The present invention relates to new synthetically generated carboxysome shells, synthetic shells encapsulating selected catalytically active moieties, cells including the synthetic shells and methods of making the synthetic shells and methods of encapsulation. The invention includes inter alia, uses of the shells as nanoreactors for enhanced catalytic performance of the encapsulated enzymes and the use of an encapsulation protein for recruiting external cargo proteins into a synthetic carboxysome shell.

BACKGROUND

Enzyme biocatalysis is generally more specific and sustainable and requires less energy than chemocatalysis. Development of enzyme-mediated biocatalysis is increasingly required across many industries in biofuel, energy, food processing, medical diagnostics, and therapeutics applications. However, enzyme instability and complexities of enzyme activation and inhibitory mechanisms are the major obstacles in enzyme biocatalysis technology. To overcome these limitations, many creative approaches have been developed, including enzyme immobilization, cross-linking, and encapsulation of target enzymes in synthetic cages.

Compartmentalisation is a ubiquitous building principle in cells, which permits segregation of biological elements and reactions. The carboxysome is a specialised bacterial microcompartment in all cyanobacteria and some chemoautotrophs. It plays an essential role in improving photosynthetic carbon fixation. The carboxysome encapsulates enzymes into a virus-like protein shell of -100 nm in diameter, which is tiled densely by proteins in the forms of hexamers, trimers, and pentamers. Each shell protein has a central pore (less than 1 nm in diameter), which is essential for selectively mediating the passage of substrate molecules into and out of the shell to boost enzymatic performance. The naturally designed three- dimensional structure, semi-permeability, protein encapsulation, and catalytic improvement of carboxysomes have inspired rational design and engineering of new nanomaterials, to incorporate desired enzymes into the protein shell for enhanced catalytic performance. However, to date, the prior art has failed to generate a synthetic carboxysome shell of sufficient interior volume as comparable to native carboxysomes, or synthetic carboxysomes having near natural features as compared to natural carboxysomes in their native hosts, and certainly not such carboxysomes comprising a cargo other than the cargo moiety of natural carboxysomes (ribulose-1 ,5-bisphosphate carboxylase-oxygenase, commonly known as Rubisco). Similarly, the prior art has failed to provide synthetic carboxysomes having a significantly reduced number of different shell proteins as compared to natural carboxysomes but comprising a cargo and having a size comparable to natural carboxysomes.

There is a need for synthetically engineered carboxysome shells having a loading capacity and characteristics equivalent to natural carboxysome shells in their native hosts, and particularly for synthetically engineered carboxysome shells that can be widely used to recruit a diverse range of cargos whilst retaining a loading capacity and characteristics equivalent to natural carboxysome shells in their native hosts. There is also a need for synthetically engineered carboxysomes having a reduced number of different shell proteins as compared to natural carboxysomes but comprising a cargo and having a size comparable to natural carboxysomes.

BRIEF SUMMARY OF THE DISCLOSURE

The present invention provides a heterologously engineered synthetic carboxysome shell having a diameter of around 100 nm.

By “around 100 nm” it is meant that the carboxysome shells of the present invention have diameters of from around 80 nm to around 120 nm. Preferably, in a population of carboxysome shells of the present invention, the average diameter of the carboxysome shells will be from around 80 nm to around 120 nm, and/or substantially all of the carboxysome shells in the population will have a diameter ranging from around 80 nm to around 120 nm.

Preferably, the synthetic shell is an a-carboxysome type shell.

Preferably, the synthetic shell comprises a number of proteins encoded by synthetic cso-1 and/or cso-2 operons. Preferably, the cso-1 operon comprises one or more genes selected from csoS2, csoS4AB, csoSICAB, csoSID and csoSCA encoding shell proteins and - carbonic anhydrase (CA). Preferably, the cso-2 operon comprises one or more genes selected from csoS2, csoS4AB, csoSICAB, and csoSID encoding shell proteins without csoSCA Preferably, the genes of the cso-1 or cso-2 operons are all Halothiobacillus genes, more preferably the genes of the cso-1 or cso-2 operons are all Halothiobacillus neapolitanus genes. In a particular embodiment, the cso-1 operon and/or the cso-2 operon comprise promotors and/or ribosome binding site sequences for one or more of the carboxysome shell protein encoding genes, and preferably the promotors and/or ribosome binding site sequences are the native promotors and/or ribosome binding site sequences for the carboxysome shell protein encoding genes. In particular embodiments the cso-1 operon and/or the cso-2 operon comprise such promotors and/or ribosome binding site sequences for two or more, three or more, four or more, five or more or six or more of the carboxysome shell protein encoding genes.

In the first preferred aspect of the present invention, the synthetic shell comprises all of the proteins encoded by the cso-1 or by the cso-2 operon.

In a particular embodiment of the first preferred aspect of the present invention, the synthetic shell does not comprise an encapsulated cargo. Optionally, in this embodiment, the outer surface of the shell comprises one or more active molecules attached thereto, and shells of this embodiment can be used to deliver molecules or drugs in biomedical research or treatment, such as doxorubicin.

Preferably, the synthetic shell of the first preferred aspect of the invention includes an encapsulation peptide (EP) to recruit an external cargo moiety into the synthetic shell. Ideally, the EP is the CsoS2 C-terminus fragment of CsoS2 and interacts with an internal surface of the synthetic shell and the external cargo moiety so as to form a link between the internal surface of the synthetic shell and the cargo moiety.

In a further embodiment of the first preferred aspect of the present invention, the synthetic shell comprises a cargo moiety encapsulated within the shell. Preferably, the cargo moiety is not a cargo found within naturally occurring carboxysomes. In particular, in this aspect of the invention, the cargo moiety is not a bacterial ribulose-1 ,5-bisphosphate carboxylaseoxygenase (Rubisco). Preferably the encapsulated cargo moiety is selected from the group comprising a protein, an enzyme, a biologically active molecule and a molecular probe. Preferably, the encapsulated cargo moiety is sensitive to a condition selected from the group comprising oxygen, redox and pH.

In some embodiments the cargo moiety is selected from [NiFe]-, [FeFe]- and [Fe]- hydrogenases, nitrogenases, nitroreductases, nitrilases, monooxygenases, cytochrome reductase, nitrile hydratase and polyethylene terephthalate (PET) hydrolase; particularly preferably the encapsulated cargo moiety is a hydrogenase enzyme selected from the group comprising [FeFe]-, [NiFe]-, and [Fe]- hydrogenase.

In a second aspect of the present invention, the synthetic protein shell consists of proteins encoded by the cso-S2 and csoSIA genes, and optionally also by the csoS4 gene, and the shell encapsulates a cargo moiety. n this aspect, the genes are preferably all Halothiobacillus neapolitanus genes. Preferably, in this aspect the encapsulated cargo moiety is selected from ths group comprising a protein, an enzyme, a biologically active molecule and a molecular probe, and may comprise Rubisco. In the second aspect of the invention, the preferred cargo moieties include all of the preferred cargo moieties of the first aspect of the invention, but in a particularly preferred embodiment of the second aspect of the invention, the cargo moiety comprises Rubisco, particularly a bacterial Rubisco and most preferably Halothiobacillus neapolitanus Rubisco.

The present invention also provides a method of constructing a heterologously engineered carboxysome shell according to the present invention, the method comprising:

(i) shell formation;

(ii) shell production and;

(iii) shell isolation.

