Amino Acid Transporter LAT3 Is Required for Podocyte Development and Function

Abstract

LAT3 is a Na+-independent neutral l-amino acid transporter recently isolated from a human hepatocellular carcinoma cell line. Although liver, skeletal muscle, and pancreas are known to express LAT3, the tissue distribution and physiologic function of this transporter are not completely understood. Here, we observed that glomeruli express LAT3. Immunofluorescence, confocal microscopy, and immunoelectron microscopy revealed that LAT3 localizes to the apical plasma membrane of podocyte foot processes. In mice, starvation upregulated glomerular LAT3, phosphorylated AKT1, reconstituted the actin network, and elongated foot processes. In the fetal kidney, we observed intense LAT3 expression at the capillary loops stage of renal development. Finally, zebrafish morphants lacking lat3 function showed collapsed glomeruli with thickened glomerular basement membranes. Permeability studies of the glomerular filtration barrier in these zebrafish morphants demonstrated a disruption of selective glomerular permeability. Our data suggest that LAT3 may play a crucial role in the development and maintenance of podocyte structure and function by regulating protein synthesis and the actin cytoskeleton.

The development of normal glomerular structure requires coordination among all glomerular cells: Endothelial cells, mesangial cells, parietal epithelial cells, and visceral epithelial cells (podocytes). The major function of the glomerulus is the ultrafiltration of blood plasma resulting in primary urine. The glomerular ultrafiltration barrier consists of three layers: The fenestrated endothelial cells, the glomerular basement membrane (GBM), and the podocyte foot processes with their interdigitating slit diaphragms.¹ Although the physiologic role of the fenestrated endothelial cells in the glomerular ultrafiltration barrier remains obscure, that the angiogenic vascular endothelial growth factor is expressed in differentiating podocytes and its receptor, flk-1, is expressed in neighboring endothelial cells suggest the pivotal role of cell–cell communication between endothelial cells and podocytes in glomerulogenesis.^2,3 In addition, it is apparent that, in the zebrafish, in avascular mutant cloche, where endothelial cell development is blocked in the early stages, podocytes are well differentiated, including WT1 expression as well as foot processes and GBM formation. This suggests that podocytes are able to differentiate morphologically and maintain GBM in the absence of endothelial cells or endothelial-derived signals, at least during its early developmental stages.⁴ Podocytes are highly specialized, postmitotic cells of the mature glomerulus.^5–7 Although several transcription factors such as FOXC2, MafB, WT1, and Lmx1b^8,9 have been known to play an important role in the podocyte differentiation, the downstream targets and regulatory machinery remain largely elusive.

Nutrients are essential factors required for cell survival and organ development. It has become more clear that amino acids not only are important as substrates for the synthesis of proteins but also function as regulators of fluxes through major metabolic pathways.^10–13 In particular, branched chain amino acids (BCAAs) have been found to exert regulatory effects on several cell functions, such as total protein synthesis and degradation, glucose metabolism, and insulin production.^14–17 Amino acid transport across the plasma membrane regulates the flow of these nutrients into cells or from cells and thus participates in interorgan amino acid nutrition. The transfer of amino acids across the hydrophobic domain of the plasma membrane is mediated by specific transporter proteins that recognize, bind, and transfer amino acids from the extracellular space into cells, or vice versa.¹³ With regard to amino acid transport in the glomerulus, it is surprising that only one study demonstrated that differentiated cultured mouse podocytes express Na⁺-dependent transport system, in this case, using reverse transcriptase–PCR (RT-PCR) and patch-clamp experiments.¹⁸ The 20 amino acids are classified on the basis of the acidic, basic, or neutral charge of their side chains. Neutral amino acids are transported through the plasma membrane via Na⁺-dependent and Na⁺-independent transport systems,¹³ in which Na⁺-independent system L (LAT) is one of the major routes to provide the cells with BCAAs. By means of expression cloning, we identified the first isoform of system L amino acid transporter, LAT1, from the C6 rat glioma cell line.¹⁹ Despite its widely known expression in many malignant cells,^20,21 in the normal tissues, the protein product of LAT1 has been shown to exist only in the plasma membrane of vascular endothelial cells forming the blood-brain barrier²² and of syncytiotrophoblastic cells in the placenta.²³ Next, LAT2 was identified by our laboratory and others.^24,25 LAT2 is ubiquitously expressed compared with LAT1 and is found in the kidney, where it is preferentially expressed in the proximal tubules but not in the glomerulus.²⁴ We also isolated a cDNA encoding a novel Na⁺-independent neutral L amino acid transporter, LAT3, from the human hepatocarcinoma cell line FLC4 by expression cloning.²⁶ Northern blot analysis of poly(A)⁺ RNA from human tissues showed predominant expression of LAT3 transcript in the liver, skeletal muscle, and pancreas.²⁶ By using mice tissues, we further confirmed LAT3 protein to be expressed in the plasma membrane of the liver and skeletal muscle, where expression was upregulated by in vivo starvation experiments.²⁷ In this study, we show for the first time that LAT3 is also expressed in both mouse and human glomerulus, and LAT3 protein is specifically localized in the apical plasma membrane of podocyte foot processes. Moreover, morpholino-mediated knockdown of zebrafish lat3 resulted in a disorganized GBM architecture and concomitant loss of selective filtration barrier permeability. These results led us to identify LAT3 as a possible regulator of podocyte function and filtration barrier integrity.

