The Scaffold Idea

The steric proximity of enzymes can cause higher yield and production rates of final goods in metabolic processes. Different catalysation cascades can be found in nature that show intermediates that are transported to their next destination through channels or enzymes that are located closely to each other (Huang et al. 2001). This leads to an increased local concentration of metabolites near the enzymes und is especially useful in case of unstable or toxic intermediates of a reaction. To make use of such possibilities in synthetic biology, the Keasling Lab developed a construction of an easily and universally applicable protein scaffold. Its potency was proven in two examples with an up to 77-fold increase in yield (Dueber et al. 2009). The scaffold consists of three protein binding domains that may be linked by any number and in any order (figure 1). The scaffold can interlink with enzymes through domain-specific tags that are added to the catalytic proteins. The scaffold offers regulatory options for the stoichiometry of enzymes in steric proximity and may thereby counteract bottlenecks and excessive production of intermediates.

Figure 1: The scaffold molecule consists of three different interaction-domains, that can be freely combined (x, y, z). Chosen enzymes get attached to the scaffold through a tag complementary to the specific desired binding domains. This arrangement creates a production chain which generates higher yields.

As far as we know, there have not been any successful attempts to clarify the structure of the scaffold molecule. NMR-structures of the separate parts (binding domains: GBD, PDZ, SH3) are known and were used to predict a homology-modeling based structure using PHYRE2 (figure 2).

Figure 2: 3D-structure of the protein scaffold, created with PHYRE2 based on structure-homology-modeling.

The use of a scaffold protein might lead to a significant increase of yield of the anthocyanidin that we successfully produced in E. coli in our 2014 iGEM project. Since especially DFR of the enzymatic cascade shows only a relatively small turnover rate and also catalyzes several side reactions, it might be the limiting step for the overall production rate of our pathway. For this reason, we tried to establish a method to produce and characterize the scaffold for a further usage as a regulatory device and an improvement of our metabolic network.  

We tried to establish a system for the covalent crosslinking of our scaffold using the thiole group of a cysteine. Cysteine is not present in the sequence of the scaffold protein and therefore enables regiospecific coupling reactions. Cysteines belong to the naturally occurring amino acids and are of great significance for protein folding. Two cysteins can form a disulfide bond that leads to structural constraints while the folding-process of a protein takes place. Cysteines are a popular target for chemical modifications of proteins and bioconjugation due to their unique reactivity among amino acids. Generally, cysteine is one of the most used amino acid residue for protein modifications. Proteins can be provided with a virtually unlimited variety of modifications using coupling reagents as maleimids, haloacetamids etc. A frequent utilization of such free cysteine residues is protein immobilization on resin particles or the labeling of proteins with fluorophors. Bioconjugation of proteins with fluorophors is of great use for binding studies. Mainly, proteins with only one accessible cysteine residue are of interest for coupling reactions since a regioselective reaction is not possible if multiple cystein residues are present in a reactive and accessible form.

Figure 3: Mechanism of an immobilization reaction. A free cysteine residue of a protein reacts with the maleimide function on the surface of a resin particle.

We improved already existing scaffold domain Biobricks for the construction of scaffold proteins by introducing BglII and BamHI restriction sites flanking the domain sequence. Additionally, C-terminal linker domains consisting of glycine and serine residues were added to the scaffold domain. New scaffold sequences can be constructed by standard cloning approaches (see Figure 2). As usual, a backbone is ligated with an insert to create the desired sequence. For example a new scaffold protein consisting of two domains can be constructed by ligating a backbone vector including the desired N-terminal scaffold domain with an insert containing the desired C-terminal domain. For this, the backbone has to be digested with the restriction enzymes BamHI and PstI cleaving the plasmid downstream of the first domain. In contrast, the insert has to be extracted from its vector by digestion with BglII and PstI. The overlap sequcences of the BglII and BamHI restriction sites are complementary. Thus, the insert can be ligated behind the first domain into the backbone. The scar sequence resulting from a combination of a BglII with BamHI restriction site cannot be recognized by nether of the enzymes. Therefore, a single ligation creates a new scaffold BioBrick immediately, which is again flanked by a BglII and BamHI sequence. Of course, more sophisticated scaffold BioBricks can therefore be constructed from composite BioBricks containing more than one domain or again by iterative cloning of single domains behind an initial domain.

Figure 2:

Cloning scheme for the construction of scaffold proteins. To assemble domains for the construction of a new scaffold protein, the backbone containing the N-terminal domain(s) can be digested with BamHI and PstI and the C-terminal domain(s) can be cut from the plasmid with BglII and PstI. The ligation of the two DNA fragments creates a new BioBrick, which can also be used for the construction of new scaffold proteins. Further scaffold proteins can be elongated by adding domains through the C-terminal BamHI site. The variability of the scaffold proteins can be increased by assembly of different domains.