In a first preferred aspect of the method of the present invention, the steps of shell formation and shell production together comprise generating synthetic cso-1 and/or cos-2 operons to heterologously express carboxysome shell proteins in a host, and this results in an empty synthetic shell comprising all of the proteins enclosed by the cso-1 or by the cso-2 operon when the expressed shell proteins undergo self-assembly.

In an embodiment of the first aspect of the method of the present invention, the method comprises the further step of attaching one or more active molecules to the outer surface of the shell. For example, a short protein peptide (for example 8 to 10 amino acids) that has a specific affinity to the targeted molecules can be fused to the outer surface of the shell, to ensure sufficient binding.

In an alternative embodiment of the first aspect of the method of the present invention, the method comprises, after the step of shell isolation, the further steps of causing the shell to disassemble and causing the disassembled shell to reassemble in the presence of a cargo moiety, so as to cause the cargo moiety to be encapsulated in the reassembled shell. Disassembly and reassembly of the synthetic shells may be carried out in nay convenient manner, for example by changing the conditions of the cells, such as by changing the pH of the medium in which the shells are held.

In a second preferred aspect of the method of the present invention, the step of shell formation comprises generating synthetic cso-1 and/or cos-2 operons to heterologously express carboxysome shell proteins in a host, and the step of shell production comprises linking an encapsulation peptide (EP) to an internal surface of the synthetic shell, preferably wherein the EP is the C-terminus fragment of CsoS2, and further includes recruiting a selected catalytically active moiety via the EP, thereby encapsulating it within the synthetic carboxysome shell as the cargo moiety to provide a synthetic shell comprising all of the proteins enclosed by the cso-1 or by the cso-2 operon and comprising an encapsulated cargo. In this aspect of the method of the present invention, the encapsulated cargo moiety is preferably not a cargo found in naturally occurring carboxysomes, and most preferably it is not a bacterial Rubisco. The cargo may preferably be any of the materials described with respect to the first aspect of the synthetic shells of the invention.

In a particular embodiment of the second aspect of the method of the present invention, the method comprises co-expressing in the same host cell cso-1 and/or cso-2 operons and an operon comprising a gene encoding a cargo moiety and at least one encapsulation peptide so that the cargo moiety is encapsulated within the synthetic shell during shell production, preferably the encapsulation peptide is the C-terminus fragment of CsoS2. This method preferably comprises expressing the operon comprising a gene encoding a cargo moiety and at least one encapsulation peptide before expressing the cso-1 and/or cso-2 operons.

In an alternative preferred embodiment of the second aspect of the method of the present invention, the step of shell formation comprises generating synthetic operons comprising all of the genes of the cso-1 and/or cos-2 operons and also comprising a gene encoding a cargo moiety, and the step of shell production comprises causing the synthetic operons to express the carboxysome shell proteins and the cargo moiety, so that the cargo moiety is encapsulated within the synthetic shell during assembly of the shell. As noted above, in this embodiment of the method of the present invention the cargo moiety is not a cargo normally found within natural carboxysomes, and most particularly the cargo is not a bacterial Rubisco.

In the first and second aspects of the method of the present invention, the cso-1 and/or cso- 2 operons preferably encode Halothiobacillus genes, more preferably the cso-1 or cso-2 operons encode Halothiobacillus neapolitanus genes.

In the first and second aspects of the method of the present invention, the cso-1 operon and/or the cso-2 operon comprise promotors and/or ribosome binding site sequences for one or more of the carboxysome shell protein encoding genes, and preferably the promotors and/or ribosome binding site sequences are the native promotors and/or ribosome binding site sequences for the carboxysome shell protein encoding genes. In particular, the cso-1 operon and/or the cso-2 operon comprise such promotors and/or ribosome binding site sequences for two or more, three or more, four or more, five or more or six or more of the carboxysome shell protein encoding genes.

In a third preferred aspect of the method of the present invention, the step of shell formation comprises generating a synthetic operon comprising the cso-S2 and csoSIA genes, and optionally also the csoS4 gene and/or by the csoSCA gene, and further comprising a gene encoding a cargo moiety, and the step of shell production comprises causing the synthetic operon to express the carboxysome shell proteins and the cargo moiety, so that the shell consists of proteins encoded by the cso-S2 and csoSTA genes, and optionally also by the csoS4 gene and/or the csoSCA gene, and the cargo moiety is encapsulated within the synthetic shell during assembly of the shell. In this aspect of the method of the present invention, the cargo may preferably be any of the materials described with respect to the second aspect of the synthetic shells of the invention. In a particularly preferred embodiment of this aspect of the method of the invention the cargo is a bacterial Rubisco, and more preferably, the gene encoding a cargo moiety comprises one or more genes encoding a bacterial Rubisco, more preferably the Halothiobacillus neapolitanus Rubisco or components thereof, such as CbbL and/or CbbS.

In the third preferred aspect of the method of the present invention, the cso genes are preferably all Halothiobacillus genes, more preferably the genes are all Halothiobacillus neapolitanus genes. Alternatively, or additionally, the synthetic operon preferably comprises promotors and/or ribosome binding site sequences for one or more of the cso genes, and preferably the promotors and/or ribosome binding site sequences are the native promotors and/or ribosome binding site sequences for the cso genes. In particular, the synthetic operon comprises such promotors and/or ribosome binding site sequences for two or more of the cso genes.

Preferably, in the methods of the present invention, the step of shell isolation comprises purification of the synthetic shells following their production in a host. The purification is achieved by any suitable method, such as by a two-step centrifugation in which step one is differential centrifugation which concentrates the shells and crudely separates them from other organelles; and step two is sucrose gradient ultracentrifugation of the concentrated shells from step one to further purify the shells. Preferably, in the methods of the present invention the host is a bacterial, yeast, algae or plant host, more preferably the host is E. coli, Salmonella or a Gram-positive bacterium such as Corynebacterium glutamicum.

According to a yet further aspect of the invention there is provided a cell containing the synthetic carboxysome shell of the present invention. Preferably, the cell is a host plant cell, a host bacterial cell, a yeast cell or an algae, such as Chlamydomonas reinhardtii.

According to a yet further aspect of the invention there is provided use of the synthetic carboxysome shell of the present invention as a metabolic nanoreactor.

According to a yet further aspect of the invention there is provided use of the synthetic carboxysome shell of the present invention to encapsulate a target protein, enzyme or biologically active molecule to enhance enzyme biocatalysis and/or as a molecule delivery vehicle in a host cell or organism.

According to yet further aspect of the invention there is provided use of an EP comprising the C-terminus fragment of CsoS2 for recruiting external cargo proteins into a synthetic carboxysome shell.