RESULTS

LAT3 Is Expressed in the Mouse Glomeruli

We conducted studies to determine whether LAT3 is expressed in the mouse kidney. RT-PCR experiments demonstrated that the LAT3 transcript was expressed in the kidney cortex as well as liver, pancreas, and skeletal muscle (Figure 1A).²⁷ In the analysis of kidney fractions, both RT-PCR and Western blotting revealed the presence of LAT3 mRNA and protein exclusively in the glomerular fractions, whereas no expression was detected in the kidney extract excluding glomeruli (Figure 1, A and B). On Western blots, glomerular LAT3 migrated as an approximately 58-kD band. Preabsorption by the immunogen peptide (peptide II, amino acids 336 to 348) and peptide III (amino acids 335 to 349) abolished these protein bands, indicating the specificity of this anti-LAT3 antibody (Figure 1C). Immunohistochemistry identified LAT3 to be specifically located in the glomerular tuft (Figure 1D).

Figure 1.

Characterization of LAT3 protein and mRNA in the mouse kidney. (A) RT-PCR of mouse LAT3. LAT3 transcript is present abundantly and specifically in the glomerulus fraction. (B and C) Western blot study of mouse LAT3. (B) Glomerulus fraction revealed the strong expression of LAT3 migrated at approximately 58 kD, whereas no protein expression was observed in the kidney cortex fraction lacking glomerular cells. (C) The intensity of 58-kD immunoband was reduced when the membrane was reacted with anti-mouse LAT3 antibody preabsorbed by synthetic peptide I (amino acids 336 to 344). Preabsorption of anti-LAT3 antibody by both peptide II (immunogen peptide, amino acids 336 to 348) and peptide III (amino acids 335 to 349) completely abrogated positive immunoband. (D) Immunohistochemistry of LAT3 in the mouse kidney. Specific antibody (LAT3 Ab) detected exclusive expression of LAT3 protein in the glomerulus (*), whereas no positive staining was seen in the glomerulus reacted with antibody preabsorbed by immunogen peptide (**). Higher magnification (*) shows LAT3 to be distributed in the capillary tufts, whereas no obvious expression was visible at higher magnification (**).

LAT3 Is Localized in the Apical Plasma Membrane of Podocyte Foot Processes

We conducted studies to verify further LAT3 expression in the human glomerulus. With the use of RT-PCR and specific primers corresponding to human LAT3 cDNA, a 511-bp long cDNA fragment could be amplified from the isolated total RNA of human glomeruli as well as the FLC4 cell line²⁶ as a positive control (Figure 2A). Direct sequence analysis confirmed the existence of human LAT3 in the amplified cDNA fragment (data not shown). The anti-LAT3 antibody defined the presence of human LAT3 protein band migrated at approximately 58 kD in the glomerular sample, whereas the same antibody preabsorbed by the synthetic peptide did not give any positive immunoband (Figure 2B). The antibody demonstrated exclusive expression of LAT3 protein in the glomeruli (Figure 2C), where that was revealed as a fine dotted pattern (arrow) along the glomerular capillary walls (Figure 2C). To verify the subcellular localization of LAT3 in the human glomerulus, we performed double-labeling immunofluorescence and confocal microscopy. As shown in Figure 2D, the staining for LAT3 was found on the urinary side of laminin α5 immunoreactivity, whereas double labeling with anti-nephrin antibody revealed partial overlapping of LAT3 and nephrin. Finally, pre-embedding immunoelectron microscopy with anti-LAT3 antibody clearly revealed that LAT3 protein is predominantly distributed in the apical plasma membrane of the podocyte foot processes, whereas no significant signal was observed in the plasma membrane of podocyte cell body (Figure 2E). The specimens reacted with immunogen peptide–preabsorbed LAT3 antibody did not show positive signals, demonstrating specificity of the foot process localization (Figure 2E).

Figure 2.

