Introduction Heteromeric amino acid transporters (HATs) are composed of a heavy subunit (rBAT or 4F2hc) and a light subunit (b0 + AT, ASC1, LAT1, LAT2, y + LAT1, y + LAT2 and xCT), joined by a disulfide bridge (Chillaron et al. 2001). rBAT and 4F2hc are type II membrane glycoproteins (N-terminal cytoplasmic). Both have a single transmembrane segment, an N-terminal intracellular tail and an extracellular domain (ectodomain). As far as we know, the role of the heavy subunit is facilitating the transit of the light subunit to the plasma membrane. The light subunits are polytopic proteins unglycosylated, with 12 transmembrane segments, and the N-and C-terminal intracellular. The light subunit is the catalytically active subunit which confers specificity to the heterodimer on the transport system (Reig et al. 2002): LAT1 and LAT2 for system L , y+LAT1 and y+LAT2 for system y + L, asc for system ASC1; xCT for system Xc -, and b0+at for system b0, + (Chillaron et al. 2001). Results Overexpression of these human proteins was carried out with the methylotrophic yeast Pichia pastoris (strain KM71H) as expression system (based on Long et al. 2005). The main objective was to generate enough protein in a high level of purity to study the structure and check their function by transport assays. The different subunits, light and heavy, were cloned into the expression vector pPICZ (Invitrogen). To facilitate the purification of the different proteins, a cluster of 10 histidines was introduced by PCR at the N-terminus of the heavy subunits and a StrepTagII (IBA) at the N-terminus of the light subunits. 4F2hc is a glycoprotein with 4 possible targets for glycosylation. The glycosylations confer heterogeneity to protein, thus glycosylation targets were eliminated by directed mutagenesis. From all these human heavy and light subunits and heterodimers, only 4F2hc for the heavy subunits, LAT2 for the light subunits, and the heterodimer 4F2hc/LAT2 were overexpressed and extracted from the yeast membrane in enough amounts to continue with the purification step. The light subunit LAT2 was successfully purified but when the stability was analysed by size exclusion chromatography showed a clear profile of aggregation, concluding further studies. In contrast, the heavy subunit 4F2hc was stable after the exclusion chromatography for two days. The heterodimer 4F2hc/LAT2 proved to be stable after gel filtration analysis during one day. Thus, the heterodimer was significantly more stable than the light subunit alone, which allowed us to make an important statement. The catalytic subunit LAT2 needed their heavy subunit (i.e. 4F2hc) to increase the stability. This statement contrasted with the results for the heterodimer rBAT/b0+AT, in which was the heavy subunit rBAT the one who needed its light subunit b0+AT to a correct folding during its biogenesis (Bartoccioni et al. 2008; Rius et al. 2011). Functional studies with human heterodimer 4F2hc/LAT2 were set up to check the role of 4F2hc in the transport. Firstly the functionality of the heterodimer 4F2hc/LAT2 and the light subunit LAT2 in the living cell was checked successfully, meaning a correct folding at expression level. The apparent KM in both cases was the same, remaining unanswered the role of the heavy subunit 4F2hc in the transport function. Next, reconstitution in liposomes was carried out successfully for 4F2hc/LAT2 but not for LAT2, due to the high aggregation tendency. 4F2hc/LAT2 showed the typical overshoot for an amino acid transporter. To carry out the structural studies and due to the difficulty to maintain a stable soluble heterodimer, it was decided to carry out the technique of Single particle -negative staining (SP-NS) in the laboratory of Prof. Fotiadis in the University of Bern (Switzerland). The 3D model technique based on transmission electron microscopy (TEM) is relatively new and has been imposed for mammalian membrane proteins, allowing structural analysis with relatively small concentration of protein. The pure heterodimer was stained in a grid with uranyl formate at 0.75% (two drops optimized for 1 second, washing with water twice). This sample was analysed by transmission electron microscopy (TEM). Different images of projections in different orientations for 4F2hc/LAT2 were kept in a library of 11,000 picks. The refinement of the whole library allowed the 3D reconstruction of this protein by Mr. Meury. The model showed two asymmetric particles, one smaller, in which the crystal of the human ectodomain 4F2hc (Fort et al. 2007) fitted pretty well, and other bigger, which showed a black hole. Thus, the smaller particle was recognized as the heavy subunit, located on top of the light subunit. The resolution was 19 amstrongs, which was in the normal range for this method (from 16 amstrongs to 25 