It will be appreciated that preferred features ascribed to one aspect of the invention applies mutatis mutandis to each and every aspect of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention are further described hereinafter with reference to the accompanying drawings, in which:

Figure 1 shows the engineering and characterization of empty a-carboxysome shells. Figure 1a shows a schematic of the carboxysome structure and metabolic pathways. The carboxysome shell serves as a physical barrier to permit passage of cytosolic bicarbonate (HCOs') and ribulose-1 ,5-bisphosphate (RuBP) into the carboxysome. The metabolic product 3-phosphoglycerate (3-PGA) is transported across the shell and is metabolized via the Calvin-Benson-Bassham cycle. Figure 1 b shows genetic organizations of the native cso operon, synthetic cso-1 and cso-2 operons. Figure 1c shows thin-section electron microscopy (EM) of E. coli cells expressing the cso-1 operon (left) and cso-2 operon (right), respectively. Arrows indicate a-carboxysome shell particles discerned in the cell. Figure 1d shows SDS-PAGE of purified cso- 1 (left) and cso-2 (right) shell structures. Figure 1e shows transmission EM of purified cso-1 (top) and cso-2 (middle) shells in the 20% sucrose fractions. Shells with CA are significantly smaller in diameter than those without CA (*** p = 0.0007, n = 150, t-test), implying the role of CA in confining shell architecture (bottom).

Figure 2 shows the roles of CsoS2 in carboxysome shell assembly and identification of CsoS2 C-terminus as the encapsulation peptide. Figure 2a shows genetic organization of the synthetic cso-3 operon. Figure 2b shows thin-section EM of E. coli cells expressing the cso-3 operon (top) and cso-2 operon (bottom). Yellow arrows indicate protein aggregates formed by non-assembled shell proteins in the cso-3 cells. White arrows indicate individual a-carboxysome shell particles in the cso-2 cells. Figure 2c shows SDS-PAGE of the 50,000x g pellets purified from E. coli cells expressing empty plasmid (control, C), cso-1 (1), cso-2 (2) and cso-3 (3) operon, respectively. Figure 2d shows confocal images revealed the formation of shell assemblies in E. coli cells in the presence of both CsoS2 and shell proteins. Figure 2e shows confocal images of E. coli cells expressing GFP, GFP-CsoS2 C- terminus (csoS2-C), co-expressing cso-2 and GFP, or co-expressing cso-2 and GFP-csoS2- C, indicating that the CsoS2 C-terminus could function as an EP to mediate incorporation of external proteins into the formed shell structures. Figure 2f shows immunoblot confirms the presence of GFP and shell proteins in the samples purified by sucrose gradient centrifugation from E. coli cells producing GFP-EP alone or GFP-EP with shells, using the anti-Histidine (top) and anti-CsoS1A/B/C (below) antibodies. Figure 2g shows transmission EM of empty a-carboxysome shells (top) and shells with GFP interiors (below) from the 20% sucrose fractions. Yellow arrows indicate GFP proteins seen in the lumen of the shells.

Figure 3 shows construction of a hydrogenase-containing nanoreactor based on the carboxysome shell. Figure 3a shows generation of the hyd operon to produce [FeFe]- hydrogenase HydA fused with ferredoxin (Fd) from the green alga Chlamydomonas, ferredoxin: NADP ⁺ -oxidoreductase (FNR) from E. coli, and the HydA maturases HydE, HydF, and HydG. The hyd and cso-2 plasmids were co-expressed to form a hydrogenasecontaining nanoreactor based on the carboxysome shell (Shell-HydA). Figure 3b shows schematic of the Shell-HydA nanoreactor encapsulating Fd-HydA and FNR. The nanoreactors were tested for hydrogen production activity using endogenous reduced nicotinamide adenine dinucleotide phosphatase (NADPH) in cells as the electron donor for in vivo assays, and for in vitro assays, using methyl viologen (MV ⁺ ) as an electron donor, which were chemically reduced by sodium dithionite. Figure 3c shows a Western blot of E. coli expressing the hyd operon alone or Shell-HydA confirming the presence of Fd-HydA, FNR, and shell proteins in the samples purified from sucrose gradient ultracentrifugation, indicating the assembly of Shell-HydA. Figure 3d shows transmission EM of empty shells (left) and Shell-HydA (right). Yellow arrows indicate cargo proteins (Fd-HydA, FNR) in the luminal side of the shell. Figure 3e shows in vivo hydrogenase activity assays. H2 production (nmol L’ ¹ h’ ¹ ) of E. coli cells expressing free HydA or Shell-HydA grown under anaerobic (left) or aerobic (right) conditions was measured using gas chromatography. Error bars represent standard errors of the mean of three biological replicates. * p = 0.0352, ** p = 0.0018.

Figure 4 shows in vitro hydrogen production of the carboxysome shell-based nanoreactor. Figure 4a shows hydrogen production activity (nmol-mg ^-1 -min ^-1 ) of isolated free HydA and Shell-HydA at pH 8 as a function of different concentrations of MV ⁺ as the electron mediator reduced by sodium dithionite, fitted with Michaelis-Menten kinetics. Error bars show standard deviations of the mean of three biological replicates. Figure 4b shows kinetic hydrogen production of free HydA and Shell-HydA using 50 mM DT reduced MV ⁺ as electron mediator at pH 8. Error bars show standard deviations of the mean of three biological replicates. Figure 4c shows relative activity of free HydA and Shell-HydA after oxygen exposure for 24 hours at 4 °C, as a relative percentage of total activities measured under anaerobic conditions. Error bars represent standard errors of the mean of three biological replicates, *** p = 0.0009.

Figure 5 shows the characterization of the internal pH and permeability of the carboxysome shell. Figure 5a shows an encapsulated fluorescent reporter (pHluorin2) within the synthetic shells. Figure 5b shows electron microscopy images of the isolated shells confirming the incorporation of pHluorin2 into the engineered shells. Figure 5c shows determination of the pH value within the shell as 7.60, significantly lower than the pH in the cellular cytosol environment of 7.96.

Figure 6 shows a bioreactor comprising the shell and encapsulated [NiFe]-hydrogenase. Figure 6a shows a schematic representation of the bioreactor. Figure 6b shows the isolated shell-based reactors and Figure 6c shows a comparison of cells producing [NiFe]- hydrogenase versus those expressing free [NiFe]-hydrogenase.

Figure 7 shows engineered functional carboxysomes to enhance plant photosynthesis and agricultural yields. Figure 7a shows the genetic constructs generated to produce carboxysome proteins and developed plant engineering strategies to transform the constructs into the model crop plant, tobacco (Figure 7b). Figure 7c shows EM images confirming the formation of carboxysome structures in plant chloroplasts. Figure 7d shows isolated formed carboxysomes observed by EM. These carboxysomes engineered in tobacco leaves have similar structures and CC>2-fixing activity as the natural carboxysomes in cyanobacteria. Figure 7e shows the potential for enhancing crop yields and food production.

Figure 8 shows the generation and characterisation of simplified carboxysomes for the host E. coli according to an aspect of the present invention. Figure 8a shows the minimal components and a structural model of simplified Halothiobacillus neapolitanus carboxysomes. Figure 8b shows an electron microscopy image of simplified carboxysomes showing the native-like polyhedral shape and size (-100 nm). Figure 8c shows a Rubisco assay revealing that the recombinant simplified carboxysomes possess carbon fixation activities.

