Team:UC Chile/Formation




Whateversisome’s technology is based on using Carboxysome, a bacterial microcompartment, as a tool for in vitro manufacturing. For this reason, studying its formation is fundamental for our work. In this section, we describe the carboxysome’s characteristics and the experiments we did in order to study its assembly. We created a fusion protein between the red fluorescent protein (RFP) and one of the major shell proteins, csoS1A, of the Carboxysome. This was done with the objective of facilitate its observation under microscopy.


Bacterial microcompatments (BMC) are protein aggregates that can generate a space or lumen isolated from the bacterial cytoplasm, where key metabolic processes occur (1). Particularly, Carboxysome correspond to one of the microcompartments most studied and its structure is well known: is based on 12 to 15 proteins, the capsule and the ones associated with it, and the internal enzymes (2).

The alpha Carboxysome from Halothiobacillus neapolitanus is a good model of study for many bacterial microcompartments. It has an icosahedral structure of about 80-150 nm (cross section) and its mass is close to 300 MDa (2). Its main function is to be part of the metabolism associated with carbon fixation. Inside the carboxysome are the enzymes Carbonic Anhydrase (CA) and Ribulose -1,5 - Biphosphate Carboxylase / Oxygenase (RuBisCO). Both of them carry out their reactions (1,3).

Another important feature of this BMC is that its heterologous expression in E. col has been achieved and the microcompartment has maintained its function both in vivo and in vitro (4). This could be carried out thanks to the characterization of the Carboxysome operon, which contains all nine essential proteins for its formation: the two subunits of RuBisCO, large (RbbL) and small (RbbS), the CsoS2 (a protein associated to the shell, necessary for the formation of the BMC), the Carbonic Anhydrase (CosS3), the major constituents of the shell (CsoS1A, CsoS1B and CsoS1B) and the lower constituents (CsoS4A and CsoS4B) (4,5). Heterologous expression of this type of microcompartment in E.coli produces some important features: first, is that the cell takes a filamentous form. Second, Carboxysomes get evenly distributed across the cell and is possible to see protein aggregates at the ends of the cells (4).

The formation and assembly of Carboxysomes in the bacterial system is not completely known. Moreover, it is postulated that the assembly is a highly complex process that requires intermolecular interactions between the subunits of RuBisCO and the proteins of the shell, forming a kind of lattice in its early stages of formation (5). Carboxysome stability and the correct stoichiometry of its parts are also fundamental for a correct assembly (4,5).

Experimental Design

pHnCBS1D plasmid was cloned and optimized by Bonacci Et al. It is a high copy plasmid that has a size of 13.242 bp, in which the whole Carboxysome operon (7.686 bp) and the coding region of CsoS1D, a protein located outside the operon and that has been reported to aid in the assembly and function of BMC, are located (4). The operon is under the regulation of an IPTG-inducible promoter from the lactose operon and has a replication origin of the pUC19 plasmid family. It also has resistance to chloramphenicol and the LacI repressor. Our team got this plasmid thanks to the collaboration of David Savage, from California Berkeley University. Since this plasmid is optimized for the heterologous expression of this microcompartment, we decided to use it. The construct of Carboxysome was sent as a BioBrick in the specified format.

Because Carboxysome is a bacterial microcompartment of a size close to 10nm (6), one way to study its expression and formation is with the use of fluorescent proteins fused to one of the major proteins of the shell (4). In order to achieve this, we fused via an Ala-Ala-Ala linker to a red fluorescent protein (RFP) (BBa_E1010) at the carboxyl terminus of CsoS1A.

In order to know which strain was better for the expression of the BMC, we performed a study in different strains of E. coli: TOP10, BL21 and DH5alpha, in order to see if there were any significant differences. All strains were made competent under the same protocol (Click HERE to go to the protocols section), as well as the bacteria transformations and inductions remained the same for the three strains.

At this point, is important to remember that the objective of this project is to target proteins into the Carboxysome: For this, we created four constructs that consisted on one of the two subunit of RuBisCO, large or small, linked to a GFP at the amino or carboxyl terminus. More information about part of the project, go to the Targeting section (Link). So, we also transformed the bacteria of each strain with the two vectors: the pHnCBS1D for Carboxysome expression (Bba_K1113100) and one of the four available targeting vectors (Bba_K1113002).

We chose the better strain (TOP10) depending on which showed better parameter of induction and less basal expression at the same conditions, after being analyzed with Fluorescent Microscopy. The selected strain was observed for proper assembly of the BMC by a Transmission Electron Microscopy (TEM), to be sure that the observed fluorescence corresponded to properly assembled Carboxysomes and not some kind of inclusion body or protein aggregates.

Another important point to analyze in the formation of Carboxysome was whether the co-transformed bacteria with our targeting construct (Link) produced any kind of problems associated with increased energy expenditure (because of the presence of two plasmids in the bacteria) or because of the presence of more quantity of one subunit of RuBisCO.

In order to check this, we decided to implement a strategy similar to the one used for the Dundee 2011 iGEM team.

Click here for more information about the strategy to check if the Carboxysome was assembling correctly.