Expression of LAT3 in the human glomerulus. (A) RT-PCR showed a cDNA fragment (511 bp) from mRNA isolated from human glomeruli, consistent with positive control FLC4 cell line. (B) Western blot analysis revealed presence of LAT3 protein migrated at approximately 58 kD, the same as in mouse, whereas peptide-preabsorbed antibody did not identify this band. (C) Immunohistochemistry of LAT3 in the fresh-frozen human kidney. Specific antibody (anti-LAT3 Ab) detected exclusively the presence of LAT3 protein in the glomerulus as a clear capillary loop pattern. Higher magnification images show LAT3 protein in a dotted pattern along the capillary wall (enlarged from rectangle). Antibody preabsorbed by immunogen peptide (Preabsorbed LAT3 Ab) did not show any positive immunostaining. (D) Double-labeling immunofluorescence and confocal microscopy. Frozen sections from human kidney were reacted with anti-LAT3 antibody and anti-nephrin antibody or anti-LAT3 antibody and anti-laminin α5 antibody. Partial co-localization of LAT3 with nephrin was observed, whereas no co-localization between LAT3 andlaminin α5 antibody. Partial co-localization of LAT3 with nephrin was observed, whereas no co-localization between LAT3 and laminin α5 was seen. (E) Subcellular localization of LAT3 in human glomerulus. Pre-embedding immunoelectron microscopy revealed the exclusive distribution of LAT3 in the apical plasma membrane of the foot processes.

Podocytes Exhibit System L Transport System

We conducted studies to examine whether podocytes indeed possess LAT3 and whether podocytes exhibit system L transport system physiologic activity. Both RT-PCR and Western blot study revealed the presence of LAT3 mRNA and protein in the samples from cultured mouse podocytes as well as isolated mouse glomeruli (Figure 3A). ¹⁴C-labeled l-leucine uptake was significantly inhibited in the presence of 2-aminobicyclo[2.2.1]heptane-2-carboxylic acid (BCH), a system L transporter–specific inhibitor, and nonradiolabeled l-leucine, indicating the presence of system L transport system in podocytes (Figure 3B). In terms of the protein expression with the system L transporter family, only LAT2 and LAT4 are known to be expressed in the tubules at a protein level but not in the glomerulus.^24,28 Thus, the amino acid transporter inhibited by system L selective inhibitor BCH in cultured podocytes was most likely LAT3.

Figure 3.

Expression of LAT3 in mouse podocytes. (A) Both LAT3 mRNA and protein were present in cultured mouse podocytes. (B) The uptake of ¹⁴C-l-leucine was distinctly inhibited in the presence of selective inhibitor BCH and nonlabeled leucine, indicating mouse podocytes express the system L amino acid transporter.

Nutrient Starvation Upregulates Glomerular LAT3, Induces AKT1 Phosphorylation, and Elongates Podocyte Foot Processes

Our previous study revealed the upregulation of LAT3 mRNA and protein in the liver and skeletal muscle of nutrient-starved mice.²⁷ We conducted studies to determine whether nutrient starvation also increases glomerular podocyte LAT3 expression. In Figure 4A, a clear increase of the LAT3 mRNA and protein levels in the isolated glomerular sample from starved mice is apparent, compared with that of fed mice, indicating that glomerular podocyte LAT3 responds to nutrient starvation. Immunofluorescence and confocal microscopy on sections from both fed and nutrient-starved mice revealed no apparent difference of the LAT3 localization (Figure 4B). Evidence suggested that nutrients, especially amino acids, regulate intracellular signaling through the mammalian target of rapamycin (mTOR) complex, controlling much of the intracellular machinery.^29–31 In addition, a recent study revealed that mTOR complex 2 exists and regulates cytoskeleton structure in podocytes³²; therefore, we examined whether glomerular podocytes undergo morphologic change by nutrient starvation. Interestingly, podocyte cell body in starved mice was observed to be flatter than that of fed mice by scanning electron microscopy (Figure 4C, arrow). Moreover, the space between neighboring foot processes in podocytes of starved mice were wider than that of fed mice (Figure 4C, higher magnification). This was most likely due to the morphologic change of the foot processes in podocytes of starved mice, because transmission electron microscopy detected that the foot processes in starved mice became thin and were elongated compared with those of fed mice (Figure 4D). Furthermore, higher magnification revealed that cortical actin network in the foot processes of starved mice was vertically rearranged (Figure 4D, arrow), and actin bundles were observed to be denser compared with those of fed mice (Figure 4D, arrowhead). Measurement of the length of the foot processes examined by transmission electron microscopy showed significant elongation in starved mice (mean ± SEM, 470.0 ± 19.8 nm) compared with those of fed mice (344.2 ± 10.2 nm; Figure 4E). Western blot analysis using an antibody against phospho-specific AKT-1, known as the specific downstream of mTOR2,³⁰ demonstrated increased phospho-AKT1 phosphorylation in the isolated glomeruli of starved mice compared with those of fed mice (Figure 4F). Immunohistochemistry by using the phospho-AKT1–specific antibody revealed that podocytes were major sites expressing AKT1 phosphorylation in the glomeruli of starved mice (Figure 4G, arrow).

Figure 4.