DETAILED DESCRIPTION

The present invention provides gene operons created using a synthetic biology approach and expression of the synthetic operons in Escherichia coli, resulting in generation of large and robust carboxysome shells with a diameter of -100 nm. The heterologously engineered shells possess similar protein composition and architectures relative to the carboxysome shells from the native hosts, suggesting great enzyme-loading capacity. The heterologously engineered shells may be produced as empty shells, or as cargo containing shells, and in this case the cargo is preferably not a cargo found in naturally occurring carboxysomes, such as Rubisco. The present invention further provides synthetic carboxysome shells comprising a reduced variety of shell proteins as compared to naturally occurring carboxysomes but comprising a cargo and still having diameters of around 100 nm. Furthermore, a carboxysome protein fragment CsoS2-C has been identified that can act as an encapsulation peptide (EP) for recruiting external cargo proteins into the synthetic shell. This novel system, combining the synthetic large protein shell and the EP, has the potential to be utilised for developing nanobioreactors and cellular factories, which can encapsulate and protect target enzymes and molecules to underpin enzyme biocatalysis and molecule delivery in diverse biotechnological applications.

The synthetic shell system of the present invention provides several and significant advantages as compared to other self-assembling nanocompartments, such as chemical cages, peptide cages, viral capsids and ferritins. For example, their large interior volume provides an enhanced cargo-loading capacity, also the synthetic shell system can be produced in industrial microorganisms, such as Escherichia coli, in large quantities at a reduced cost. In addition, taking advantage of the naturally occurring permeability of the carboxysome shell, it possesses an extraordinary advance in selectively confining transport of metabolites. Since the shell permeability can be modified and tailored through genetic engineering, it can generate a favourable interior environment (such as 02-free and redox environment, pH) and provide specific substrates to facilitate targeted enzymatic processes. Furthermore, as the synthetic shell system of the present invention is a purely protein-based system, it presents minimal toxicity to human beings in therapeutics applications.

Repurposing carboxysome shells to build microbial cell factories has the potential to underpin enzyme packaging, with the aim of protecting enzymes and boosting enzymatic activities for a range of industrial applications. Enzymes are densely packed inside the carboxysome shell, which provides a means of concentrating and protecting enzyme proteins. The intrinsic internal microcompartment created by the natural carboxysome shell contains the reduced levels of O2, thereby beneficial to oxygen-sensitive molecules, products and pathways. Through protein engineering, the shell permeability can also be tailored to provide favourable environment for specific target molecules. Therefore, it can be modified to underpin production of bioenergy, renewable chemicals and high-value products.

The metabolic nanobioreactor of the present invention also has the potential to be installed in other host organisms to improve/optimise the metabolism of host cells. For example, it can be engineered into crop plants for enhanced plant photosynthesis and agricultural yields (up to 60% of crop productivity improvement by estimation). It thus has potential societal impacts on food production and security.

Reference herein to a “cargo moiety” is intended to include proteins, enzymes and biologically active molecules.

Reference herein to “a biologically active molecule” is intended to include a compound that exerts a direct physiological effect on a host cell and includes for example, and without limitation a colourant, a fragrance, an antibacterial agent, an antifungal agent, an antiviral agent or an oil.

Reference herein to “a metabolic nanoreactor” is intended to refer to an entity, that when installed in a host organism is capable of affecting metabolism of a host cell. For example, when engineered into crop plants it is capable of enhancing plant photosynthesis, agricultural yields and plant productivity. Materials and Methods

Generation of Constructs

The nucleotide sequence of the a-carboxysome shell operon encoding CsoS2, CsoSCA, CsoS4A, CsoS4B, CsoSIC, CsoSIA, CsoSI B, and CsoSI D was amplified from the genome of Halothiobacillus neapolitanus and was cloned into the pBAD vector (cso-1 vector). The csoS2 gene and the nucleotide sequence encoding CsoS4A, CsoS4B, CsoSIC, CsoSIA, CsoSI B, and CsoSI D were assembled into pBAD cso-2 vector). The genes encoding CsoS4A, CsoS4B, CsoSIC, CsoSIA, CsoSI B, and CsoSI D were inserted into pBAD to generate the cso-3 vector. The nucleotide sequence of the C-terminus of full- length CsoS2 was amplified from the cso-1 vector. The hydA and fd genes from Chlamydomonas reinhardtii were codon-optimized for heterologous expression in E. coli. The Fd protein was fused to the N-terminus of HydA with a 15 amino acid linker composed of GGGGSGGGGSGGGGS [SEQ ID NO: 1], The hydGX and hydEF genes from Shewanella oneidensis, which encode the maturases HydE, HydF and HydG were synthesized. The fnr gene was amplified from the genomic DNA of E. coli BL21(DE3). The fd-hydA and fnr genes were separately fused with the nucleotide sequence of EP and then ligated to pCDFDueT-1 , together with the hydGX and hydEF gene fragments, to generate the hyd vector. The gfp gene was cloned to pCDFDueT-1 to generate the pCDF-gfp vector. The gfp gene fused with the nucleotide sequence of EP was cloned into pCDFDueT-1 and in frame with the nucleotide sequence encoding 6* polyhistidine tag to create pCDF-gfp-EP. The gfp gene fused to full length csoS2 gene was inserted into pCDFDueT-1 to construct pCDF-csoS2-gfp. The gfp gene fused to the truncated CsoS2 C-terminal region was inserted into pCDFDueT-1. All of these constructs were verified by PCR and DNA sequencing and transformed into E. coli DH5a and BL21(DE3) cells. The nucleotide sequence of the genes encoding Halothiobacillus neapolitanus CbbL, CbbS, CsoS2, and CsoSIA proteins was amplified from the cso-1 vector and was assembled into the pBAD vector for the production of simplified carboxysomes comprising only two different shell proteins.

Heterogeneously generation of a-carboxysome shells

E. coli strains containing the cso-1, cso-2, or cso-3 vectors were cultivated at 37 °C in Lysogeny Broth (LB) medium containing 100 pg-mL ^-1 ampicillin. The expression of these vectors was induced by L-Arabinose (1 mM, final concentration) once the cells reached early log phase (ODeoo = 0.6). Cells were grown at 25 °C for 16 hours with constant shaking and then were harvested by centrifugation at 4,000* g for 10 minutes. The cell pellets were washed with TEMB buffer (5 mM Tris-HCI pH = 8.0, 1 mM EDTA, 10 mM MgCh, 20 mM NaHCOs) and resuspended in TEMB buffer supplemented with 10% (v/v) CelLytic B cell Lysis reagent (Sigma-Aldrich) and 1 % Protein Inhibitor Cocktail (100*) (Sigma-Aldrich). The cell suspensions were lysed by French Press (Stansted Fluid Power, UK). Cell debris was removed by centrifugation, followed by centrifugation at 50,000* g to enrich a-carboxysome shells. The pellets were resuspended in TEMB buffer and then loaded onto sucrose gradients (10-30% or 10-50%, w/v) followed by ultracentrifugation (BeckMan, XL100K Ultracentrifuge) at 105,000* g for 30 minutes. Each sucrose fractions were collected and stored at 4 °C.

Recombinant expression of GFP, GFP-EP, and CsoS2-GFP

E. coli strains containing pCDF-gfp, pCDF-gfp-EP or pCDF-csoS2-gfp were grown in LB containing 50 pg-mL ^-1 spectinomycin at 37 °C. The expression of these vectors was induced by 0.5 mM (final concentration) isopropyl p-D-thiogalactopyranoside (IPTG) once the cells reached early log phase (ODeoo = 0.6), followed by cell cultivation for 16 hours at 25 °C.

