User:Drew C. MacKellar/Notebook/ARPA-E/2011/06/28

{| width="800"
 * style="background-color: #EEE"|[[Image:owwnotebook_icon.png|128px]] Project name
 * style="background-color: #F2F2F2" align="center"|  |Main project page
 * style="background-color: #F2F2F2" align="center"|  |Main project page


 * colspan="2"|
 * colspan="2"|

KAS Engineering
1. Primer design for mutating FabB to produce dodecanoate:

I tried another approach within Swiss PDB Viewer, using FabF as a test case: I opened the file containing the dimer, made both dodecanoate molecules part of the same polypeptide chain, then selected residues within 6 angstroms of the 12th carbon of one of them, colored them blue, and the nearby residues of the other dodecanoate molecule, coloring them red. This allowed me to see which nearby residues were from which monomer. I hid all other residues, and labeled the ones visible. Then, to exclude those near the proximal portion of dodecanoate, I selected residues within 6 angstroms of the Threonine 164 (the visible residue farthest from the distal carbon of dodecanoate), and labeled them yellow, then hid them. The results show 17 residues lining the binding pocket, including the 5 chosen for mutation.

FabF binding pocket distal residues.

I’ll try the same thing with FabB:

This covers the main residues chosen by the previous method. The PDB file actually designates the binding sites as G106, A162, S163, M197, F291, G391, F392, and (on the opposite monomer) M138. Of course, these don’t necessarily designate the best residues to mutate to suit our purposes. Still, it’s a place to start; I’ll design primers to mutate the closest residues (110, 113, 134, 137, 197, 200).

Olsen 2001 describes the structure of FabB, and lavishes the most attention on the active site cysteine, histidine, histidine catalytic triad. They do mention the acyl binding pocket in the discussion, however, although it doesn’t shed a whole lot of light on which residues to mutate. It suggests that the unique ability of FabB to elongate C10:1 substrates is due to recognition of the incoming acyl-ACP, rather than by the conformation of the binding pocket. It also says that the binding pocket has no obvious limitation by space, and that both FabB & FabF can elongate substrates beyond C18, so it’s unclear what confers the specificity of the accepted substrates (perhaps thioesterase activity? Could there be a different thioesterase or condensation enzyme that’s expressed under the glycerol starvation conditions mentioned?):

“Substrate Specificity of KAS I and KAS II

The synthesis of membrane fatty acids in E. coli differs from that in plastids in that a double bond is built into the growing chain during elongation in E. coli, whereas double bonds are inserted into the finished chain in plastids. In E. coli, the branch between synthesis of saturated and unsaturated fatty acids occurs at a carbon chain with ten carbon atoms. The major differences in the substrate profiles of KAS I and KAS II are that only the former allegedly elongates a cis-3 decanoyl (C10:1) primer, while only the latter elongates a cis-9 hexadecanoyl (C16:1) primer. Both carry out all elongations be- tween C4 and C16 in the saturated pathway [8, 9]. The noted differences have been presumed to have a structural basis. We had anticipated clues to solving this puzzle from analyses of the structure of the KAS I–primer substrate complexes, especially with respect to the initiation of the unsaturated pathway by KAS I when it elongates C10:1. The architecture of the binding pockets of the two enzymes, however, is identical in the area proximal to the nucleophilic cysteine residue that binds the substrates. This is where one would expect KAS I and not KAS II to be able to harbor the double bond. Both saturated decanoyl and dodecanoyl substrates have an E configuration in this position, which can be readily modeled as a Z bond to resemble the cis bond without encountering any steric hindrance in either isozyme. This points to the possibility that acceptance or rejection of the C10:1 fatty acyl chain takes place before transfer from ACP. We have earlier suggested that the patch of positively charged side chains that delineate the active site entrance is involved in the specific ACP recognition [19].

That only KAS II elongates C16:1 to C18:1 is part of a more general question with respect to substrate specificity, namely, how the maximum length of a fatty acid that can fit into a binding pocket is determined. While preliminary soaking experiments with C16-CoA did not yield usable substrate–KAS I complexes, we have previously reported that KAS I can accept 16 carbon acyl chains as readily as 14 carbon ones in a transfer assay that measures the first step in the Claisen condensation [29]. This is in accord with the observation that E. coli, under conditions of glycerol starvation, synthesizes prominent amounts of C20:1, C18, and C20 as well as traces of C22 that are found among the free fatty acids [40]. This infers that either KAS I and/or KAS II has the ability to carry out as many as three additional elongation steps. The KAS I binding pocket seems to be effectively blocked at the dimer interface by the charged/ polar residues described. Contrarily, KAS II seems to be able to harbor longer substrates. Its binding pocket is hydrophobic and does not have an obvious end where the binding pockets meet on the dimer interface. Very recently, the replacement of two residues lining the substrate pocket of KAS II, one with a bulky phenylalanine, the other with a methionine with a longer hydrophobic chain, was shown to reduce dramatically the size of fatty acyl chains that could be elongated. This confirmed modeling predictions that the substrate binding pocket would be markedly reduced in size [41]. While this demonstrates very beautifully that such an approach may enable the plant oil biotechnologists to control the length of fatty acids synthesized in oil seeds, it does not help solve the puzzle of the structural basis of the KAS I and KAS II substrate specificity differences.”

2. A thought about detecting dodecanoate production:

It has occurred to me that another option for detecting FA production in cells that could be incorporated into a screen is bioluminescence. Luciferase enzymes use aldehydes as a substrate, together with an electron donor like FMNH2. Most work with specific chain length aldehydes, and some are supposed to work with dodecanal. One of these enzymes comes from Photobacterium phosphoreum, at least according to Meighen 1991, but Watanabe 1972 suggests longer chains work better with this enzyme. Of course, we could do a screen based on decreasing luminescence, but there would probably be a lot of other reasons that this would decrease than just shortening of the FAs produced.


 * }