Upregulation of glomerular LAT3 in food-starved mice. (A) RT-PCR and Western blot analysis revealed the increase of both mRNA and protein with LAT3 in the glomerulus from 48 h starved mice compared with fed mice. (B) Immunofluorescence and confocal microscopy showed increased fluorescence intensity of LAT3 in the glomerulus of mice that were starved for 48 h compared with that of fed mice. There was no distinct difference of their spatial immunostaining pattern. (C) Scanning microscopy revealed flatter podocyte cell bodies in starved mice compared with that of fed mice. (D) Transmission electron microscopy detected that the foot processes in starved mice became thin and were elongated compared with those of fed mice. Higher magnification revealed that cortical actin network in the foot processes of starved mice was vertically rearranged (arrow), and actin bundles were observed to be denser compared with those of fed mice (arrow head). (E) The length of the foot processes was measured as described in the Concise Methods section. Significant elongation of the foot processes in starved mice was revealed. (F) Western blot study with phospho-AKT (serine 473; p-AKT1) detected increase of phosphorylated AKT1 in the glomeruli in starved mice. (G) Immunohistochemistry identified podocyte to be major cells expressing phosphorylated AKT1 in the glomeruli in starved mice.

Expression of LAT3 in the Fetal Kidney

There is no doubt that amino acids are essential components used in organ development.³³ We conducted studies to determine when LAT3 expression commences in the fetal kidney. Cryostat sections from fetal human kidneys of 16 wk of gestation were immunostained with anti-LAT3 antibody. LAT3 was observed intensely in the capillary loop stage and very faintly in the S-shaped body stage of human glomerulus (Figure 5, arrow).

Figure 5.

Immunofluorescence study in the human fetal kidney revealed LAT3 protein to be strongly expressed in the capillary stage glomerulus and faintly in the S-shaped body glomerulus (arrow). Blue, DAPI staining for the nucleus; C, capillary stage; S, S-shaped stage.

Knocked Down lat3 Causes Collapsed Glomerulus and Disrupts Ultrafiltration Barrier in Zebrafish

Next, we studied whether LAT3 plays a crucial role during glomerular development. We chose to determine the functional requirement of LAT3 during glomerulogenesis by morpholino-mediated gene knockdown in the zebrafish. The lat3/slc43a1 gene (hereafter referred to as lat3) is present as a single copy gene within the zebrafish genome. BLAST searches against NCBI EST databases indicated that the lat3 gene gives rise to a 536–amino acid protein that shares 60% amino acid identity with human LAT3 (Supplemental Figure 1). We performed whole-mount in situ hybridization to determine the expression pattern of lat3 during zebrafish embryogenesis. We found lat3 to be expressed in the pharyngeal arches and intermediate mesoderm during the pharyngula stage (Supplemental Figure 2). Expression became downregulated throughout the mesoderm during the hatching stage.

Morpholinos were designed against the lat3 ATG translation start site (lat3 ATG MO) and the exon 2 splice donor site (lat3 SP MO; Supplemental Figure 3A). Injection of lat3 SP MO caused a mis-splicing of the lat3 transcript in morphant embryos. RT-PCR and sequencing demonstrated that intron 2 sequences were unspliced, resulting in an increase in amplicon size from 501 bp in wild-type to 1379 bp in lat3 SP MO–injected embryos (Supplemental Figure 3, B and C). Translation of the morphant PCR product predicts a premature stop codon within the third transmembrane domain. Knockdown of lat3 in zebrafish resulted in pericardial edema by 96 hours postfertilization (hpf) followed by general edema involving the entire larval body (Supplemental Figure 3D). These phenotypes were rescued by co-injection with wild-type lat3 mRNA (Supplemental Figure 3E). Histologic analysis by periodic acid-Schiff (PAS) staining showed that morpholino targeting the splice donor led to the mild collapsed glomerulus where the GBM was still visible (Figure 6A). Conversely, morpholino targeting the ATG led to drastic collapsed glomerulus where the GBM staining was completely absent (Figure 6A). To test the filtration barrier function of these glomeruli, we perfused the vasculature with 500-kD FITC-dextran. Embryos injected with control antisense morpholino showed an absence of tracer in pronephric tubule cells (Figure 6B), indicating selective retention of the large dextran in the vasculature. In contrast, embryos injected with lat3 SP MO showed abundant FITC-positive endosomes in the pronephric tubules, indicating a disruption of filtration barrier. Finally, we examined electron microscopy to evaluate the ultrastructural basis for the breakdown of ultrafiltration barrier function. In the wild-type, the foot processes were regularly arranged along the GBM with consistent spacing between foot processes spanned by slit diaphragms (Figure 6C, arrow). In lat3 SP MO–injected embryos, the intercellular junctions and the organization of the foot processes were still preserved (arrow), whereas the GBM was thicker and the intensity of the lamina densa was reduced compared with those of control injected embryos. In the lat3 ATG MO–injected embryos, there were drastic changes in both the podocytes and the GBM. Although intercellular junctions of the foot processes were partly preserved (Figure 6C, arrows), the shape and size of the foot processes were variable and a pile of debris from podocytes could be seen (Figure 6C). Moreover, podocytes were invaginated into the pitted GBM that was drastically thickened and its lamina densa was split into multiple strands (Figure 6C). With regard to the actin constitution of the foot processes, a short linear pattern was visible in wild-type (Figure 6D, arrow). In contrast, in ATG MO, the actin network appeared as a large dotted pattern, and intense filamentous actin was visible in the invaginated processes (Figure 6D, arrowhead).

