Mammalian fatty acid synthase in complex with NADP (PDB ID: 2VZ9) from Sus scrofa

Created by: Kaiwen Su

            Mammalian fatty acid synthase (mFAS) complexed with NADP (PDB ID: 2VZ9) is a large multienzyme transferase that catalyzes all steps of fatty acid synthesis and shuttles the essential substrates needed for elongation. Its crystal structure contains five catalytic domains, two nonenzymatic domains and a shuttle protein domain. Overexpression in FAS can be linked to development of cancer cells (1). Activity of fatty acid synthase (FAS) in mammals is often low in most body tissue. It is the sole protein in the human genome that is capable of synthesizing long-chain fatty acids from acetyl-CoA, malonyl-CoA and nicotinamide adenine dinucleotide phosphate (NADPH). However, its overexpression is correlated with tumor malignancy (2, 3). FAS inhibitors have shown progress for anti-tumor treatment and FAS has emerged as an important drug target for treatment of human cancer. FAS inhibition uses cerulenin or C75 to induce apoptosis in human cancer cells both in vitro and in vivo, implying the dependence of cancer cell survival on FAS activity (3, 4). Consequently, it is important for scientists to continue to study mFAS to find a way to potentially treat cancer.

            Studying the structure of mFAS is an important step in fully understanding the function of the protein and potential treatment for cancer (1). The mFAS assembles into an intertwined dimer, resembling an “X” shape. The structure mainly contains alpha helices and beta sheets, and has evenly spread polarity groups. The mFAS is segregated into a lower condensing portion, containing the ketoacyl synthase (KS) and malonyl-acetyl transferase (MAT) domains, and an upper portion, including dehydratase (DH), NADPH-dependent enoyl reductase (ER), and ketoreductase (KR) responsible for β-carbon modification. Two nonenzymatic domains (ΨME and ΨKR) are located at the periphery of the modification. The two polypeptides (condensing and modifying) dimerize through an extended contact area which involves more than 150 residues per chain. Linker between KS and MAT domains is composed of amino acids 420 to 490.  The linker includes two short α-helices facing the KS and a three-stranded antiparallel β-sheet on the MAT side. Connection between the condensing and modifying parts is provided by residues 838 to 858 between the KS-MAT linker domain and DH domain. KR domain acts as central connecter for modifying part of mFAS, and interacts with DH, ER and nonenzymatic domains (1). ExPASy is a bioinformatics resource portal that provides access to scientific databases and software tools. One of the tools is the computation of theoretical isoelectric point and molecular weight for user entered sequences. According to the database, the mFAS protein has a total molecular weight of 272,250 Da and an isoelectric point of 5.98 (5).

            The noncatalytic domains contains ΨKR and ΨME. ΨKR domain has KR character, but it has lost the ability to bind NADPH because of extensive truncation of its core. Consequently, it functions mainly to support the active site of catalytic KR domain (1). ΨME domain is structurally closely related to S-adenosyl-methionine (SAM)-dependent methyltransferases. Its core consists of a seven-stranded β-sheet with three helices on each side and a C-terminal strand in anti-parallel orientation (6). This domain contains a short linker (residues 1125 to 1224) which leads into the adjacent ΨKR fold. In mFAS, prevention of cofactor binding results in absence of methyltransferase activity and products in systems. Thus, ΨME domain represents an inactive version of a functional enzyme in a precursor of mFAS (1).

            In mammals, FAS has only a single KS domain for all steps of fatty acid elongation. It is highly specific for saturated acyl chains and discriminates on basis of β-carbon atom status. It does not accept β-ketoacyl, β-enoyl- or β-hydroxyacyl substrates. A possible explanation for this phenomenon is found in structure organization of domain’s substrate-binding channel. As KS associate with FAS, the substrate channel divides into halves, corresponding to the substrate-binding and phosphopantetheine-binding regions. The amino acid sequence of the substrate-binding region is important for defining substrate specificity (7, 8). For phosphopantetheine binding pocket, the number of residues at the narrow tunnel is highly conserved (1).

            The MAT domain for mFAS has slightly different orientations of two subdomains that are selected by crystal-packing interactions. It has rather broad specificity for malonyl-CoA derivative and uses both acetyl-CoA and malonyl-CoA with equal efficiency (7). Phe-682, Phe-553 and Met-499 create a hydrophobic active site. The two phenylalanines form a hydrophobic cavity which allows Met-499 to flip onto the methyl group of an acetyl substrate. The dual specificity of MAT results from a conserved arginine for salt-bridging malonyl substrate and a hydrophobic nature of active site (1).

