models 17 results rather from failure in trophoblast differentiation than from reduced SLC7 transport activity. Hence, small molecules which induce only short-term inhibition of transporter activity are a valuable experimental tool to study placental transfer mechanisms as they will not affect trophoblast differentiation and associated processes like trophoblast fusion. Therefore, studying placental AA transfer by using specific small molecule inhibitors of AA transporters instead of silencing or knock-out could help to reveal the relevance of LAT1 or LAT2 in materno-foetal leucine transfer without affecting cell differentiation.
Since LAT1 was found to be selectively expressed and up-regulated in various rapidly proliferative cancer types [18][19][20] and has a putative role in drug delivery across the blood-brain barrier, 21 efforts have been made to pharmaceutically target this transporter using substrate-mimicking or virtual screening approaches. [22][23][24][25][26] The substrate-mimicking tyrosine analog JPH203 (also known as  was tested in several in vitro and in vivo cancer cell proliferation experiments and described as potent LAT1-specific inhibitor. [27][28][29][30][31] To delineate the contribution of LAT1 from LAT2 in materno-foetal leucine transport, we synthesized the LAT1-specific inhibitor JPH203, the structurally closely related inhibitor (JG336), as well as a third small molecule inhibitor (JX009) with comparable leucine uptake inhibition efficiency but lower LAT1-specificity (structures see Figure 4).
In this study, we assessed the contribution of LAT1-mediated Na + -independent leucine uptake into trophoblasts and investigated the transfer of this essential AA across the placental barrier by transient inhibition. To achieve these goals, we (a) tested primary human trophoblast and BeWo (clone b30) cell models for LAT1, LAT2 and 4F2hc expression and leucine uptake capacity under Na + -free conditions, (b) investigated whether the differentiation status of primary trophoblasts and BeWo cells has an impact on leucine uptake and (c) assessed the contribution of LAT1 for the uptake and transfer of leucine by application of small molecule inhibitors.

| MATERIAL S AND ME THODS
All chemicals and reagents were purchased from Sigma-Aldrich in Switzerland unless otherwise stated.

| Isolation of primary human trophoblast cells
Placentae from normal healthy pregnancies were collected after elective Caesarean section at the Division of Gynecology and Obstetrics, Lindenhofgruppe, Bern, Switzerland. Details on the study subjects are found in Table 1. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of the Canton of Bern (Basec Nr2016-00250). The collected tissue was used to isolate primary cytotrophoblasts as previously described. 32 Isolated cells were cultured on Cell-BIND plates in Dulbecco's modified Eagle's medium containing 4.5 g/L glucose (DMEM-highGlucose; Gibco, Paisley, UK) and characterized by analysing the expression of cytokeratin-7 and vimentin as previously described. 33 Primary trophoblasts were evaluated for leucine uptake at the cytotrophoblast (CTB) and syncytiotrophoblast (STB) stage (ie after 12 and 48 hours of culture, respectively) when spontaneous differentiation and fusion has occurred.

| BeWo cell culture
BeWo cells (clone b30) were cultured in DMEM containing 1.0 g/L glucose (DMEM-lowGlucose; Gibco). BeWo cells require stimulation by forskolin to induce STB formation. 34 Unstimulated BeWo cells representing the CTB stage were studied after 24 hours of culture; differentiated, syncytialized BeWo cells were analysed after stimulation with 100 μmol/L forskolin for 48 hours when markers for syncytialization are significantly increased. 32 For leucine transfer experiments, BeWo cells were seeded at a density of 100 000 cells/cm 2 onto permeable membranes (0.4 μm pore size) mounted in 12-well format from the Transwell ® system (Corning Inc., Corning, NY, USA). Cells were cultured in DMEM-lowGlucose at 37°C with 5% CO 2 atmosphere for 7-14 days.

| Placental membrane protein isolation
Placental tissues from healthy pregnancies were used to simultaneously isolate microvillous membranes (MVM) and basal membranes (BM) by Mg 2+ precipitation based on a published method 35