Figure 6.

Functional analysis of lat3 in the zebrafish pronephros. (A) Transverse histologic sections of 96 hpf embryos stained with PAS reveal the glomerulus. Both lat3 SP and lat3 ATG morphants revealed collapsed glomeruli. Note that PAS staining in the GBM was absent in the lat3 ATG morphant glomerulus. (B) FITC-dextran (500 kD) was injected into the circulation of larvae. In the control larvae, FITC fluorescence was present in the vasculature surrounding the pronephric ducts; however, no endosomes containing filtered dye were visible in the pronephric duct epithelial cells. In the lat3 SP MO, dye uptake was detectable in the form of small apical endosomes adjacent to the duct lumen (arrow). (C) In control podocytes, the foot processes were regularly arranged along the GBM with consistent spacing between foot processes spanned by slit diaphragms (arrows). In the lat3 SP MO, the GBM was thicker compared with the wild-type, but foot processes and slit diaphragms were preserved (arrows). In the lat3 ATG MO, podocytes were invaginated into the pitted GBM that was drastically thickened, and its lamina densa was split into multiple strands. Bar = 500 nm. (D) Actin constitution of the foot processes in wild-type was a short linear pattern (arrow), whereas in ATG MO, the actin network was observed to have a large dotted pattern (arrow) and intense filamentous actin was visible in the invaginated process (arrowhead). Bar = 100 nm.

DISCUSSION

The major novel observation in the present study is that Na⁺-independent system L amino acid transporter LAT3 is specifically expressed in the apical side of the plasma membrane of the podocyte foot processes. Moreover, nutrient starvation induces mTOR2 activation, affects podocyte actin organization, and is associated with the upregulation of LAT3. Finally, the lack of LAT3 causes a break down of ultrafiltration barrier accompanied with severe disorganization of the GBM. We believe that this is the first in vivo study providing evidence that the system L amino acid transporter actually exerts a crucial function in podocyte biology, in particular, an essential role in the organization of developmental glomerulus.

The molecular nature of human and mouse LAT3 was previously demonstrated by our recent studies.^26,27 Although the deduced amino acid sequence of mouse LAT3 showed 83% identity with that of human LAT3, its functional property was the same as that of human LAT3, preferring BCAAs: l-leucine, l-isoleucine, l-valine, and l-phenylalanine as substrates.^26,27 Starvation actually induced upregulation of mouse LAT3 protein and mRNA in both the liver and skeletal muscle.²⁷ These results suggested that LAT3 might indeed function as an amino acid transporter, transporting BCAAs from the liver and skeletal muscle to the blood stream and thereby participating in the regulatory system of interorgan amino acid nutrition.²⁷ Our previous study with human LAT3 did not detect the positive expression of its transcript in the whole-kidney polyA⁺RNA from Multiple Tissue Northern Blots Membrane,²⁶ whereas in this study, we clearly identified LAT3 protein and mRNA in the samples from the glomerulus. The discrepancy of these two results could be explained by the relatively lower ratio of mRNA of the whole glomerulus in kidney polyA⁺RNA from Multiple Tissue Northern Blots. Regarding the molecular weight of LAT3, our previous article showed that mouse liver LAT3 expresses two isoforms of LAT3, migrated at approximately 58 kD and approximately 48 kD, and both the skeletal muscle and pancreas express only the approximately 48-kD form.²⁷ This study revealed glomerular LAT3 to be mainly approximately 58 kD. Because deduced molecular weight of human LAT3 is estimated to be 62.6 kD, this relative molecular migration on SDS gels is reasonable; however, although our ExPasy search showed that LAT3 contained the motifs for potential N- or O-glycosylation on its amino acid sequence, it seems not to be a glycoprotein, because we could not find any shifted migration of LAT3 when the protein lysates from isolated human and mouse glomeruli were treated with N-glycosidase F and O-glycosidase (data not shown). Therefore, we speculate that some other type of posttranslational modification may be involved in the plasma membrane assembly of LAT3 in the podocyte plasma membrane.