Dundee 2011 strategy to check if a microcompartment is assembling correctly.
Old BioBrick characterization: ssrA tag

The ssrA tags are peptide signals that are incorporated into proteins in the carboxyl terminal that are recognized by proteases, such as ClpX. Enzymes which in the presence of the ssrA degrade the entire protein. These tags consist in an amino acid sequence that binds to the protein SspB, a protein adapter that improves protease recognition of the degradation tag. At the end of the tag there is a sequence that can vary and it is the one recognized by the protease ClpX. This final sequence is composed of three amino acids; it can be LAA, DAS or LDD (A: Alanine, D: aspartic acid, L: leucine, S: serine). Besides these two tags sequences, it can have; 2, 4 or 8 amino acids, which can change the rate of degradation of the protein labeled.

Diferent sequences of ssrA-like Tags
During 2011, the Dundee iGEM team worked with Pdu, a bacterial microcompartment (BMC) like the Carboxysome. In their project they attached a ssrA tag to a GFP targeted to Pdu. By observing the fluorescence activity of the GFP despite having the degradation tag, they could prove that their BMC was able to protect the proteins from the protease, and thus, prove that their compartment was functional and well assembled. We wanted to replicate the experiment in order to see if the Carboxysome is also able to protect the GFP with the ssrA tag from degradation and by that affirming their hypothesis of formation of the BMC by protection of the GFP with the degradation tag. Also, we wanted to confirm that this brick, from Parts Registry, is functioning properly by observing degradation of the fusion protein containing the ssrA tag.

Under the concept just mentioned, we wanted to test the theory incorporating to our targeting BioBrick that already have the GFP (BBa_I746916) the ssrA degradation tag: moderately fast (Bba_M0052). This tag was chosen because it is not fastest, so it gives the fusion protein time to enter the BMC before being degraded. Eventually we could test the other degradation tags, by modifying induction rates of our constructs.

Thus, when the RuBisCO large subunit joined with the GFP (RubL+ GFP) construct is induced, we can observe the protein in the bacterial cytoplasm. If the same construct is also induced with the Carboxysome formation plasmid, we would observe GFP colocalized with the BMC and nothing in the cytoplasm. Furthermore, if we use the RubL + GFP + ssrA construct, we should not observe GFP in the bacteria's cytoplasm if the protease system is working correctly. Finally, if we induce the Carboxysome formation plasmid, again we should see GFP colocalized with the BMC, therefore this GFP would be protected by the Carboxysome from protease activity.

Diagram of the mecanism of action of the ssrA tag
Right now, we continuing working on the genetic construct of RubL + GFP + ssrA tag. We are working hard to achieve the correct construct for further experiments in order to corroborate the hypothesis about the ssrA tags. For obtaining results, once it's ready the construct, we need to transform E. coli to be induced and see if the tag for the protease is working. In case of a positive result, we will transform E. coli with the construct with ssrA in addition of Carboxysome construct. This is needed, to prove if this particular BMC is capable of protecting GFP from the protease as Pdu did it.

Further Work

One of the limiting steps in the formation of Carboxysome corresponds to the stoichiometry of the constituent proteins. To see if the stoichiometry is correct, you can make a protein extraction and run a gel to analyze the correct expression of the 10 proteins involved in the assembly. This way we can check that the fusion protein is not interfering with this balance.

Another thing we were planning on doing is to knockout the different subunits of RuBisCO to see if can obtain a correct formation of the microcompartment but with more available space inside for the metabolic reactions to occur.
  • 1. Kerfeld C a, Heinhorst S, Cannon GC. Bacterial microcompartments. Annual review of microbiology [Internet]. 2010 Jan [cited 2013 Jun 7];64:391–408. Available from:
  • 2. Cheng S, Liu Y, Crowley CS, Yeates TO, Bobik T a. Bacterial microcompartments: their properties and paradoxes. BioEssays : news and reviews in molecular, cellular and developmental biology [Internet]. 2008 Nov [cited 2013 Jun 7];30(11-12):1084–95. Available from:
  • 3. Frank S, Lawrence AD, Prentice MB, Warren MJ. Bacterial microcompartments moving into a synthetic biological world. Journal of biotechnology [Internet]. 2013 Jan 20;163(2):273–9. Available from:
  • 4. Bonacci W, Teng PK, Afonso B, Niederholtmeyer H, Grob P, Silver PA. Modularity of a carbon-fixing protein organelle. 2011
  • 5. Iancu C V, Morris DM, Dou Z, Heinhorst S, Cannon GC, Jensen GJ. Organization, structure, and assembly of alpha-carboxysomes determined by electron cryotomography of intact cells. Journal of molecular biology [Internet]. 2010 Feb 12 [cited 2013 Sep 23];396(1):105–17. Available from:
  • 6. Shively JM, Ball F, Kline B. Electron Microscopy of the Carboxysomes (Polyhedral Bodies) of Thiobacillus neapolitanus. Journal of bacteriology. 1973;116(3):1405–11.
  • 7. McGinness, K. E., Baker, T. A., & Sauer, R. T. (2006). Engineering controllable protein degradation. Mol Cell, 22(5), 701-707.