F I G U R E 1
Schematic representation of Na + -independent leucine transfer across the placental barrier with asymmetric LAT1 expression. Scheme of the placental villous structures as interface between the foetal capillaries and the maternal circulation (A, upper left panel) and the human placental barrier (A, upper right panel) involving three different cell types: endothelial cells lining the foetal capillaries (EC), varying intercellular space (pink), single-nucleated cytotrophoblasts (CTB) and syncytiotrophoblasts (STB) forming a multi-nucleated monolayer, which is in direct contact with maternal blood and mainly responsible for materno-foetal nutrient transport. In (A) lower panel, Na + -independent System L exchanger and System L-like facilitators known to be expressed in the human placenta are shown. They are present either at the apical microvillous membrane (MVM) or at the basal membrane (BM) of STB. LAT1 (SLC7A5), expressed at the MVM, and LAT2 (SLC7A8), present at both the MVM and BM, are SLC7 family members and colocalize with their heavy chain partner 4F2hc (SLC3A2, CD98). LAT3 (SLC43A1) and LAT4 (SLC43A2) were described as System L-like facilitators and are most likely expressed at the BM. Increased intracellular leucine concentrations stimulate mTORC1-mediated cell proliferation and survival, and repress the ATF4-mediated amino acid balance sensing system, which reduces global translation and increases biosynthesis. (B) Representative picture showing apical LAT1 localization in immunohistochemistry of human term placenta. Arrows indicate the apical MVM, that is, maternal blood orientated side; arrowheads depict the BM of the STB. There was no signal in negative control. (C) upper panel, representative immunoblot of 6 purified membrane protein samples isolated from term placental tissues. (C) lower panel, LAT1 was highly expressed in MVM, but negligible in BM protein fractions isolated from a total of 11 human term placentae. The data is normalized to the respective tissue homogenate before membrane separation and shown as boxplots with Tukey whiskers (Mann-Whitney test, α = 0.05; ***P < 0.0001) For the reference signal (loading control) a mouse anti-beta-actin antibody (A2228, Sigma Aldrich, Buchs SG, Switzerland) was used.
Details regarding the immunoblotting procedures are described in Supporting Information. Densitometrical analysis of immunoblots was performed using the LI-COR OdysseyW Imaging System. LAT1 expression in placental tissue was visualized by immunohistochemistry using the same LAT1-specific antibody (10 µg/mL). The preparation of cryosections and the staining procedure including visualization are described in Supporting Information.

| Expression changes of LAT1, LAT2 and 4F2hc during trophoblast differentiation
To compare changes in expression of LAT1, LAT2 and 4F2hc during cell differentiation, mRNA and protein levels were determined by RT-qPCR and immunoblotting, respectively. RNA isolation, firststrand cDNA synthesis and RT-qPCR analysis were performed as previously described. 36 Expression results were normalized to the mean of the reference genes YWHAZ, GAPDH and β-actin. Primer nucleotide sequences are listed in Table S1. For protein quantification 40 µg cell lysates were loaded on 10% acrylamide gels and separated by SDS-PAGE. The immobilized bands were semi-dry transferred to nitrocellulose membranes (GE Healthcare, Glattbrugg, Switzerland). Blots were blocked with 5% w/v non-fat milk in Tris Buffered Saline with 0.1% Tween-20 (TBST). Proteins were detected by antibodies against LAT1 (see above), LAT2 (Anti-SLC7A8, AV43930, Sigma-Aldrich) and 4F2hc (CD98 (E-5), sc-376815, Santa Cruz, Biotechnology Inc., Heidelberg, Germany). Protein content was measured using a commercial Pierce™ BCA Protein Assay Kit.

| Leucine uptake assay
The leucine uptake protocol was based on a recently published method 30

| Leucine transfer across a polarized trophoblast monolayer
Before starting the transfer experiment, the formation of a tight trophoblast monolayer in BeWo cells was monitored by measuring transepithelial electrical resistance (TEER, [Ω*cm 2 ]) and passive diffusion as reported previously. 38

| Statistical analysis
Anthropometric and clinical data are expressed as mean ± standard deviation (SD) for normal distribution or median with interquartile range for not normal distribution. Student t tests were performed to detect differences in mRNA levels between CTBs and STBs.
MVM and BM membrane protein fractions were compared by using Mann-Whitney test. A P-value <0.05 was considered as statistically significant. Statistical comparisons were performed using GraphPad Prism software, La Jolla, USA.

| LAT1, LAT2 and 4F2hc are asymmetrically expressed in the human placenta
Histological investigation of healthy placental tissue demonstrated strong expression of LAT1 at the apical membrane of STB (arrows in Figure 1B) which was confirmed by the strong signal found in the MVM fraction ( Figure 1C). The asymmetric expression of LAT1 in syncytialized trophoblasts was further confirmed by immunoblot analysis in paired MVM/BM isolated from term control placentae. LAT1 was predominantly expressed at the MVM as reflected in a mean MVM to BM ratio of 6.9 (range 3.5-14.9; Figure 1C). Western blot analysis of 12 MVM and BM pairs revealed a significantly higher expression of LAT1 in MVM compared to BM (P = 0.0015; Figure 1C). The results of these two independent experimental approaches suggest almost exclusive LAT1 expression at the apical membrane which is in direct contact with maternal blood. In contrast, LAT2 was found to be expressed in both MVM and BM ( Figure S2), with an apparent predominance of the two LAT2 variants (30 and 50 kDa) in BM as reported before. 4 Expression of 4F2hc, the heavy chain partner protein of LAT1 and LAT2, was higher in MVM as compared to BM ( Figure S2).

