The AMP-activated protein kinase (AMPK) is characterized by its ability to bind to AMP, which enables it to adjust enzymatic activity by sensing the cellular energy status and maintain the balance between ATP production and consumption in eukaryotic cells. It also has important roles in the regulation of cell growth and proliferation, and in the establishment and maintenance of cell polarity. These important functions have rendered AMPK an important drug target for obesity, type 2 diabetes and cancer treatments. However, the regulatory mechanism of AMPK activity by AMP binding remains unsolved. Here we report the crystal structures of an unphosphorylated fragment of the AMPK alpha-subunit (KD-AID) from Schizosaccharomyces pombe that contains both the catalytic kinase domain and an autoinhibitory domain (AID), and of a phosphorylated kinase domain from Saccharomyces cerevisiae (Snf1-pKD). The AID binds, from the 'backside', to the hinge region of its kinase domain, forming contacts with both amino-terminal and carboxy-terminal lobes. Structural analyses indicate that AID binding might constrain the mobility of helix alphaC, hence resulting in an autoinhibited KD-AID with much lower kinase activity than that of the kinase domain alone. AMP activates AMPK both allosterically and by inhibiting dephosphorylation. Further in vitro kinetic studies demonstrate that disruption of the KD-AID interface reverses the autoinhibition and these AMPK heterotrimeric mutants no longer respond to the change in AMP concentration. The structural and biochemical data have shown the primary mechanism of AMPK autoinhibition and suggest a conformational switch model for AMPK activation by AMP.

LAT1 (SLC7A5) is one of the representative light chain proteins of heteromeric amino acid transporters, forming a heterodimer with its heavy chain partner 4F2hc (SLC3A2). LAT1 is overexpressed in many types of tumors and mediates the transfer of drugs and hormones across the blood-brain barrier. Thus, LAT1 is considered as a drug target for cancer treatment and may be exploited for drug delivery into the brain. Here, we synthesized three potent inhibitors of human LAT1, which inhibit transport of leucine with IC50 values between 100 and 250 nM, and solved the cryo-EM structures of the corresponding LAT1-4F2hc complexes with these inhibitors bound at resolution of up to 2.7 or 2.8 Å. The protein assumes an outward-facing occluded conformation, with the inhibitors bound in the classical substrate binding pocket, but with their tails wedged between the substrate binding site and TM10 of LAT1. We also solved the complex structure of LAT1-4F2hc with 3,5-diiodo-l-tyrosine (Diiodo-Tyr) at 3.4 Å overall resolution, which revealed a different inhibition mechanism and might represent an intermediate conformation between the outward-facing occluded state mentioned above and the outward-open state. To our knowledge, this is the first time that the outward-facing conformation is revealed for the HAT family. Our results unveil more important insights into the working mechanisms of HATs and provide a structural basis for future drug design.

The human SLC26 transporter family exhibits various transport characteristics, and family member SLC26A9 performs multiple roles, including acting as Cl-/HCO 3 exchangers, Clchannels, and Na + transporters. Some mutations of SLC26A9 are correlated with abnormalities in respiration and digestion systems. As a potential target colocalizing with CFTR in cystic fibrosis patients, SLC26A9 is of great value in drug development. Here, we present a cryo-EM structure of the human SLC26A9 dimer at 2.6 Å resolution. A segment at the C-terminal end is bound to the entry of the intracellular vestibule of the putative transport pathway, which has been proven by electrophysiological experiments to be a gating modulator. Multiple chloride and sodium ions are resolved in the high-resolution structure, identifying novel ion-binding pockets for the first time. Together, our structure takes important steps in elucidating the structural features and regulatory mechanism of SLC26A9, with potential significance in the treatment of cystic fibrosis.

In Gram-negative bacteria, phospholipids are major components of the inner membrane and the inner leaflet of the outer membrane, playing an essential role in forming the unique dual-membrane barrier to exclude the entry of most antibiotics. Understanding the mechanisms of phospholipid translocation between the inner and outer membrane represents one of the major challenges surrounding bacterial phospholipid homeostasis. The conserved MlaFEDB complex in the inner membrane functions as an ABC transporter to drive the translocation of phospholipids between the inner membrane and the periplasmic protein MlaC. However, the mechanism of phospholipid translocation remains elusive. Here we determined three cryo-EM structures of MlaFEDB from Escherichia coli in its nucleotide-free and ATP-bound conformations, and performed extensive functional studies to verify and extend our findings from structural analyses. Our work reveals unique structural features of the entire MlaFEDB complex, six wellresolved phospholipids in three distinct cavities, and large-scale conformational changes upon ATP binding. Together, these findings define the cycle of structural rearrangement of MlaFEDB in action, and suggest that MlaFEDB uses an extrusion mechanism to extract and release phospholipids through the central translocation cavity.

• The ABC transporter, ABCC4, localizes to the platelet plasma membrane and regulates aggregation by exporting cAMP and antithrombotic drugs.Controlling the activation of platelets is a key strategy to mitigate cardiovascular disease. Previous studies have suggested that the ATP-binding cassette (ABC) transporter, ABCC4, functions in platelet-dense granules. Using plasma membrane biotinylation and superresolution microscopy, we demonstrate that ABCC4 is primarily expressed on the plasma membrane of both mouse and human platelets. Platelets lacking ABCC4 have unchanged dense-granule function, number, and volume, but harbor a selective impairment in collagen-induced aggregation. Accordingly, Abcc4 knockout (KO) platelet attachment to a collagen substratum was also faulty and associated with elevated intracellular cyclic AMP (cAMP) and reduced plasma membrane localization of the major collagen receptor, GPVI. In the ferric-chloride vasculature injury model, Abcc4 KO mice exhibited markedly impaired thrombus formation. The attenuation of platelet aggregation by the phosphodiesterase inhibitor EHNA (a non-ABCC4 substrate), when combined with Abcc4 deficiency, illustrated a crucial functional interaction between phosphodiesterases and ABCC4. This was extended in vivo where EHNA dramatically prolonged the bleeding time, but only in Abcc4 KO mice. Further, we demonstrated in human platelets that ABCC4 inhibition, when coupled with phosphodiesterase inhibition, strongly impaired platelet aggregation. These findings have important clinical implications because they directly highlight an important relationship between ABCC4 transporter function and phosphodiesterases in accounting for the cAMPdirected activity of antithrombotic agents. (Blood. 2015;126(20):2307-2319