These parts were utilized in combination with existing basic parts for the assembly of multi-gene constructs suitable for expression of the fungal bioluminescence pathway in Synechocystis sp. PCC 6803. Our assemblies were performed through golden gate cloning using MoClo standardized fusion sites. Every step of the cloning process was screened and sequence-verified through antibiotic selection, blue-white screening, colony PCR, and analytical restriction digestions.

After designing the coding sequences with BbsI restriction sites, BsaI restriction sites, and fusion sites on the 5’ and 3’ flanks (as seen on BBa_K4350814) and codon-optimizing for expression in Synechocystis sp. PCC 6803, all eight coding sequences were synthesized as gBlocks by IDT.

Plasmids from the CyanoGate Kit and MoClo Toolkit were acquired from AddGene for assembly backbones, vectors, promoters, and terminators. The necessary plasmids from these kits were extracted and verified through analytical digestions, however, visualization of analytical digestions through gel electrophoresis yielded unexpected bands. To investigate this, we designed experiments to vary the digestion times, which revealed that the unexpected bands were the result of incomplete digestions.

After extraction and sequence-verification of these plasmids, The linear fragments synthesized by IDT were cloned into the pAGM9121 universal level 0 backbone (BBa_K2703011) through digestion with the Type IIS restriction endonuclease BbsI.

The first round of level 0 transformations successfully produced white colonies on blue-white screening plates. However, colony PCR results of H3H, NpgA, nnLuz, and CPH plasmid transformants only indicated proper amplification of the NpgA level 0 plasmid, and amplification of the H3H level 0 plasmid at a lower than expected band height. To troubleshoot this, the golden gate assemblies of all 8 level 0 parts were directly amplified through PCR to assess if the results were due to improper assembly or transformation. These results indicate proper amplification of the nnLuz, H3H, NpgA, nnLuz V49A, nnLuz V49A d33, and psLuz level 0 plasmids, but no amplification of CPH or HispS. Furthermore, it was determined that the annealing temperatures utilized for PCR were miscalculated, which may have contributed to the poor results. Colony PCR of new colonies were attempted at the correct temperatures for nnLuz, HispS, nnLuz V49A, nnLuz V49A d33, and psLuz transformants, which indicated successful amplification of nnLuz, nnLuz V49A, nnLuz V49A d33, and psLuz, but no amplification of CPH or HispS. Miniprep plasmid extractions of nnLuz, H3H, NpgA, nnLuz V49A, nnLuz V49A d33, and psLuz were digested with restriction endonucleases, and all produced expected bands, thus sequence-verifying all Level 0s except CPH and HispS.

Lack of successful PCR results for CPH and HispS indicated that another assembly was necessary. Transformants of newly assembled plasmids were analyzed through colony PCR, using re-calculated annealing temperatures for each plasmid. All colonies indicated amplification at the same, improper height, and no amplification of properly-sized DNA, indicating that other nucleic acids present in the colonies were being amplified.

Despite these poor results, plasmid extraction and analytical digestions were performed, which yielded expected results for the successful assembly and transformation of CPH, but the bands produced by the HispS digest were far too low, indicating improper assembly.

Another HispS assembly was attempted alongside all other level 1 constructs, and assembled DNA was transformed. The level 1 transcriptional units were cloned into the pICH47732, pICH47742, pICH47751, pICH47761, and pICH47772 backbones included in the MoClo Toolkit (Weber et al., 2011). Colony PCR of level 1 nnLuz transformants and level 0 HispS transformants demonstrated successful amplification of the nnLuz level 1 plasmid, but no amplification of the HispS level 0 plasmid. Digestion of the level 1 constructs produced expected results, thus sequence-verifying level 1 constructs of all genes besides HispS. After re-assessing the annealing temperatures, re-visiting the golden gate assembly simulations, and analyzing previous gel results it was determined that the HispS level 0 assemblies were failing due to improper assembly of the pAGM9121 backbone lacking both the LacZ insert and the HispS insert, perhaps due to DNA impurities or mismatches.

Assembly simulations demonstrated successful assembly of HispS without issue, so the HispS assembly was attempted again with 5 replicates to ensure that at least one was successful. Colony PCR results of these transformants demonstrated successful amplification in at least two colonies, which were then sequence-verified through extraction and digestion. This was followed by assembly, transformation, extraction, and digestion of the level 1 HispS transcriptional unit.

Our final multi-gene constructs were assembled into the pCAT.000 and pCAT.015 plasmids provided by the CyanoGate Kit (Vasudevan et al., 2019) and cloned in E. coli, thus completing the assembly and cloning process.

---