Acetyl-CoA Carboxylase 2 - Biotin Carboxylase Domain
Created by Jason Liao
Acetyl-CoA carboxylase 2 (ACC2) (PDB ID: 3JRX) in Homo sapiens has a crucial roles in fatty acid synthesis and fatty acid oxidation. ACC2 catalyzes the ATP-dependent carboxylation of acetyl-CoA to malonyl-CoA with carbonate serving as the CO2 source. Eukaryotes possess two isoforms of ACC: ACC1 and ACC2. Both isoforms have virtually identical sequences, but ACC2 contains an additional N-terminal mitochondrial targeting domain, so it is localized to mitochondria of liver, heart, and skeletal muscle tissue. ACC2 produces malonyl-CoA that inhibits carnitine palmitoyltransferase (CPT 1), a mitochondrial membrane protein that shuttles fatty acids into the mitochondrial matrix for β-fatty acid oxidation (Tong, 2005). Starting from the N-terminus, ACC2 contains three distinct domains: the biotin carboxylase (BC), the biotin carboxyl carrier protein (BCCP), and the carboxyl transferase (CT) domain. The BC domain catalyzes the ATP-coupled carboxylation of biotin. The CT domain transfers the carboxyl group from biotin to acetyl-CoA which produces malonyl-CoA. The BCCP domain is covalently bonded to biotin and translocates carboxylated biotin between the BC and CT domains (Cho Y. , et al., 2008). ACCs have gained widespread attention as therapeutic targets for obesity and diabetes because malonyl-CoA inhibits fatty acid oxidation and is necessary for fatty acid synthesis. ACC2 knockout mice demonstrate reduced malonyl-CoA levels, continuous fatty acid oxidation, decreased de novo lipogenesis, and subsequent protection from obesity and insulin resistance (Abu-Elheiga, Matzuk, Abo-Hashema, & Wakil, 2001).This paper primarily focuses on the structure of the human ACC2 BC domain.
The ACC2 BC domain is regulated through acute allosteric inhibition. Monomeric ACC is catalytically inactive. The active form exists as a long linear polymer of 20-40 ACC polypeptides. Eukaryotic ACC2 is inhibited by reversible phosphorylation at Ser-222 (PDB ID: 3JRW) by AMP-activated protein kinase (AMPK). The phosphorylation prevents the polymerization of ACC polypeptides. The exact mechanism of human ACC polymerization is unknown, but it appears that polymerization occurs in a head-to-head, tail-to-tail manner by non-covalent associations.
Current literature characterizes the human ACC2 BC domain (residues 217-775) as a crystal structure separate from the other ACC2 CT and BCCP domains. The 587 residue protein has a molecular weight of 55407.37 Daltons and a pI is 5.59 (Expasy, 2012). The protein contains 16 α-helices, 3 3/10-helices, and 17 β-strands. Most β-strands are antiparallel to each other. The BC domain contains four subdomains, the A- (residues 217-375), B- (residues 436-496), C-domain (residues 497-775), and an AB linker (residues 376-435) that connects the A- and B-domains.
The BC domain is a cylindrical peptide with a hinged lid. It possesses an outer shell of aliphatic α-helices and nonpolar β-sheets in the interior. The exterior aliphatic character allows ACC2 to effectively function in the cytosol of cells. The ATP binding site is located between the B- and C-domains, and the site of polymerization is situated between the A- and C-domains. The B-domain has flexible connecting loops allowing it to open and close like a hinged lid
ACC2 polymerization is inhibited by phosphorylation of Ser-222. Unphosphorylated Ser-222 is in a disordered loop and does not appear in the crystal structure, but phosphorylated Ser-222 (pSer-222) and its four adjacent residues (Ser-220, Met-221, Gly-223, Leu-224) appear in the phosphorylated crystal structure. pSer-222 forms a strong ionic interaction with the Arg-277 of the BC domain polymerization site. Arg-277 is stabilized by an ionic interaction with Glu-671. The five residue segment containing pSer-222 binds and inhibits polymerization at the active site. AMPK phosphorylation at Ser-222 has demonstrated considerable inhibition of polymerization and catalytic activity (Cho Y. , et al., 2009).
Polymerization inhibition has also been achieved by soraphen A binding to the BC domain (PDB ID: 3JRW). Soraphen A and the inhibitory five residue pSer-222 segment have similar chemical shapes. Like pSer-222, soraphen A also binds to the BC polymerization site at the Arg-277 residue. Soraphen A is also stabilized by ionic interactions with Glu-593, Lys-274, Ser-278 and hydrophobic interactions with Trp-681 and Met-594. Trp-681 and Met-594 undergo drastic conformational changes to accommodate soraphen A’s bulky carbon backbone. Polymerization is inhibited by soraphen A binding, suggesting that AMPK phosphorylation and soraphen A inhibition occur by the same mechanism (Cho Y. , et al., 2009).
In both phosphorylative and ligand binding inhibition mechanisms, Arg-277 and Glu-671 play crucial roles. A normal unphosphorylated BC domain will easily polymerize, but polymerization does not occur when Arg-277 or Glu-671 is converted to alanine, suggesting the two residues are essential for pSer-222 or soraphen A binding (Cho Y. , et al., 2009).
To find a suitable comparison protein, queries were made in the BLAST and Dali databases. The goal of the BLAST server is to find proteins that have similar primary structures. The position-specific-iterated basic local alignment search tool (PSI-BLAST) compares the primary structure of the queried protein to all of the proteins in the database. It then assigns an E value to other proteins that have similar sequences. A lower E value corresponds to greater sequence homology. A PSI-BLAST query of the BC domain of ACC2 revealed that ACC2 BC domain from Saccharomyces cerevisiae (yeast), ACC from Brassica juncea and ACC1 BC domain from Homo sapiens gave 0.0 E values (Altschul, et al., 2005). The Dali database compares the tertiary structures of two proteins and assigns a Z value. A higher Z value corresponds to greater tertiary structure similarity. A Dali query revealed that the yeast ACC2 BC domain (PDB ID: 1W96) and human ACC1 BC domain gave Z scores of 52.0 and 60.3, respectively (Holm & Rosenstrom, 2010). The yeast ACC2 BC domain has a similar structure to the human analog. The yeast domain catalyzes the same reaction, and it shares 67% sequence homology with the human ACC2 BC domain. The yeast domain contains 21 α-helices and 20 β-strands. It has a virtually identical ATP-grasp fold and polymerization/inhibition site (Shen, Volrath, Weatherly, Elich, & Tong, 2004). Although human ACC2 polymerization is inhibited by AMPK phosphorylation at Ser-222, yeast BC does not have an equivalent serine residue. However, yeast can still efficiently bind soraphen A via stabilizing ionic interactions with Glu-392, Lys-73, Ser-77, Arg-76, and Glu-477. The human analogs of these residues are Glu-593, Lys-274, Ser-278, Arg-277, and Glu-671. Although Trp-681 and Met-594 have to undergo a conformation change to bind soraphen A in human ACC2 BC, the yeast BC domain already assumes a sterically favorable binding pocket (Cho Y. , et al., 2009).