Expression of mature [FeFe]-hydrogenase and generation of a-carboxysome shells with encapsulated hydrogenases

For the expression of mature, functional [FeFe]-hydrogenases as well as Fd and FNR, E. coli cells containing the hyd vector grown in LB medium containing 0.2 mM ferric ammonium citrate and 50 pg-mL ^-1 spectinomycin was induced by the addition of 0.5 mM IPTG, as well as 0.2 mM L-cysteine and 2.5 mM sodium fumarate (final concentration) at ODeoo = 0.7-0.8. For the co-expression of a-carboxysome shells and mature hydrogenases to produce Shell- HydA, E. coli strains containing the hyd vector and the cso-1 or cso-2 vector were cultivated in LB medium containing 0.2 mM ferric ammonium citrate, 50 pg-mL ^-1 spectinomycin, and 100 pg-mL ^-1 ampicillin. The hyd expression was induced by the addition of 0.5 mM IPTG at ODeoo = 0.7. After 4-hour induction of the hyd expression, the shell expression was induced by 1 mM L-Arabinose, and cells were then grown at 25 °C for 16 hours. The hyd induction was performed before expression of shells to allow hydrogenase maturation prior to shell encapsulation. This temporally separated expression strategy ensures maturation and activation of hydrogenases prior to shell formation and encapsulation. The purification of HydA and Shell-HydA was carried out following the method for shell purification mentioned above.

Hydrogenase activity assay

For in vivo activity assay, E. coli strains were first grown aerobically at 37 °C until ODeoo reached 0.7-0.8. Cells were then transferred to falcon tubes, sealed with rubber turn-over closures (Sigma-Aldrich), and degassed by 100% nitrogen before the addition of IPTG, L- cysteine, and sodium fumarate. For aerobic treatment, culture tubes were sealed without nitrogen degassing process. L-Arabinose was added 4 hours after IPTG induction. Then, cells were grown at 25°C for 16 hours. Hydrogen produced in the culture tubes was detected by gas chromatography. 1 mL gas samples were taken with a gas-tight syringe and a sample loop was flushed (100 pL) with the sample. The sample loop was then switched and run on a Bruker 450-GC gas chromatograph. The system was equipped with a molecular sieve 13 x 60-80 mesh 1.5 m x 1/8 in. x 2 mm ss column at 50 °C with an argon flow of 40.0 mL-min- ¹ . Hydrogen was detected by a thermal conductivity detector referencing against standard gas with a known concentration of hydrogen. The hydrogen yields were normalized by the HydA content in each tested sample quantified by Immunoblot. For each experiment, at least three biological replicates were examined.

For in vitro activity assay, strains were grown aerobically at 37 °C until ODeoo reached 0.7-0.8, then degassed by 100% nitrogen before the addition of IPTG, L-cysteine and sodium fumarate. L-Arabinose was added 4 hours after IPTG induction. Cultures were then cultivated at 25 °C with constant shaking for 16 hours. Cell harvesting and protein purification were carried out in an anaerobic chamber (Don Whitley Scientific, MACS-MG-500). The in vitro hydrogen evolution assays were performed inside an anaerobic glove bag (The Versatile AtmosBag, Sigma-Aldrich) flushed with 100% N2 before measurements. In vitro activity assay of free HydA was performed as a control for comparison with the hydrogen evolution activity of Shell-HydA.

For in vitro hydrogenase kinetics assays, the protein amount of HydA in the samples containing free HydA or Shell-HydA were quantified by immunoblot using purified HydA as the reference. Then, samples (0.5 mL, 10-15 mg-mL' ¹ ) containing equal amount of HydA in TMB buffer (5 mM Tris-HCI pH 8.0, 10 mM MgCL and 20 mM NaHCOs) were mixed with 100% nitrogen degassed MV (0-200 mM, final) and sodium dithionite (500 mM, final) in serum vials (Agilent Technologies) inside the anaerobic glove bag. The vials were incubated at 37 °C for 100 minutes with constant shaking and were then assayed for hydrogen production. Hydrogenase activity at a range of MV concentrations was plotted and fit using a standard Michaelis-Menten model. In addition, hydrogen evolution of free HydA and Shell- HydA at 50 mM MV and 500 mM sodium dithionite was measured every 20 minutes using gas chromatography. For each experiment, at least three biological replicates were examined.

Rubisco activity assay Isolated carboxysomes were diluted to 1 mg -ml' ¹ protein concentration by Bradford Assay using Rubisco assay buffer (100 mM EPPS, pH 8.0; 20 mM MgCh). The samples were then added into scintillation vials containing NaH ¹⁴ COs final concentration 25 mM and incubated at 30°C for 2 mins before the addition of D-ribulose 1 ,5-bisphosphate sodium salt hydrate (RuBP, Sigma Aldrich, US) with a final concentration of 1 mM. The reaction ensued for 5 mins before being terminated by adding 2:1 by volume 10% formic acid. Samples were dried for at least 30 mins at 95 °C to remove unfixed ¹⁴ C before re-suspending the fixed ¹⁴ C pellets with ultra-pure water and adding 2 ml of scintillation cocktail (Ultima Gold XR, Perkin Elmer, US). Radioactivity measurements were then taken using a scintillation counter (Tri-Carb, Perkin Elmer, US). Raw readings were used to calculate the amount of fixed ¹⁴ C, and then converted to the total carbon fixation rates. Results are presented as mean ± standard deviation (SD) using at least three biological replicates.

Oxygen exposure treatment

The HydA and Shell-HydA samples (0.5 mL, 10-15 mg-mL' ¹ ) used in the in vitro activity assays were exposed to the air for 24 hours at 4°C and were then sealed with rubber turnover closures, followed by 100% nitrogen degassing. The degassed samples were mixed with MV (50 mM, final) and sodium dithionite (500 mM, final). The vials were incubated for 16 hours at 37 °C with constant shaking and were then assayed for hydrogen production. All buffers used in the experiments were pre-degassed by 100% nitrogen. For each experiment, at least three biological replicates were examined.

Mass spectrometry

The shell samples collected from sucrose fractions were washed with PBS buffer and were treated as previously described for mass spectrometry analysis. Data-dependent LC-MS/MS analysis was conducted on a QExactive quadrupole-Orbitrap mass spectrometer coupled to a Dionex Ultimate 3000 RSLC nano-liquid chromatograph (Hemel Hempstead, UK). The raw data file was imported into Progenesis QI for Proteomics (Version 3.0 Nonlinear Dynamics, Newcastle upon Tyne, UK, Waters Company). Peak picking parameters were applied with the sensitivity set to maximum and features with charges of 2 ⁺ to 7 ⁺ were retained. A Mascot Generic File, created by Progenesis, was searched against the H. neapolitanus carboxysome protein database from UniProt.

SDS-PAGE and immunoblot analysis

SDS-PAGE and immunoblot examination were performed following the previously described procedure, 30 pg of total protein was loaded into each well. Immunoblot analysis was performed using primary mouse monoclonal anti-His (Life Technologies), rabbit polyclonal anti-CsoS1 (Agrisera), and horseradish peroxidase-conjugated goat anti-mouse IgG secondary antibody (Promega) and anti-rabbit IgG secondary antibody (GE Healthcare). Signals were visualized using a chemiluminescence kit (Bio-Rad). Immunoblot protein quantification was performed using Imaged. For each experiment, at least three biological repeats were examined.

