One of the main hypothesis on which our technology is based on is that in the sequence of one of the subunits of RuBisCO can be used as a destination sequence to the inside of Carboxysome. This assumption is supported mainly on the idea that it has been reported that other bacterial microcompartments do have targeting sequences associated with enzymes that are inside of them (1).

At present, the exact destination sequence that guides enzymes into this particular microcompartment is unknown (1,2), but it has been described that the long subunit of RuBisCO is the best candidate for proper sequestration of enzymes  to the inside of Carboxysome (3). This is because RuBisCO’s sequence has a high conservation value between different microcompartments, so there is more possibility that in its sequence the targeting tag is contained. It has also been shown that expressing mutated RuBisCO’s large subunits inside bacteria expressing Carboxysomes resulted in subunits not going into the microcompartment.

Been able to find the destination sequence has been ranked as one of the most necessary progress to be made ​​in this area (4,5). In fact, for Whateversisome to work, we needed to find this sequence.

To answer our hypothesis , we designed  several experiments based on the use of fluorescent markers to determine by direct observation our results.

1) Marking Carboxysome :

We created a construct in which we modified the Carboxysome operon isolated from H. neopolitamus and optimized by David Savage and his collaborators, who kindly shared it to us (more information about this step here, by adding a red fluorescent protein (RFP) (BBa_E1010). RFP was coupled by an Ala -Ala- Ala linker at the 3 'end CsoS1A protein, which is one of the main shell protein from the Carboxysome. This was done using the Gibson Assembly technique (6).

2) Marking RuBisCO:

Four genetic constructs were made ​​ based on binding a green fluorescent protein (superfolder GFP) (BBa_I746916) to one of the subunits of native RuBisCO of H. neopolitanus. The variants correspond to: the long and short subunit of RuBisCO with GFP bound via a linker (BBa_K105012), at the 3 'and 5' end in both cases. The linker is a ten aminoacid flexible sequence, chosen to avoid potencial protein interference.

The expression of these genetic constructs is under the L-arabinose induced promoter, pBAD (BBa_K206000 ). This allows more control over the expression of the construct. We used the RBS corresponding to BBa_B0034, characterized by Elowits, 1999. They were contained inside the pSB4K5 vector (BBa_J04450), designed and optimized by GEM2006_Bangalore. It was chosen because it possesses a different antibiotic resistance than the one used on David Savage’s plasmid.
To build this BioBrick in the standard form required in order to summit it to the Registry of Standard Biological Parts, we had to mutate an EcoRI restriction site that was in the middle of sequence of the large subunit of RuBisCO. To achieve this, we used the Gibson Assembly method.