Table 1: Plasmid design for the AlcR system Transcriptional terminators included for each gene. Plasmid copy numbers adapted from [4].

Table 2: Plasmid design for the XylR system Transcriptional terminators included for each gene. Plasmid copy numbers adapted from [4]. For a detailed description of the pCK04AxylR plasmid see [1].

For the biological implementation of the SmoColi system, we constructed a three-plasmid system with different antibiotic resistances. This setup was chosen in order to clone the whole SmoColi system into the same bacterial cells.

The vectors used in our setup are the parts pSB6A5, pSB3C5, pSB3K3 and pSB4K5. These plasmids contain different origins of replication, thus resulting in different copy numbers of the correlating plasmid in E.coli. pBR322, p15A and pSC101 origins are used as a stable three-plasmid system [3]. The copy number influences the amount of protein produced by the cell and plays a crucial role in our design, especially for the bandpass filter. Depending on the function of a particular protein in our system we had to evaluate on which plasmid the corresponding gene should be cloned.

In the xylene-responsive system we included an artificial degradation pathway (xylWCMABN) so that we could establish a xylene gradient. For this purpose the plasmid pCK04AxylR was added [1]. To avoid incompatibilities of plasmids we adapted the rest of our system and changed the composition of the genes on each vector compared to the AlcR system.
pSB6A5

Figure 1: Improvement of the plasmid pSB6A1 to pSB6A5

The pSB6A5 plasmid we constructed by reducing the size of the antibotic (Ampicillin) and orign (pMB1) casette of pSB6A1 and introducing of Terminators around the multiple cloning site. PMB1 is a medium copy orign with about 15-20 copies per cell. For the characterization of pSB6A5 see experimental results

General procedure

Figure 1: Cloning strategy for SmoColi

By cloning our parts we tried to follow the strategies proposed in the registry of standard biological parts [5] as much as possible . The general cloning scheme was performed as follows:

• If available, genes and promoters in SmoColi were obtained from the iGEM spring 2011 distribution kit. A codon-optimized version of alcR and P_tet-lacI_M1 were synthesized. P_alc- and P_U-promoters were obtained by PCR.

• After isolation of the DNA, ribosome binding sites and prefix (5' gaattcgcggccgcttctagag 3') respectively suffix (5' tactagtagcggccgctgcag 3') were added to the gene by polymerase chain reaction. The PCR products were purified and afterwards digested with XbaI-PstI.

• Plasmid vectors containing the corresponding promoter were linearized by SpeI-PstI and the XbaI-PstI fragment containing the gene of interest was ligated into the backbone.

• In a next step we isolated a EcoRI-SpeI fragment with promoter, ribosome binding site and gene and inserted it into a EcoRI-XbaI cut vector containing a transcriptional terminator.

• In the end all the constructed parts were assembled on the corresponding vectors for the final SmoColi system. Some parts (especially lacI in an actively expressed form) appeared to be negatively selected by E.coli. So we had to assemble the parts in a way that lacI was always repressed.

Sources of genes and promoters

Table 4: Promoters of SmoColi: source and final constructs of the different used promoters

Table 3: Genes of SmoColi: source and final constructs of the different used genes

Cloning successes and problems

All parts of the system were successfully assembled by this strategy, except for lacI and cI under control of the Tet promoter. Whenever we were sequencing a clone of either of those constructs, we always found protein inactivating mutations within the primer annealing part - either being deletions of the Shine-Dalgarno sequence of the RBS or frame shift mutations within the first bases of the coding regions. We first believed the problem to lie in the synthesis of the very long oligonucleotides, for adding the whole RBS and prefix in front of the genes. But also performing a whole plasmid PCR with short primers did not solve the problem and we still only got non-protein-generating sequences. We believe this to be due to lethality of high overexpression levels of LacI or CI, leading to a selection for clones mutated in a way that lacI or cI is not expressed. Thus one always has to make sure to clone lacI or cI behind downregulated promoters, for example for P_tet you have to make sure that there is TetR present in the cell.

Table 5: Different ribosome binding sites and their estimated strengths which are included in our design. The strengths of the sequences were estimated with a RBS strength calculator [7]. In the last column it is listed in front of which genes a particular RBS was added.

Another important role in the synthesis of proteins is played by the ribosome binding site (RBS), the location where the ribosomes are bound to the mRNA at the initiation of the translation process. The ribosome binding sites which we included in our system were designed in several different ways:

• For the bandpass filter we implemented the ones which are also used for the plasmids in [2].

• For the other parts we either tried to use naturally ocuring ribosome binding sites or took specific ones from the Anderson library of ribosome binding sites [6].

The strength of all the ribosome binding sites used was estimated with a RBS calculator. Although the efficiency of protein synthesis is not directly proportional to the estimated values, they can still indicate whether a certain RBS is more or less strong. Below you can find a list of the different ribosome binding sites used, their estimated binding strengths and in front of which genes they were put.