Dynamic light scattering analysis

1 .5 mL (20 mg-mL' ¹ total protein) of individual sucrose fractions (10%, 15%, 20%, 25%, 30%) containing shell particles were analyzed by Dynamic light scattering (DLS ZetaSizer) to measure the size distribution and average size of the shells. For each experiment, at least three biological repeats were examined.

Transmission Electron Microscopy

Thin-section EM was performed to visualize the reconstituted shell structures in E. coli strains. Isolated shell structures were characterized using negative staining EM. Images were recorded using an FEI Tecnai G2 Spirit BioTWIN transmission electron microscope equipped with a Gatan Rio 16 camera. Image analysis was carried out using Imaged software. Statistical analysis was performed using Student’s t-test.

Confocal microscopy

Overnight induced E. coli cells were immobilized by drying a droplet of cell suspension onto LB agar pads. Blocks of agar with the cells absorbed onto the surface were covered with a coverslip and placed under the microscope. Laser-scanning confocal fluorescence microscopy imaging was performed on a Zeiss LSM780 confocal microscope with a 63x/1.4 NA oil-immersion objective with excitation wavelength at 488 nm and emission at 520 nm. Live-cell images were recorded from at least five different cultures. All images were captured with all pixels below saturation. Image analysis was carried out using Imaged.

EXAMPLE 1

According to the types of the encapsulated Rubisco, carboxysomes can be categorized into a-carboxysomes and p-carboxysomes. Distinct from the “inside out” de novo assembly of p- carboxysomes, the assembly of a-carboxysomes appears to start from shell formation or a simultaneous shell-interior assembly, highlighting the possibility of generating and reprogramming entire a-carboxysome shells.

The a-carboxysome proteins of the chemoautotrophic bacterium Halothiobacillus neapolitanus are mostly encoded by genes that are located in a single cso operon, including cbbL and cbbS that encode the large and small subunits of Rubisco, respectively, and genes that encode shell proteins (Fig. 1 b). In addition, csoSID is ~11 kbp downstream of the cso operon and encodes pesudohexameric shell proteins that are most likely responsible for controlling passage of metabolite molecules. We generated the synthetic cso- 1 and cso-2 operons, which were modified based on the native H. neapolitanus cso operon, to heterologously express a-carboxysome shell proteins in E. coli BL21(DE3) (Fig. 1 b). The cso-1 operon contains the genes encoding carboxysome shell proteins (csoS2, csoS4AB, csoSICAB, csoSID), as well as csoSCA that encodes p-carbonic anhydrase (CA). The cso- 2 operon comprises only the shell protein-encoding genes (csoS2, csoS4AB, csoSICAB, and csoSID), without csoSCA Expression of the cso-1 and cso-2 operons both led to production of highly ordered shell architectures in the E. coli hosts, as exhibited by thin- section electron microscopy (EM) (Fig. 1c). Sucrose gradient ultracentrifugation and SDS- polyacrylamide gel electrophoresis (SDS-PAGE) indicated the presence of major shell proteins consisting of CsoSIC, CsoSIA, CsoSI B, and CsoS2 (exists as two isoforms CsoS2A and CsoS2B, see detailed description below) enriched in the 20% sucrose fraction (Fig. 1d). Mass spectrometry further showed the presence of all seven shell proteins and CA encoded by the cso-1 operon and seven shell proteins encoded by the cso-2 operon in the 20% sucrose fraction (Table 1), confirming the self-assembly of expressed carboxysome shell proteins to form shell supramolecular structures.

Negative-staining EM of the 20% sucrose fraction shows that the recombinant a- carboxysome shells exhibit a polyhedral shape, with a diameter of 80-110 nm (cso-1 shell: 85.93 ± 15.74 nm; cso-2 shell: 97.12 ± 20.84 nm; n = 150) (Fig. 1e), resembling the native a-carboxysomes from H. neapolitanus. This shows that empty a-carboxysome shells can be constructed in the non-native host E. coli, by expressing only shell proteins without cargo proteins, consistent with in vivo observations. This in turn implies the specific assembly pathway of a-carboxysomes, either “shell first” or “concomitant shell-core assembly”, is distinct from the “inside-out” mode of p-carboxysome biogenesis.

Without cargos, the shells in the same sucrose fraction exhibit variable structures (Fig. 1e). Likewise, the shell size varies among shell structures collected from different sucrose fractions. Dynamic Light Scattering (DLS) analysis revealed that the shell size gradually increases from 10% to 30% sucrose density (cso-1 shell: from 82.8 nm to 104.3 nm; cso-2 shell: from 82.1 nm to 147.8 nm), in agreement with EM results.

Results also showed that stable polyhedral shells can be formed in the absence of CA, implicating that CA is not an essential component for shell assembly. This is consistent with previous results, which illustrated that a-carboxysomes can be assembled in H. neapolitanus without CA. Moreover, the shells lacking CA are on average larger than their counterparts containing CA from individual sucrose fractions (Fig. 1e), suggesting that CA plays a role in defining the shell architecture, likely by tight association with shell proteins. The effect of CA on shaping the shell structure is restricted in the shells that possess the minimal size (~82 nm, 10% sucrose).

EXAMPLE 2

In the native a-carboxysome, CsoS2 is an intriguing disordered protein with relatively high content and appears as two distinct isoforms in H. neapolitanus'. the long form CsoS2A (-130 kDa) and the short form CsoS2B (-85 kDa). CsoS2 interacts with shell proteins and its N-terminus is crucial for organising Rubisco inside the a-carboxysome, suggesting the important role of CsoS2 in carboxysome assembly, which is functionally analogous to CcmM, the linking proteins of p-carboxysomes for bridging cargo and the shell. To verify the necessity of CsoS2 in the assembly of empty a-carboxysome shells, a cso-3 operon was generated by deleting csoS2 from the cso-2 operon (Fig. 2a). Thin-section EM showed that no shell structures were discerned in the cso-3 E. coli cells, in contrast to the EM results of the cso-2 cells (Fig. 2b). Non-assembled shell proteins were prone to form aggregates at the pole of the cso-3 cell (Fig. 2b, orange arrow). SDS-PAGE analysis of the pellets of cell extracts after 50, 000* g centrifugation showed that shell structures were formed only in the presence of CsoS2 (as indicated by the presence of shell proteins CsoS1A/C and CsoSI B) (Fig. 2c). To further examine the CsoS2-mediated formation of a-carboxysome shells, enhanced green fluorescence protein (GFP) was fused to the C-terminus of full-length CsoS2. Without CsoS2 or shell proteins, GFP signal was evenly distributed throughout the cytosol of E. coli (Fig. 2d, top and middle). When CsoS2-GFP and cso-3 shells were coexpressed, spotty fluorescence signal was visualized in the cell’s cytosol (Fig. 2d, bottom), signifying the formation of highly ordered shell structures. Collectively, our results revealed that CsoS2 is essential for the formation of empty a-carboxysome shells.

