Created by Michael Chi 

Mammalian fatty acid synthase or mFAS (PDB ID: 2VZ8) is a large multifunctional enzyme that catalyzes all steps of fatty acid synthesis; this is known as a type I fatty acid synthase (FAS) system (5). The protein of interest is isolated from the pig Sus scrofa and exists physiologically as an alpha-2 homodimer (5). In contrasts, plant and bacterial fatty acid biosynthesis is accomplished by monofunctional proteins in a dissociated type II FAS system (4). FAS is responsible for producing fatty acids that fulfill vital biological functions. Fatty acids are central constituents of biological membranes, serve as energy storage compounds, and act as second messengers or covalent modifiers regulating the localization of proteins. In addition human FAS has been associated with obesity and obesity-related diseases including diabetes and cardiovascular disorders, and the overexpression of FAS has been found in analysis of many forms of cancer cells and is correlated with tumor malignancy (3, 6, 7, 9) . Further research with FAS inhibitors have shown potential for weight reduction in animal models, though their exact mode of action has not been established, and have also demonstrated antitumor activity both in vivo and in vitro. FAS inhibitors therefore are important antiobesity and antineoplastic drug targets. 

Purification of mammalian fatty acid synthase yields a 270-kDa polypeptide chain that consists of all seven required domains (5). This is consistent according to ExPASy, a bioinformatics resource portal to many other scientific databases and tools, indicating the molecular weight of the protein to be 272251.98 Da, and its isoelectric point to be 5.98 (8). The peptide chain assembles into a intertwined homodimer approximating an "X" structure as the catalytically active form (5). In all organisms there is a conserved set of reactions for fatty acid biosynthesis (4, 5). The growing fatty acid is attached to acyl carrier protein (ACP) and is released by thioesterase (TE) after reaching 16 or 18 carbon atoms in length. The malonyl and acetyl transferases (MAT) are combined in mammals and is responsible for transferring the malonyl and acetyl group of malonyl-CoA and acetyl-CoA. The beta-ketoacyl synthase (KS) then catalyzes the decarboxylative condensation of the malonyl and acetyl moieties. The NADPH-dependent beta-ketoreductase (KR), dehydratase (DH), and NADPH-dependent enoylreductase (ER) are utilized to undergo a sequential beta-carbon modification pathway that results in a saturated acyl product elongated by two carbon units. The cycle is repeated for further elongation (Figure 1). 

The homodimer, mFAS, has two distinct portions: a lower condensing portion containing the KS and the MAT domains and an upper beta-carbon modifying portion containing the DH, ER, and KR domains (5). These two parts are loosely connected and form only tangential contacts (5). Two additional non-enzymatic domains pseudo-methyltransferase (psiME) and pseudo-ketoreductase (psi-KR) are located at the periphery of the upper modifying domains (5). The linear arrangement of the polypeptide chain is as followed: KS, LD, MAT, DH, psi-ME, psi-KR, ER, KR, ACP, and TE. The linear and folded arrangement can be seen in Figure 2a and 2b respectively. 

The two polypeptides dimerize through an extended contact area of 5400 angstrom² which involves more than 150 residues per chain (5). Homophilic interactions of the KS and ER domains, homophilic interactions between the double "hot dog" folds via a loop around Ser-941 of the DH domain, and the remaining interactions formed by the C-terminal part of the linker domain (LD) (residues 846 to 860 which are SAADFPSGSSCSSVA) with the KS domain of the other chain all contribute to dimerization interaction (5). The linker domains are regions of connection and are not catalytically active themselves. They are split into two regions comprised of residues 420-490 (SAADFPSGSSCSSVA) which form 2 short alpha-helices and residues 809-837 (SVNPNGLFPPVEFPAPRGTPLISPHIKWD) which form a three-stranded antiparallel beta-sheet (5). The LD act as an adaptor to prevent direct interaction between the KS and MAT domains. Further connection between the condensing and modifying portions is provided by residues 838 to 858 (HSQAWDVPSAADFPSGSSCSS) (5). In the modifying portion, the KR domain acts as central connector and interacts with the second hot dog domain of DH, ER, psi-ME, and psi-KR. Residues 1513 to 1518 (AFRHFP) and 1117 to 1123 (EKFCFTP) are beta-strand forming regions also buried in the modifying portion that contribute to the integrity of the dimer (5). 

There are seven catalytically active domains however only five have been resolved. The seven domains are as followed: KS, MAT, DH, ER, KR, ACP and TE. The two unresolved domains are ACP and TE and is suggested to lack resolution due to their inherent flexibility (4). There are two additional nonenzymatic domains psi-ME, an homolog of methyltransferase in polyketide synthases (PKS), and psi-KR an non-functional, truncated KR. 

The ketoacyl synthase domain (KS) of mFAS is a single domain highly specific for saturated acyl chains and does not accept beta-ketoacyl, beta-enoyl, or beta-hydroxyacyl substrates (5). This specificity is explained by a considerable constriction at the base of the phosphopantetheine binding pocket leading into the large acyl chain substrate binding tunnel that is formed between both KS domains of the two subunits (5). His-331, Cys-161, and His-293 are conserved active residues. Tyr-222, Met-205, and Phe-395 of one unit and Glu-136' of the second unit are responsible for the constriction of the binding site (5). There are approximately 5 alpha helices and 5 beta strands found in the active site of the domain.