            The NADPH-dependent KR domain acts as a central connector for the modifying part of mFAS and interacts with the DH, ER domains (1). It interacts with the second subdomain of DH, forming an 800Ų interface. The contact between KR and ER domain extends over an area of 400 Ų. The KR domain has single-domain proteins that have a Rossmann fold and a substrate binding extension inserted before the last helix. The arrangement of residues in the active site of the KR domain is consistent with a proton-relay mechanism. Loops (residues 1975 to 1990) in the vicinity of active site are disordered and become stabilized by interactions of Met-1973 with active site Lys-1995 and substrate binding extension upon cofactor binding. The direction of substrate entry into the active site can be inferred from the stereospecificity of KR domain (1).

DH domain in mFAS adopts a pseudodimeric fold. Pseudo-dimeric DH domain harbors a single active site and substrate binding tunnel with an open end (1). Each pseudodimer contains a single active site formed by His-878 from N-terminal, and Asp-1033 and His-1037 from C-terminal. His-1037 is only present in chicken and pig FAS whereas the corresponding amino acid in all other mFAS is glutamine. The histidine (or glutamine) at this position is oriented toward Asp-1033 at hydrogen-bonding distance, indicating a stabilization function. The hydrophobic substrate binding tunnel starts at pseudodimer interface, stretches the C-terminal domain and has an open end toward top of FAS assembly. In mFAS, the second catalytic site is inactive because the loop is reduced to a short turn and 30 residues less at the N-terminus, resulting in absence of hydrophobic tunnel (1).

The mFAS ER is composed of medium-chain dehydrogenases/reductases (MDRs) (1). It contains two subdomains, a nucleotide binding Rossmann-fold (residues 1651-1794) and a substrate binding part (residues 1530 to 1650 and 1795 to 1858). It binds the NADP+ cofactor in open extended conformation. Lys-1771 and Asp-1797 are candidate donors for substrate protonation after hydride transfer from NADPH because they are close to the hydride donor nicotinamide C4. The two residues are strictly conserved in mFAS. The active site of ER is located in the narrow opening created by the bound cofactor. Substrate entry occurs through a tunnel along cofactor toward the nicotinamide ring. The tunnel continues toward back of ER domain where the long acyl chains are allowed to exit (1).

            Substrates are shuttled by acyl carrier protein (ACP) domains to entry sites (1). The entry sites are grouped into two lateral clefts. Entry sites of MAT, DH, and ER are oriented toward one face of mFAS, and those of KS and KR are oriented toward the other face. Anchor point of ACP at residue 2113 (not visualizable in structure) is located at the center of the upper portion of mFAS. During elongation cycle, the ACP is loaded with substrates at the lateral MAT domain. ACP then delivers substrates to KS pocket on opposite side of mFAS. After condensation at the KS, ACP reaches the KR on the same side before approaching DH domain. Then, ACP proceeds to ER for reduction and delivers fully saturated substrate to KS active site before reloaded at MAT site for the next cycle (1).

            To compare mammalian FAS’s structure with similar proteins in other species, the mFAS’s structure was entered into PSI-BLAST and Dali Server. The PSI-BLAST compares proteins’ primary structures and assigns an E-value to each protein where 0 is the best score, indicating similar primary structures. The E-value is assigned based on gaps of amino acids that the query does not share with the subject. It decreases with total sequence homology and increases with presence of gaps. Similarly, Dali Server compares secondary structures and assign a Z-score where 65 is the best score, indicating similar secondary structures. The comparison protein must have an E-value below 0.05 and Z-score above 2.0 (9, 10).

            The mFAS is compared to thioesterase (TE) domain of human fatty acid synthase (PDB ID: 1XKT) since the comparison protein has a low E-value of 2x10-153 (9, 10). No corresponding Z-score is found using Dali Server. Comparisons found using PSI-BLAST do not correspond to the comparisons found on Dali Server (using a different but highly similar structure). The TE domain of human fatty acid synthase differs from mFAS since it exists as a monomer. The mFAS exists as a homo 2-mer-A2 with two subunits. The entire structure of TE domain of human FAS is two subdomains, made up of 9 α-helices and 8 β-strands. The human TE domain is close to the ER domain of the mFAS based on overlap. Since TE domain of human FAS is a monomer, it has no symmetry. The mFAS of interest, however, has a symmetry with a cyclic C2. Both FAS catalyzes formation of palmitic acid as its major product and show no significant activity toward fatty acyl chain lengths shorter than 14 carbon atoms. TE domain of human FAS revealed the presence of a hydrophobic groove with a distal pocket that constitutes a candidate substrate binding site. The structure sets the identity of Asp residue located at the proximal end of the groove. In ER domain of mFAS, Asp-1797 is a candidate proton donor for substrate (11, 12). Comparison between TE domain of human FAS and mFAS demonstrates how differences in structure can result in different functions of proteins. The mFAS complexed with NADPH is a transferase, and the TE domain of human FAS is a hydroxylase (1, 11).