It is great of interest that LAT3 was specifically located at the apical side of the plasma membrane only in the foot processes, not in the primary processes and cell bodies. Although it is difficult to explain this specific pattern of the localization, the starvation experiment may suggest some clues. Namely, in addition to the liver and skeletal muscle,²⁷ starvation seemed to have upregulated mouse LAT3 mRNA and protein of podocyte. Substrate selectivity of LAT3 is distinct from that of LAT1 and LAT2.²⁶ LAT3 shows narrower substrate selectivity than LAT1 and LAT2 and mainly transports BCAAs and phenylalanine. It is known that the concentration of BCAAs in the blood is increased during starvation in human^34–37 and rat.³⁸ Because BCAAs easily filtered through ultrafiltration barrier as a result of its small molecular weight, the level of BCAAs of primary urine in the Bowman's space should be increased during starvation. On the basis of the functional property of LAT3, in which transport of BCAAs are electroneutral and mediated by facilitated diffusion,²⁶ BCAAs are possibly transported from Bowman's space into podocyte cytoplasm via LAT3, especially during starvation.

Given the data with the starved mice and the knocked-down zebrafish, the main contribution of LAT3 function for podocyte biology may be through the mTOR pathway, but it may depend on the developmental stage of the glomerulus. With regard to the implication of LAT3 in the immature podocyte, our zebrafish study showed that not only ATG morphant but also donor SP morphant resulted in the apparent collapsed glomeruli. Especially in the ATG morphant, electron microscopy demonstrated the severely altered structure of the GBM. The relation between the lack of LAT3 and the disorganized GBM remains unclear; however, a recent concept concerning the role of BCAAs, especially leucine, in protein synthesis may be relevant. It has become apparent that leucine is involved in the protein synthesis step by increasing mRNA translation initiation ratio through stimulating the mTOR complex 1.^15,39 Because LAT3 acts to transport leucine, as described already, the lack of LAT3 possibly resulted in the reduction of intracellular leucine level in podocytes, which might cause the decreased or altered production of matrix proteins, thereby leading to the alteration of the GBM construct. Indeed, because PAS recognizes glycans of the proteins, the lack of PAS staining in the GBM of the ATG morphant may be explained by the deficient level of matrix protein or immature protein in the GBM, which may be caused by deficiency of intracellular leucine in developmental podocyte. The synthesis ratio and the turnover ratio between matrices and membrane proteins may be different, which could explain why the SD structure exists despite the altered GBM.

In the mature glomerulus, our data clearly revealed that starvation upregulated LAT3 of podocytes accompanied with elongation of the foot processes and increased phosphorylation of AKT1 on serine 473. mTOR complex 2 is known to phosphorylate AKT on the hydrophobic motif site S473 and to regulate the actin cytoskeleton, possibly through the Rho small GTPase family and protein kinase C.³⁰ Thus, it is most likely that starvation induced LAT3 in podocytes, which possibly led to the induction of the mTOR2 pathway and thereby affected actin organization in podocytes. Our data suggest that LAT3 may play a crucial role in the podocyte biology through regulating the protein synthesis machinery and actin network at both stages of developmental and mature glomeruli. This also suggests that dysfunction of LAT3 may be one mode of pathogenesis in the podocyte injury of the acquired glomerular diseases.

CONCISE METHODS

Tissue Preparation

The experimental protocol was approved by the Animal Care Committee of Kyorin University School of Medicine. Six-week-old male ICR mice (26 to 28 g; Saitama Experimental Animals Supply Co., Ltd., Sugito, Saitama, Japan) were anesthetized by intraperitoneal injection with pentobarbital. For the histologic experiments, kidneys were harvested and embedded in Tissue-Tek OCT compound (Sakura Fine technical Co., Tokyo, Japan). Kidneys were also rapidly frozen and stored in liquid nitrogen for the isolation of total RNA and protein lysate. Glomeruli were isolated from adult ICR mice by means of the magnet beads perfusion method.⁴⁰ For the starvation studies, 15 mice were deprived of food for 48 h, and another 15 fed mice were examined for the controls. Four human kidney cortex samples were obtained from histologically normal regions of fresh kidneys from patients who underwent nephrectomy for cancer, as described previously.^41–45 The isolation of human glomeruli was performed by a differential sieving technique.^41–45 Fetal kidney was obtained from autopsy sample (16 wk of gestation) at the Department of Pathology, Institute of Clinical, University of Tsukuba. Informed consent was obtained from all cases.

Reverse Transcriptase–PCR

Total RNA was extracted from mouse total kidney, isolated glomeruli, liver, pancreas, and skeletal muscle and from isolated human glomeruli, immortalized mouse podocytes (a gift from Dr. Peter Mundel, Mount Sinai School of Medicine, New York, NY),⁴⁶ and FLC4 cells²⁶ using Isogen (Wako Life Science Reagents, Osaka, Japan) according to the manufacturer's instruction. RT-PCR for mouse LAT3 was carried out as described previously.²⁷ For the human LAT3, the sense primer 5′-GCAGCACCAACACCACCCAG-3′ and antisense primer 5′-GTGGGGGAGCAGAGGCTCTT-3′ were used. One microgram of total RNA was amplified under the following conditions: 30 amplification cycles at 94°C for 1 min, 63°C for 1 min, and 72°C for 1 min. Amplification was completed with prolonged synthesis at 72°C for 10 min. PCR products were visualized by ethidium bromide staining after electrophoresis on a 4% NuSieve 3:1 agarose gel.