EXAMPLE 3

Given the necessity of CsoS2 in shell formation and the binding of CsoS2 N-terminus with Rubisco, we speculate that the C-terminal fragment of CsoS2 (278 amino acids) links the shell and cargo assemblies, and thereby, may serve as the encapsulation peptide (EP) to direct cargo enzymes into the a-carboxysome shell. To verify this hypothesis, GFP was fused to the N-terminus of the CsoS2 C-terminal fragment and co-expressed with the cso-2 shells. Confocal images showed that the cells expressing only GFP, GFP-EP, or cso-2 shells and GFP presented diffuse fluorescence signal throughout the cell, whereas the cells expressing cso-2 shells and GFP-EP exhibited explicitly dispersed fluorescent foci (Fig. 2e), demonstrating the EP-mediated cargo encapsulation into the synthetic shell. Recombinant cso-2 shells with incorporated GFP-EP were purified from E. coli by sucrose gradient ultracentrifugation. Immunoblot analysis revealed that the shells encapsulating GFP-EP were detected in the 10%-50% sucrose fractions, with the strongest GFP signal observed in the 20% sucrose fraction (Fig. 2f). Without shells, by contrast, free GFP-EP were only detected at the top of the sucrose gradient (Fig. 2f), consistent with the quantification of GFP fluorescence in these sucrose fractions. EM of the 20% sucrose fraction revealed the incorporation of GFP-EP in the shell interior, and that the size of the GFP-encapsulated shells is relatively comparable to that of the empty shells (Fig. 2g). These results demonstrate that the C-terminal region of CsoS2 directly interacts with the shell, consistent with the previous finding, and this peptide could serve as an EP to recruit external proteins into the empty shell.

The C-terminal region of CsoS2 has three repeats (R1-R3) and a conserved C-terminal domain. To determine the roles of these domains in cargo encapsulation, we systematically generated the GFP-tagged CsoS2 C-terminus variants that differ in the number of the repeats and the C-terminal peptide. Coexpressing these GFP-CsoS2 C-terminus peptides with the cso-2 shells and confocal images showed that the punctate fluorescence signal relative to the cytosolic fluorescence reduced along with the decrease in the number of the C-terminal repeats. The C-terminal peptide alone was incapable of mediating incorporation of GFP into the shell. These results indicate that the R1 R3 repeats of the CsoS2 C-terminal region are important for cargo encapsulation, probably by interacting with shell proteins; more repeats ensure higher efficiency of cargo encapsulation. Additionally, removal of the C-terminal peptide from the CsoS2 C-terminus (only R1R2R3) has no significant effects on shell formation and GFP incorporation, supporting the assumption that the C-terminal peptide may be exposed on the outside of the carboxysome shell.

EXAMPLE 4

Though catalytically efficient for hydrogen evolution, [FeFe]-hydrogenases are highly sensitive to O2, representing a barrier for biological H2 production. The naturally developed O2-limited microenvironment in the carboxysome shell is ideal for enhancing the catalytic activities of oxygen-sensitive enzymes. Defining the role of the CsoS2 C-terminus as the EP paves the way for generating a nanoreactor based on the intact a-carboxysome shell, which is capable of incorporating heterologous enzymes for new catalytic functions.

The [FeFe]-hydrogenase from the green alga Chlamydomonas reinhardtii, HydA, is one of the simplest [FeFe]-hydrogenases and represents the “minimal unit” for biological H2 production. We generated a hyd plasmid to express the following proteins: (1) algal HydA that are fused with algal ferrodoxin (Fd) at the N-terminus, with a 15-aa linker, and the EP at the C-terminus. Fd serves as the native electron donor of algal HydA, and the Fd-HydA fusion increased the rate of H2 production; (2) E. coli ferrodoxin: NADP ⁺ -oxidoreductase (FNR) fused with EP at the C-terminus. FNR originally from E. coli could catalyse the transfer of electrons from NADPH to Fd; (3) the maturase enzymes HydE, HydF, and HydG (Fig. 3a), which are crucial for the formation and activation of HydA. In addition, polyhistidine tags (His-tags) were linked to the N-terminus of Fd-HydA and the C-terminus of FNR to facilitate biochemical identification. This plasmid and cso-2 were both transformed into E. coli (Fig. 3a). Expression of the hyd plasmid was induced by addition of isopropyl p-D-1- thiogalactopyranoside (IPTG) for 4 hours before the expression of the shell induced by L- arabinose. This temporally separated expression strategy ensures maturation and activation of hydrogenases prior to shell formation and encapsulation. During shell assembly, EP fusion to both Fd-HydA and FNR directs packaging of Fd, HydA, and FNR into the shell interior, allowing for installation of a functional electron transfer pathway into the a- carboxysome shell and generation of a hydrogenase-containing nanoreactor (Shell-HydA) for consecutive H2 production using electrons from NADPH (Fig. 3b).

After 50,000x g centrifugation, the components of Shell-HydA including shell proteins and HydA were detected in the pellet, whereas unencapsulated HydA were only present in the supernatant, implicating the formation of Shell-HydA assemblies. Similar to the shells containing GFP-EP (Fig. 2f), Shell-HydA were predominantly enriched in the 20% sucrose fraction, whereas expressed HydA in the absence of shells were not detectable in the 10- 50% sucrose fractions (Fig. 3c). EM of the Shell-HydA assemblies in the 20% sucrose fraction revealed the incorporation of HydA into the recombinant carboxysome shells mediated by the EP (Fig. 3d, orange arrows).

The H2 production activity of the generated E. coli cells expressing Shell-HydA and free HydA were assayed using endogenous NADPH in the cells as the electron source. Cells were cultivated and induced for 16 hours in either aerobic or anaerobic conditions, and the produced H2 was measured by gas chromatography. Under anoxic conditions, cells expressing Shell-HydA had a -20% greater H2 yield than those expressing free HydA, standardized by the amount of HydA (Fig. 3e). The discrepancy in H2 evolution between the E. coli hosts producing Shell-HydA and free HydA is explicitly more pronounced under aerobic conditions. Cells producing Shell-HydA yielded -4.1 times more H2 than those expressing free HydA (Fig. 3e). The results strongly indicate that the Shell-HydA nanoreactors are catalytically active for H2 production, and encapsulation of the carboxysome shell is beneficial to the H2 production activity of the hydrogenases, by concentrating enzymes and substrates and reducing oxygen levels in the shell lumen taking advantage of the shell semi-permeability. No notable difference in the H2 production activity between HydA encapsulated inside the shell with CA and without CA, under either anaerobic or aerobic conditions was observed.

Furthermore, the Shell-HydA catalysts were anaerobically purified for in vitro activity assays using methyl viologen (MV) as an electron donor, which were chemically reduced by sodium dithionite. In the assays, the HydA content was measured using purified HydA as the reference. The maximum hydrogen evolution rate of Shell-HydA is 603.33 ± 44.87 nmol-mg" ¹ -min" ¹ at pH 8, ~5.5-fold greater than that of free HydA (109.32 ± 29.28 nmol-mg" ¹ -min" ¹ ) (Fig. 4a), comparable with the results of in vivo activity assays (Fig. 3e). The amount of hydrogen produced by Shell-HydA measured at 50 mM MV increases linearly over time, indicating that the process is indeed catalytic (Fig. 4b). Moreover, Shell-HydA produces remarkably more hydrogen under these conditions than free HydA, emphasising the advantage of the carboxysome shell-organized nanoreactor in enhancing the H2 production activity of encapsulated hydrogenases.