The malonyl-acetyl transferase domain (MAT) has an uniquely equal specificity for acetyl-CoA and malonyl-CoA with broad specificity for malonyl-CoA derivatives (5). There is a conserved Ser-His dyad active site determined to be His-683 and Ser-581 that contribute to catalytic functions (5). In mFAS, Ser-682 is replaced with Phe, Glu-553 is replaced with Phe, and Glu-499 is replaced with and Met (5). These three substitutions create a more hydrophobic active site and is responsible for the double specificity unique to mFAS. Approximately 7 alpha helices and 6 beta strands were determined to be within the active site of the domain.

The ketoreductase domain (KR) is a short chain with a characteristic Rossmann fold and a substrate binding extension (residues 2072 to 2075 which could not be resolved) before the last alpha helix (5). Residues 1975 to 1990 (LRDAVLENQTPEFFQD) are part of the beta4/alpha4 loop in the active site that are disordered in the apo form but is stabilized with cofactor binding (5). This domain is NADPH-dependent and after binding to the substrate, the interaction between Met-1973 and Lys-1995 help stabilize the structure (5).

The dehydratase domain (DH) adopts a pseudo-dimeric hot dog fold. The active site is formed by residues His-878 from the N-terminal hot dog fold and Asp-1033 and His-1037 from the C-terminal fold (5). His-878 and Asp-1033 is suggested to participate in substrate protonation and deprotonation (4, 5). His-1037 is only found in chicken and pig FAS while other mammals utilize Glu (5). His-1037 is oriented with Asp-1033 at hydrogen-bonding distance indicating a stabilizing effect (5). There is a hydrophobic substrate binding tunnel that stretches through the C-terminal fold. In plants and bacteria there are two active sites within this tunnel; in mammals only the first site is active and the second site is truncated (5). Two beta strands and one alpha helix was approximated in the DH domain.

The enoylreductase domain (ER) in contrast to all other domains has a unique mammalian structure. The mFAS ER contains two subdomains, a nucleotide binding Rossmann fold (residues 1651 to 1794), and a substrate binding portion (residues 1530 to 1650 and 1795 to 1858) (5). It binds the NADP+ cofactor in an open extended conformation between the two subdomains. Lys-1771 and Asp-1797 are considered to be candidate donor residues for substrate protonation after hydride transfer from NADPH (5). The active site of ER is located in a narrow crevice created in part by the bound nucleotide cofactor. The ER active site contained 3 alpha helices and 8 beta strands.

The acyl-carrier protein domain (ACP) has an anchor point at Glu-2113 in the center of the upper portion of the lateral clefts of mFAS (5). Besides the anchor point further resolution was not obtained. The linker to ACP is 12 to 14 amino acids, corresponding to a maximum length of approximately 40Å (5). The linker connecting ACP with thioesterase (TE) is 23 to 26 residues and approximately 80Å (5). Whereas the motion of fungal ACP is constrained by double-tethering, no second anchor point is apparent for mFAS ACP. Secondary and tertiary  structure could not be established due to the flexibility of the domain during crystalization. 

A query was run in two algorithmic protein databases, the Position-Specific Integrated Basic Local Assignment Search Tool (PSI-BLAST) and the Dali server, to find a comparison protein for mFAS and resulted in the human fatty acid synthase KS-MAT didomain (PDB ID: 3HHD). The PSI-BLAST runs a query based on primary structure similarity. The program incorporates gaps in the amino acid sequence to produce sequences of best fit when comparing the subject protein with that of the query proteins. The generated output is an E value which is indicative of high similarity if the value falls below 0.05. The E-value generated by PSI-BLAST for human FAS KS-MAT domain is 0 (1). This indicates the primary sequence of the two proteins have high similarity and may be identical, the obvious distinction between the two is that one is from pigs and the other human. The Dali server runs a query based on tertiary structure similarity via a sum-of-pairs method to calculate intermolecular distances. The output is a Z-score which indicates significant similarity if the value is greater than 2. The Z-score given by the Dali server for the comparison protein was 68 indicating high tertiary similarity between the two (2). The biggest difference here is the absence of the remaining domains. The human FAS KS-MAT protein is significant for a number of clinical reasons mentioned earlier. Due to the size and flexibility of the large mFAS protein, drug screening and other experimental assays are difficult (7). The human protein utilizes only the KS-MAT domains which allow for more rapid inhibition screening assays and a more focused study of the structure (7). Inhibition of FAS using cerulenin or synthetic FAS inhibitors such as C75 has shown reduction in food intake and profound but reversible weight loss (9). In addition human breast carcinoma cells were purified and found the overexpression of mFAS (3). The use of inhibitors demonstrated growth inhibition to normal levels (3). In the past decade, FAS inhibition as a drug target has been studied to provide novel therapeutic opportunities for metabolically treating and preventing cancer. FAS activity has shown a network of pathways that regulate several aspects of tumor function including metabolism, cell survival and proliferation, DNA replication and transcription, and protein degradation (6). Oncogenic effects thus have yet to be clearly established and further research is required for a concrete conclusion.