Team:Open Science Global/Wetware/BaClo-Assembly-Standard
Assembly: The Plasmid Backbone Cloning (BaClo) Assembly Standard
Although B. subtilis and P. pastoris have advantages over E. coli for recombinant protein secretion, the high-copy vectors and origins of replication it is much easier to propagate and extract genetic parts from cloning vectors in E. coli. In other words, even if our wetware is designed for use in B. subtilis and P. pastoris, we will still need to be able to store both basic parts and composite genetic devices in E. coli prior to assembly and transformation or conjugation into the target manufacturing cell type.
To address this challenge, we implemented the Open Yeast Collection’s extension of the MoClo assembly standard, which we call the plasmid Backbone Cloning assembly standard, or BaClo. Level 0 parts are inserted into and propagated in cloning vectors like FreeGenes’ pOpen_v3 using BbsI Golden Gate assembly; while level 1 genetic devices are assembled from these level 0 part plasmids with BsaI Golden Gate. The BaClo assembly standard contains the standard MoClo part definitions and overhangs for promoter (Prom), ribosome binding site (RBS), protein coding sequence (CDS), and transcriptional terminator (Term) part types, as well as the overhangs for the holding vector. However, in addition to these traditional MoClo part types, BaClo contains level 0 definitions for eight additional part types, that comprise the backbone of a flexible shuttle vector between E. coli and other cell types.
These part types include E. coli selection markers (EcoSel) and origins of replication (EcoOri) so the assembled plasmids can replicate in E. coli; ‘Flank Forward’ (FF) and ‘Flank Reverse’ (FR) parts, which can contain homology arms or origins of replication for non-E. coli cell types; selection markers for target cells (TSel) other than E. coli; 5’ and 3’ ‘left’ and ‘right’ assembly connector parts (ACL and ACR), which each contain a single non-BsaI Type IIS restriction site and enable higher-level assembly of multiple transcription units; and finally, a ‘packaging’ part (Pkg), which can hold an origin of transfer for shuttling a vector directly from E. coli cells into target cells of interest via conjugation. Each of the overhangs that define these new part types was selected using the NEBeta GetSet Golden Gate tool; and a one-pot assembly with all 12 part types is predicted by NEBeta’s Ligase Fidelity Viewer to have over 98% assembly fidelity.
BaClo is the same assembly standard as is used for the FreeGenes Open Yeast Collection (OYC), except that the OYC, being designed for eukaryotes, lacks RBS parts. Because we chose to use the same assembly standard for our B. subtilis toolkit, the species-agnostic components of the OYC such as the multi-gene assembly connectors will be interoperable with our wetware and usable to build genetic devices in B. subtilis.
Refactoring pHT43
We set about applying the BaClo assembly standard to pHT43, an E. coli-B. subtilis shuttle vector that stably replicates in B. subtilis, and is designed express inserted CDSs under lactose/IPTG induction and secrete the recombinant protein out of the cell.
pHT43 E. coli Selection Marker
pHT43’s E. coli selection marker conveys the same resistance (to ampicillin-family antibiotics) as the selection marker for the pOpen_v3 holding vector, so we replaced it with two E. coli selection markers (one for kanamycin resistance, one for spectinomycin resistance) whose sequences we drew from vector sequences in the Standard European Vector Architecture (SEVA) repository. These were defined as BaClo EcoSel parts and synthesized as IDT gBlocks with the following architecture:
-->BbsI>GGAG-->BsaI>GCAA---EcoOri ColE1---AAAA<ACTA<--CGCT<BbsI--
-->BbsI>GGAG-->BsaI>TACT---EcoOri ColE1---AAAA<BsaI<--CGCT<BbsI--
Where BbsI digestion assembles the gBlock into pOpen_v3 or an equivalent cloning vector, and BsaI prepares the level 0 part for level 1 Golden Gate assembly into a composite genetic device.
pHT43 E. coli Origin of Replication
pHT43’s E. coli origin of replication (ColE1, used in pMB1, pBR322 and pUC vectors) was defined as an EcoOri part, in a gBlock with this architecture:
-->BbsI>GGAG-->BsaI>TACT---EcoOri ColE1---AAAA<BsaI<--CGCT<BbsI--
pHT43 B. subtilis Origin of Replication
pHT43’s B. subtilis origin of replication is quite large (~2.9 kb), adding to the considerable size of the plasmid (~8 kb) even without a recombinant gene inserted. Titok et al. showed that the origin of replication used by pHT43 can be reduced to ~1.9 kb while maintaining function; our wetware designs include this ‘Titok fragment’ which can serve as the origin of replication. We created designs in which both the full-length and the Titok fragment were defined as BaClo FF parts in gBlocks:
-->BbsI>GGAG-->BsaI>AAAA---FF pHT43 Ori---AAGG<BsaI<--CGCT<BbsI--
-->BbsI>GGAG-->BsaI>AAAA---FF Titok Fragment---AAGG<BsaI<--CGCT<BbsI--
Unfortunately, gene synthesis of both the full-length and the Titok fragment of the B. subtilis origin of replication from pHT43 has failed thus far; we are currently awaiting testing of an alternative assembly method that leverages Golden Gate rather than polymerase cycling assembly. Without this origin of replication, we have not yet been able to assemble and test a shuttle vector that replicates in both E. coli and B. subtilis. In the meantime, we have additionally designed 5’ and 3’ homology arm parts (defined as BaClo FF and FR parts respectively) derived from the loci around the B. subtilis amyE, thrC, and lacA genes. We used homology arms from pre-existing integrative B. subtilis plasmids (pBs1C, pBs2E, and pBs4S respectively) as a starting point, but changed the specific loci of some of the arms to avoid Type IIS and other restriction sites that are either used in our assembly standard, or could be used in extensions of it.
pHT43 B. subtilis selection marker
pHT43’s B. subtilis selection marker CatA9, which confers chloramphenicol resistance, was defined as a BaClo Tsel part, with the relevant BsaI overhangs. The CatA9 gene is adjacent to, but on the opposite strand and facing away from, the LacI CDS and inducible promoter that drives recombinant protein expression.
-->BbsI>GGAG-->BsaI>AAGG---<TSel CatA9<---ATGA<BsaI<--CGCT<BbsI--
pHT43 Recombinant Protein Promoter
The LacI/LacO inducible promoter was defined as a BaClo Prom part. In order to make expression from our constructs more predictable, to make it so that transcription and translation levels are more independent of the specific coding sequence downstream of the promoter, we added a RiboJ54 transcriptional insulator downstream of the LacO.
-->BbsI>GGAG-->BsaI>GGAG---Prom LacI/LacO---TACT<BsaI<--CGCT<BbsI--
pHT43 Transcription Unit Machinery
The original pHT43 architecture has a ribosome binding site, start codon, a signal peptide secretion tag from B. amyloliquefasciens, and an in-frame multiple cloning site for palindromic type II restriction enzymes. There are two primary limitations to this architecture. First, it only allows a single specific SecTag to be appended to a CDS of interest, rather than testing many secTags, as is likely necessary to find the few that secrete protein efficiently. And second, pHT43’s multiple cloning site does not support Golden Gate assembly. This means one-pot, multi-part assemblies (for instance, of tags with protein coding sequences) are not possible with the original pHT43 backbone. Our wetware collection aims to enable the construction of pHT43-like E. coli-B. subtilis shuttle vectors with Golden Gate compatibility.