What is the SubtiToolKit?

Bacillus subtilis is a valuable industrial protein production platform, with its spore and biofilm forming capabilities garnering attention due to their potential application in areas ranging from biosensing to smart materials. As such, B. subtilis is a major synthetic biology chassis, yet it lacks a standardized toolbox for genetic manipulation similar to those available for Escherichia coli. EcoFlex is a commonly-used, versatile modular DNA assembly toolkit for Golden Gate cloning in E. coli. BacilloFlex is an extension of the EcoFlex assembly standard to B. subtilis. A Golden Gate toolkit can accelerate and simplify synbio research projects conducted in the chassis of choice, thanks to this method conferring several advantages such as: possibility of one pot reaction, of assembling multiple parts in a single reaction, scarless assembly and cost-efficacy by virtue of only requiring a few TypeIIS restriction enzyme to clone all genetic elements. Hence, Dr. Joaquin Caro Astorga, part of the Ellis lab, has embarked on a project to develop a gram positive bacteria Golden Gate Toolkit, inspired from EcoFlex, called the SubtiToolkit, with a primary focus on cloning in B. subtilis.

How does the SubtiToolKit work?

The SubtiToolkit involves a hierarchical system with 4 levels, similar to EcoFlex:

• L0 - storage of bioparts eg. promoters, RBSs, CDSs, terminators and spacers
• L1(A-D) - transcriptional units 
• L2 - vector hosting up to 4 multiple transcriptional units, in order L1A to L1D
• EXP L1 and L2 - final destination vectors that can be expressed in B. subtilis, featuring an appropriate origin of replication and antibiotic resistance marker. Designed to host 4 inserts. 
• MAD L2 - integrative vector, designed to have two transcriptional units flanked by homology arms of target locus assembled in L1 backbones (5’ end in 1A and 3’ end in 1D).

 The entire system works on the basis of 3 TypeIIS restriction enzymes: 

• BpiI - for assembly into L0
• BsaI -  for assembly into L1 (including EXP L1 vector), as well as L2s into L1
• BsmbI -  for assembly into L0  (including EXP L1 vector)

The toolkit design dictates a fixed number of lower-level assemblies to be fit into higher level vectors, namely 4, which if not met means that a mismatch will occur between the backbone’s overhang and the overhang of the final assembly. In the event in which less than 4 assemblies need to be inserted in a higher level vector, the toolkit provides the use of spacers, short sequences featuring the standardized overhangs of the missing assemblies, allowing successful closing of the higher level construct. As a result, spacers 1B, 1C and 1D would be used to “fill” alongside the newly produced construct 1A a L2 assembly, allowing expression in B.subtilis.

Our Contribution

Within the scope of our project, the SubtiToolkit was used for all cloning aspects in each of the three subsections of our wet lab, namely: spore surface display of chitinases, self-digesting plasmid assembly and germinant receptor engineering. As such we have been the first users of the toolkit testing its usability and functionality outside of the creator. We have also been the first to utilize it in the context of a wider project. Hence, we have been able to expand the kit to include useful parts for users exploring different applications in B. subtilis eg. germinant receptor work. Through our experience we have also been able to optimize protocols, identify areas of improvement to increase efficacy and user-friendliness and propose different avenues to achieve this through to the creator.

Extension of the Toolkit

Throughout our project we developed L0 and L1 parts that can prove useful to future users of the toolkit, for applications far beyond our own. Below is a list of the same:

• Protein Display: B. subtilis spores represent robust protein display solutions. Throughout our summer work, we have expanded STK to include spore coat proteins fusions with chitinase (ChiS). Both L0 and L1 parts based on this fusion protein system can be used by future users that are exploring the use of B. subtilis in biocontrol, biosensing and bioconversion. More significantly, we introduced a new promoter and RBS combination that is key to spore surface display. The sporulation process in B. subtilis is characterized by a tightly controlled and specific gene expression system, with anchor protein expression regulated by sigmaK, a transcriptional regulator active in the mother cell. Spore coat proteins are not naturally expressed constitutively, infact most spore surface display projects utilize a native anchor protein promoter in their designs. Prior to our work, the SubtiToolkit did not contain any promoters activated by sigmaK. The addition of this part will be highly beneficial to any future users interested in protein display.
 