Antibodies

Anti-mouse LAT3 polyclonal antibody²⁷ and anti-nephrin mAb^44,45,47 were previously described. The following antibodies were purchased from the suppliers as indicated: Anti-human LAT3 polyclonal antibody (Trans Genic, Kumamoto Japan), anti-laminin α5 mAb (clone 4C7; Chemicon Int., Temecula, CA), anti–β-actin mAb (Sigma, Tokyo, Japan), anti-AKT1 phospho-specific (pS473) polyclonal antibody (Rockland, Gilbertsville, PA), horseradish peroxidase–labeled goat anti-rabbit Ig (Dako, Carpinteria, CA), Alexa Fluor 488–conjugated goat anti-rabbit and anti-mouse IgG (Molecular Probes, Eugene, OR), and Texas Red-X goat anti-rabbit and anti-mouse IgG (Molecular Probes).

l-[¹⁴C]Leucine Uptake Experiment in Podocytes

Conditionally immortalized mouse podocyte cell line was maintained as described previously.⁴⁶ l-[¹⁴C]leucine uptake (100 μΜ) was measured in the absence or presence of 10 mM nonradiolabeled l-leucine and BCH, a specific functional inhibitor of the system L amino acid transporter.^26,27

Western Blot Analysis

Protein samples homogenized in ice-cold lysis buffer were separated by 10% SDS-PAGE under reducing conditions.²⁷ Membranes were reacted with anti-mouse LAT3 antibody (0.5 μg/ml) and anti–β-actin antibody (0.1 μg/ml). For verification of the specificity of this antibody for the mouse glomerulus, anti-LAT3 antibody was preabsorbed by three peptides, including synthetic peptide.²⁷ For human glomerular sample, anti-human LAT3 antibody (0.5 μg/ml) and synthetic peptide–preabsorbed LAT3 antibody were reacted. For the mTOR study, samples from isolated glomeruli of fed and starved mice were used and the membrane was reacted with anti-AKT phospho-specific antibody (1:500). Immunocomplex was developed by using the Western Lightning Chemiluminescence reagent (Life Science Products, Boston, MA).

Immunohistochemistry

Human frozen kidney sections and mouse kidney sections fixed with 4% paraformaldehyde-PBS for 30 min were incubated with blocking buffer and then reacted with primary antibodies (anti-human LAT3 antibody, anti-mouse LAT3 antibody, and anti-mouse LAT3 antibody preabsorbed with antigen peptide II, 5 μg/ml each) overnight at 4°C. Paraffin-embedded kidney sections of the fed mice and the mice starved for 48 h were dewaxed and autoclave heated at 120°C for 10 min in Target Retrieval Solution (Dako) for antigen retrieval. The slides were washed with PBS and reacted with anti-AKT phospho-specific antibody (1:100). After washing with PBS, the slides were incubated with horseradish peroxidase–labeled goat anti-rabbit antibody (1:200) at room temperature, and then the slides were developed by immersion in 1.4 mmol/L 3,3′-diaminobenzidene tetrahydrochloride (Sigma Chemical Co., St. Louis, MO) in PBS.²⁷

Immunofluorescence and Confocal Microscopy

For the dual immunostaining of LAT3 and nephrin or laminin, the frozen sections of human adult kidney cortex were incubated with blocking buffer for 1 h at room temperature. The reaction of the primary antibodies LAT3 (5 μg/ml) and nephrin (2 μg/ml) or laminin α5 (5 μg/ml) was conducted for 1 h at room temperature, and then immunocomplexes were visualized using Alexa Fluor 488–conjugated goat anti-rabbit IgG and Texas Red-X goat anti-mouse IgG, respectively. Frozen sections of the kidney samples from fed and starved mice were incubated with blocking buffer for 60 min and then reacted with anti-LAT3 antibody. After washing with PBS, the slides were incubated with Texas Red-X goat anti-rabbit IgG (5 μg/ml). The frozen sections of human fetal kidney were incubated with antibodies against human LAT3 (5 μg/ml) at room temperature and reacted with Alexa Fluor 488–conjugated goat anti-rabbit IgG. Signals were examined under a confocal laser scanning microscope equipped with a Krypton/Argon laser (MRC1024; Bio-Rad, Hercules, CA).