| Trophoblast differentiation induces upregulation of LAT1 and 4F2hc expression
To study the role of LAT1 at the placental barrier on the cellular level,

| Leucine uptake increases with trophoblast differentiation
We further investigated whether the expression changes caused by trophoblast differentiation resulted also in an increased leucine uptake efficiency in both primary (n = 3; Figure 3A) and BeWo cells (n = 3; Figure 3B). Indeed, both spontaneously differentiated primary tropho- on functional level leucine uptake capacity was higher in BeWo cells compared to primary trophoblasts, but the kinetic changes due to cell differentiation were similar in both cell models (Figure 3).

| Leucine uptake is modulated by System L-specific small molecule inhibitors
Based on the high leucine uptake capacity and the high expression

| Transfer of leucine across the placental barrier is reduced by inhibitors of System L transporters
The small molecule inhibitors JPH203, JG336 and JX009 were also used to assess the relevance of LAT1 and LAT2 in leucine transport across the placental barrier using the Transwell ® system ( Figure 5).  Figure 5C). In contrast, JX009 had no effect on the intracellular 3 [H]-leucine concentration within the tested time course.
The higher retention of leucine in JX009 treated cells suggests inhibition of leucine secretion towards the lower foetal compartment.

| D ISCUSS I ON
The placenta plays a crucial role in the distribution of essential nutrients from the mother to the growing foetus. In this context, which results in an increased uptake, transport or transfer capacity as previously shown for the alanine-serine-cysteine transporters and for alpha-aminoisobutyric acid transport. 49,50 This concept is in line with recent findings that differentiation processes such as syncytium formation resulted in an up-regulation of MVM associated membrane proteins. 16 Based on the validation of expression in the two trophoblast cell models ( Figure 2) and comparable uptake behaviour (Figure 3), the BeWo cell line was chosen to test the effect of different SLC7-specific inhibitors. However, to our knowledge LAT1 and LAT2 localization in MVM or BM of BeWo cells after forskolin-induced differentiation has not been reported yet. Thus localization differences between primary trophoblasts and BeWo cells cannot be completely excluded and therefore the Transwell ® data using polarized BeWo cells should be interpreted with caution.
The low molecular weight inhibitors JPH203, JG336 and JX009 were assessed for their capacity to reduce leucine uptake into BeWo cells (Figure 4) and leucine transfer across the placental barrier ( Figure 5). Due to the varying SLC7 specificity, different inhibition patterns were expected. JPH203 was previously reported as potent, LAT1-specific inhibitor. 27-31,46 JX009 had been described in the patent literature 37 to inhibit LAT1-mediated transport with similar efficacy as JPH203, albeit with significantly lower specificity, that is, it is also a potent LAT2 inhibitor. As no other LAT2 inhibitors have been described in the literature, JX009 was considered the best available tool for the assessment of LAT2-related transport of leucine. JG336 was prepared based on a limited structure activity relationships (SAR) study on JPH203 that investigated the effects of electron-donating (such as the methoxy group in JG336, see dashed red box in Figure 4B) BCH which has been also tested in this study ( Figure 4D) and other investigations. 51,52 In comparison to JX009 this inhibitor is >100- The leucine in the upper compartment was spiked with 1 µCi/mL 3 H-L-leucine for transfer quantification. C, The intracellular leucine content was measured after washing the cell layers with DPBS at the end of the experiment (after 6 h). A significant reduction of intracellular leucine contents was detected for the LAT1-specific inhibitors JPH203 (P = 0.020) and JG336 (P = 0.0001), but not for the System L inhibitor JX009-(P = 0.215). The detected intracellular leucine retention caused through LAT1 and LAT2 inhibition by JX009 suggests a System L-dependent leucine efflux across the BM as ratelimiting step. Statistical analyses were performed using a parametric one-way ANOVA analysis (α = 0.05) The selective expression and overexpression in various cancer types with poor survival expectancy 20,56-60 made LAT1 an interesting pharmaceutical target. 27 This study demonstrates that the choriocarcinoma cell line BeWo could serve as useful model system to test putative LAT1 inhibitors and characterize their effect at an active AA transporting and physiologically relevant cell barrier.
It was shown that high and asymmetric expression of SLC7 family members makes the placental leucine uptake sensitive to LAT1-and LAT2 inhibition. 8,41 Furthermore, the application of compounds like the LAT1-specific inhibitor JPH203 or the less specific SLC7transporter inhibitor JX009, allows for distinguishing the contribution of single leucine transporters across the placental barrier. The short-term treatment with small molecule inhibitors also reduces the risk of affecting trophoblast fusion and differentiation which could occur after long-term gene silencing or constitutive knockout. 16

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest. writing -review and editing (lead).