amstrongs). Discussion It was observed that the heavy subunit was located on top of the light subunit LAT2, and not in contact with the cell membrane as was firstly though. The size for the heavy subunit coincided with the existing 3D crystals of the human ectodomain which can fit quite accurately, always assuming the presence of the transmembrane segment in the 3D model. By contrast, the light subunit did not fit with the crystal structure of the prokaryotic homolog AdiC in the APA family (APC superfamily) (Gao et al. 2009) (Kowalczyk et al. 2011) due to the large amount of detergent surrounding this highly hydrophobic subunit in SP-NS method. In spite of that, when the size was compared with AdiC and Stet (a prokaryotic homolog in the LAT family with 30% of homology) studied in the same SP-NS method (Casagrande et al. 2009) the light subunit LAT2 coincided in size with its homologs, demonstrating that the increased volume was due to the detergent effect. Supporting the 4F2hc/LAT2 model, interaction studies with integrins (Feral et al 2005; Feral et al. 2007) and other membrane proteins involved in cell growth (ICAMI; Liu et al. 2003) and / or overexpressed in tumours (CD147/MCT1; Xu et al. 2005) suggest an effect in the transport function through the heavy subunit 4F2hc, which may be explained with an orientation on top of the light subunit and interaction by the external loops. New Evidences: Recently, the 4F2hc/LAT2 heterodimer model in which the heavy subunit is located on top of the light subunit has been corroborated by cross-linking experiments by Miss Helena Alvarez in our laboratory. This fact, allow us to imagine how interactions between both subunits will carry out also when the disulphide bridge is missing. Analyzing the external loops in AdiC atomic structure (the closest paradigm with LATs at present) is found that the external loop 3 and the external loop 4 are the longest (around 25 residues). These loops are even longer in LAT2, which make possible the interaction between both subunits being the separation of 16 amstrongs in the 3D model. Both loops have important roles in the transport cycle based in LeuT fold. The external loop 3 has an important movement in the transition from outward-open conformation to occluded-outward conformation due to the tilt of 40o of the transmembrane 6. The external loop 4, moves down to lid the substrate pathway during the transition from occluded-outward conformation to the occluded-inward conformation. Our new 3D model of a human heteromeric aminoacid transporter offers the opportunity to study new aspects about the role of the heavy subunit in the holotransporter. If the external loops join 4F2hc and LAT2 modulating the transport function in presence of other transmembrane proteins, or if 4F2hc only acts as a bollard of a multiproteic complex, rest to be studied in the future. / Els transportadors heteromèrics d'aminoàcids (HATS) de metazous estan formats per una subunitat pesada (4F2hc o rBAT) (N-glicoproteïna amb 1 segment transmembrana i un gran ectodomini en el seu extrem C-terminal), i una subunitat lleugera (d'entre 10) unides covalentment per un pont disulfur, fent aquests transportadors únics entre els metazous. En humans, 6 subunitats lleugeres es troben formant heterodímers amb 4F2hc (LAT1, LAT2, y+ LAT1, y + LAT2, XCT i asc1) i una (b0, + AT) amb rBAT. Els HATs tenen incidència en la salut, ja que mutacions en qualsevol de les subunitats ocasionen aminoacidúries (cistinúria, lisinúria amb intolerància a proteïnes), són receptors virals o estan sobre expressats en cèl • lules tumorals. El nostre grup va determinar l'estructura de l'ectodomini de 4F2hc a 2.1 Å (Fort J et al. 2007), i recentment ha resolt l'estructura d'un homòleg procariota (AdiC d' E. coli, amb ~17% d´homologia) de les subunitats lleugeres a 3.0 Å de resolució (Kowalczyk et al. 2011). Per contra no hi ha informació estructural sobre els holo-transportadors HAT. El present treball ens mostra el primer model estructural a 19 Å d'un transportador HAT humà, el transportador 4F2hc/LAT2. La importància de 4F2hc, a part de tenir un paper important en immunologia, es troba en la seva sobreexpressió en cèl•lules tumorals, el que la converteix en una important diana per a tractaments i desenvolupament de vacunes contra el càncer. El model ens mostra com en aquest transportador, l´ectodomini de 4F2hc està situat sobre LAT2, suggerint interacció amb els bucles extracel•lulars del transportador i nos sols interacció a través del pont disulfur del segment transmembrana com es pensava anteriorment. Aquesta nova topologia explica la necessitat i la importància de que l'ectodomini de 4F2hc formi part de l´heterodímer 4F2hc/LAT2 i presenta un escenari estructural per al paper "chaperone-like" de 4F2hc sobre les subunitats lleugeres.