Scanning and Transmission Electron Microscopy

For scanning electron microscopy, mouse kidney tissues were fixed in 2.5% glutaraldehyde and postfixed in 1% OsO₄ for 2 h. Then the samples were dehydrated in a graded series of ethanol (50, 70, 90, 99.5, and 100%), freeze-dried with t-butyl alcohol, mounted, and coated with gold using a sputter coater (JFC-1300 Auto Fine Coater; JEOL, Tokyo, Japan). Finally, the samples were observed under a scanning electron microscope (JSM-5600/ LV; JEOL) operated at 25 kV.

For transmission electron microscopy, tissue samples were fixed in phosphate-buffered 2.5% glutaraldehyde (pH 7.4), postosmicated, and dehydrated with graded alcohols as described previously.⁴⁸ Briefly, after immersion in propylene oxide, the specimens were embedded in Epon 812. Ultrathin sections were cut perpendicularly to the epithelium, doubly stained with uranyl acetate and lead citrate, and examined with a transmission electron microscope (TEM-1010C; JEOL). The length of the foot processes was determined by measurement from the apical tip side to the attachment side on the GBM in 250 of the foot processes from five glomeruli of starved or fed mice. Data were expressed as means ± SEM. The significance of differences was calculated by the two-tailed Mann-Whitney U test.

Immunoelectron Microscopy

Frozen sections of human kidney cortex samples were incubated with anti-human LAT3 antibody as mentioned already. After washing with PBS, sections were fixed in 4% formaldehyde-PBS for 1 h at room temperature and further prepared for electron microscopic observation using the pre-embedding procedure.^41,42 Briefly, after DAB-H₂O₂ treatment, the sections were treated with 1% osmium tetroxide in 0.1 M PB for 10 min, dehydrated by passage through a series of graded ethanols, and embedded in Epon on glass slides. Ultrathin sections were made, stained with 0.1% lead citrate for 7 min, and examined at 80 kV with a transmission electron microscope (JEM-1010; JEOL).

Zebrafish Studies

Wild-type zebrafish were maintained and mated as described previously.^49,50 Dechorionated embryos were kept at 28.5°C in E3 solution with or without 0.003% 1-phenyl-2-thiourea (Sigma, St. Louis, MO) to suppress pigmentation and staged according to hpf. Zebrafish embryos were fixed in Bouin's fixative (Polysciences, Warrington, PA), dehydrated, and embedded in glycolmethacrylate (JB-4 resin; Polysciences) for plastic sectioning. Sections of approximately 3- to 4-μm thickness were cut and stained in PAS and mounted in Permount (Sigma Chemical Co., St. Louis, MO) for light microscopy. Whole embryos were observed using Leica MZ12 dissecting stereomicroscope. Embryos at 96 hpf were fixed for electron microscopy in 2.5% glutaraldehyde/2% paraformaldehyde as described already. The morpholino oligonucleotides were obtained from GENE TOOL, LLC (Philomath, OR). The sequence of the translation blocking oligonucleotide was lat3 ATG MO: 5′-AGCCATTCCCCTGCGAAACACAGAG-3′; the exon 2 splice donor blocking oligonucleotide sequences was lat3 SP MO 5′-TTGAAAGTGTGTGTTCACCTTCCAG-3′. The morpholinos were diluted in injection containing 200 mM KCl and 0.1% Phenol Red (Sigma Chemical Co., St. Louis, MO) and titrated to determine the lowest concentration sufficient to induce abnormal mRNA splicing consistently. Morpholinos for both lat3 and control were injected at a final stock concentration of 0.05 to 0.08 mM into one- to two-cell stage embryos. For each morpholino, more than 400 individual embryos were injected. Standard control oligonucleotide was 5′-CCTCTTACCTCAGTTACAATTTATA-3′ and showed no effect on development. Injections were performed using a microinjector PLI-90 (Harvard Apparatus, Cambridge, MA). The effect of the splice morpholino was verified by RT-PCR from single embryo total RNA with nested primers in flanking exons yielding a 300- to 500-bp amplicon. RT-PCR primers were designed from flanking exon coding sequence to confirm MO oligo efficacy and characterize the altered mRNA splicing products. The primer sets were lat3 sense primer 5′-CTCTCTAGAACAGGCGTTCA-3′ and lat3 antisense primer 5′-ATGACGCATAAGATCCAATC-3′, respectively. Fluorescent dye injection was performed by a solution of 1% lysine-fixable FITC-conjugated dextran in phosphate buffer (500 kD; Molecular probes). Fluorescence tracer was injected into the common cardinal vein of 80 hpf larvae anesthetized with 0.2 mg/ml tricane (3-amino benzoic acid ethyl ester; Sigma, Schnelldorf, Germany) in egg water. After overnight incubation, the embryos were fixed, embedded in plastic, and sectioned. Uptake of filtered fluorescence dextran by duct cells was evaluated in serial sections using a Leica fluorescence microscope.

DISCLOSURES

None.