‍

• Germinant receptor engineering:  During the summer, we delved deep into the world of germinant receptors trying to understand their function. (link to dry lab page). Extensive computational work was conducted to identify key residues responsible for the interaction with native and foreign ligands in GerA. Based on this research it was clear that the optimal approach towards the engineering of a receptor involved screening of a mutant library generated on the basis of the binding site location within the receptor’s amino acid profile .  The creation of such a mutant library is generally conducted by DNA synthesis providers such as TWIST. Another alternative is error prone PCR, a cheap and accessible mutant generation strategy which finds easy adoption in labs worldwide. In both cases, a pool of linear fragments is produced. Being interested in receptor repurposing,  we started developing our our experimental design when we identified the assembly of such fragments into the germinant receptor operon to be a bottleneck. As a solution, we designed in silico a destination vector for convenient mutant library assembly into a plasmid backbone. This design would allow for higher efficacy and prevent loss of complexity of the mutant library by having a single insert into a backbone ready for expression in B. subtilis.   This is useful for receptor engineering experiments because these would need to be conducted in a KO strain, so you would need a plasmid operon. This design would be useful for any future user interested in achieving targeted germination in response to a non-native ligand.
 
‍

• Self-digesting plasmid circuitry: Thanks to the gene expression pattern in sporulation, it is possible to selectively degrade plasmid DNA so as to have spores free of foreign DNA. In the past months, we have successfully cloned and introduced into SubtiToolKit L1 parts designed to enable degradation of plasmid in the forespore during T3 sporulation phase. These consists of a Cas9 gRNA targeting the ori used in the STK backbones, Cas9 and D15 protein sequences which have both been optimised for B. subtilis. These parts can be assembled together in a L2 expression vector along with a forth L1 transcriptional unit of choice, allowing future users to create their own spore surface display systems with the same biocontainment strategy.
  
‍

• HAs for integration at GerA locus or to perform a KO of GerA operon: L1 parts featuring 5’ and 3’ homology operon which can be used  in the pMAD integrative vector to either perform a gene KO or genomic integration at this specific locus.
 
‍

• HAs for integration at GerK locus or to perform a KO of GerK operon: L1 parts featuring 5’ and 3’ homology operon which can be used  in the pMAD integrative vector to either perform a gene KO or genomic integration at this specific locus.

Troubleshooting and Improvements

With our experience as the first heavy users of SubtiToolKit, we identified key bottlenecks limiting the user friendliness and efficacy of the toolkit. 

Problem:  

• In several cases, moving from a lower level plasmid to a higher order one does not involve a change in antibiotic resistant marker. This complicates and lengthens screening for identification of successful transformants,  leading to, in some extreme cases, hundreds of colony PCRs and single digit hit rate. 
• Short term solution: two different strategies could solve the issue key vectors, such as STK-EXP, could be modified to present distinct antibiotic markers than their lower order counterpart. This can be achieved via PCR mediated addition of the marker of interest. 

Inability to express L1A-D  plasmids directly in B. subtilis. The toolkit’s high throughput assembly pipeline restricts users from testing individual transcriptional units assembled in L1A-D directly in B. subtilis. For example, in our self digesting plasmid circuitry we required the use of L1A-D backbones so as to have transcriptional units that we could assemble into a single L2 construct. However, although we wanted to characterize the expression of individual TUs in B. subtilis to evaluate burden, and identify optimal design parameters, this was not possible. To do the same, we would have had to perform a separate L1 assembly using L0 parts into the L1-EXP backbone. The addition of a B. subtilis ori and antibiotic marker would greatly increase the flexibility of the toolkit. 

In our experience, STK-EXP has been difficult to successfully transform into bacillus. One possibility is that this is due to its tendency to dimerise. Across our final work weeks, we have not been able to successfully pinpoint the root issue with STK108. As a consequence, we suggest that whilst research is being conducted to improve this part, different homology arms should be provided to perform genomic integration of target TUs in diverse  and well characterized loci. Furthermore, we recommend the introduction of a restriction digest site flanking the TU, such as BsmBI. This would not only improve the ease-of-use of the toolkit, but enable the adoption of a consistent and reliable pipeline for cloning in SubtiToolKit.