To further investigate the O2 tolerance of the shell-based nanoreactor, the Shell-HydA catalysts purified under anaerobic conditions were exposed to the air for 24 hours followed by activity assays. After oxygen exposure, Shell-HydA retained 13.3 ± 2.1 % activity relative to the activity measured under anaerobic conditions (Fig. 4c). By contrast, only 0.97 ± 0.1% activity was maintained for free HydA exposed to O2. The results support the hypothesis that the carboxysome shell could act as a selective barrier to efficiently scavenge oxygen, providing elevated O2 resistance of this hydrogenase-containing nanoreactor. Additionally, it was proposed that the carboxysome shell allows passage of negatively charged metabolites and protons across the shell. The H2 production experiments suggested that the positive protons and negative NADPH, as well as the catalytic product H2, can diffuse across the shell (Fig. 3b). Precisely how the carboxysome shell excludes O2 diffusion merits further investigation.

EXAMPLE 5

Using the constructs and procedures described herein before to produce empty carboxysome shells for enzyme encapsulation, we focused on the characterisation of the internal environment and permeability of the shell. We encapsulated a fluorescent reporter pHluorin2 within the synthetic shells (Fig. 5a). EM images of the isolated shells have confirmed the incorporation of pHluorin2 into the engineered shells (Fig. 5b). This reporter protein is a ratio pH-sensitive green fluorescent protein (GFP) with a pH- dependent bimodal excitation spectrum. The ratio of the fluorescence peak values at 392 nm and 470 nm is used to measure the local pH conditions in solution. Based on the spectral feature of pHluorin2 and the capacity of the engineered shells in protein encapsulation, we have determined the pH value within the shell is 7.60, significantly lower than the pH in the cellular cytosol environment of 7.96 (Fig. 5c). We further proved that the lower pH of the shell interior is due to the specific permeability of the shell to bicarbonate (HCOT) and protons. These findings advanced our understanding of the internal environment and permeability of the synthetic shell. It also demonstrated the reproducibility of generating this bio-factory and its capacity of encapsulating external enzymes.

EXAMPLE 6

We have also encapsulated another type of hydrogenase, [NiFe]-hydrogenase, into the shell (Fig. 6a and Fig. 6b). Compared to the [FeFe]-hydrogenase described above, [NiFe]- hydrogenase is more oxygen-tolerant and stable. Cells producing Shell+[NiFe]-hydrogenase yielded ~11 times more H2 than those expressing free [NiFe]-hydrogenase (Fig. 6c). These results show the potential utility of the shells of the present invention encapsulating enzymes to boost enzyme activity and enhance bioenergy production.

EXAMPLE 7

Carboxysomes are natural CC>2-fixing machinery in all cyanobacteria and many microorganisms. It encapsulates the CC>2-fixing enzyme Rubisco into the carboxysome shell to improve Rubisco carboxylation and carbon assimilation. Taking advantage of this structural feature, cyanobacterial carboxysomes have an improved capacity of carbon fixation than crop plant Rubisco enzymes (carboxysomes are not present in any crop species). Therefore, engineering carboxysomes into plants provide great promise for enhancing crop photosynthesis and productivity.

We generated the genetic constructs to produce carboxysome proteins (Fig. 7a) and developed plant engineering strategies to transform the constructs into the model crop plant, tobacco (Fig. 7b). EM revealed that the production of carboxysome proteins resulted in the formation of carboxysome structures in plant chloroplasts (Fig. 7c), where photosynthetic carbon fixation occurs. We further isolated the formed carboxysomes and characterised them using EM. These carboxysomes engineered in tobacco leaves have similar structures and CC>2-fixing activity as the natural carboxysomes in cyanobacteria (Fig. 7d). This embodiment demonstrated that the carboxysome structures can be reconstituted in not only bacterial cells, but also eukaryotic systems, such as plants. It also provides a novel scientific solution for enhancing crop yields and food production in the long term (Fig. 7e).

EXAMPLE 8

Ribulose-1 ,5-bisphosphate carboxylase/oxygenase (Rubisco) is the most abundant protein on Earth, and is a key enzyme for CO2 fixation. However, Rubisco is an inefficient enzyme, and the catalytical reactions of Rubisco are essentially the limiting steps in photosynthetic CO2 fixation. This is mainly because of its poor capability in discriminating between the competing substrates, CO2 and O2. The fixation of O2 leads to a process termed photorespiration, which wastes energy and decreases sugar synthesis. To overcome the inherent limitations of Rubisco, carboxysomes in all cyanobacteria and many chemoautotrophs use a protein shell to encapsulate Rubisco enzymes. The permeability of shells allows a high-CC>2 level around the enclosed Rubisco to enhance Rubisco carboxylation. Given the substantial contributions of carboxysomes in global carbon cycle, introducing carboxysomes into plants has attracted increasing attention in recent years. However, engineering a full set of bacterial carboxysomes genes into an eukaryotic system and ensuring their proper expression are technically challenging. Thus, there is a pressing need to optimise the genetic constructs and engineer simplified, functional carboxysomes.

To achieve this, we tested the minimal requirement of building active carboxysomes in a E. coli system. We assembled four genes that encode shell proteins CsoSIA and CsoS2 along with Rubisco enzymes CbbLS into one expression vector (Fig. 3A), and allowed E. coli to produce these proteins in cells. We then isolated these formed protein assemblies and characterised them using electron microscopy and enzyme assays. Electron microscopy showed that the production of carboxysome proteins resulted in the formation of carboxysome structures (namely simplified carboxysomes) with a native-like structure (-100 nm) (Fig. 3B). Rubisco assays confirmed that these simplified carboxysomes have the CO2- fixing activities (Fig. 3C).

This study demonstrated that based on shell structure engineering, we can further modify the carboxysome composition to build desired carboxysome structures, highlighting the amendability and modularity of carboxysomes.

The simplified carboxysomes are composed of four families of components, instead of nine as in the native carboxysomes, and are thus ideal systems in bioengineering of CC>2-fixing factories for biotechnological applications, for example, introducing simplified carboxysomes in plants to enhance plant photosynthetic carbon assimilation and crop yields in the long term.

In summary, we generated and characterized large and robust carboxysome shells with a diameter of -100 nm. We also demonstrated the feasibility of reprograming the synthetic protein cages, by sequestering a catalytically active, ^-producing pathway within the shell, to boost production of hydrogen. The heterologously engineered shells possess similar protein composition and architectures relative to the carboxysome shells from the native hosts, suggesting the great capacity of cargo loading. We provide evidence that CsoS2 is essential for the assembly of the empty shell, and the C-terminus of CsoS2 could serve as an EP for recruiting external cargo proteins into the synthetic shells. Taking advantage of the defined EP and self-assembly of the carboxysome shells, we installed active hydrogenases and Fd from green algae together with the E. coli FNR inside the empty shell to create a large proteinaceous nanoreactor for hydrogen production. The shell encapsulation and the specific microenvironment created in the engineering nanoreactor were demonstrated to favour the catalytic activity of oxygen-sensitive [FeFe]-hydrogenases and improve O2 tolerance of the generated catalysts significantly. Advanced understanding of the assembly principles of carboxysomes and shells, and the developed engineering systems for precisely tuning enzyme activation, shell formation and encapsulation will inform rational design and engineering of carboxysome shell-based nanomaterials and nanostructures in biotechnological applications for catalytic enhancement, enzyme protection, and molecule delivery. We also see the potential to combine these biological assemblies with abiotic cocatalysts in the future. We also generated and characterised carboxysome shells having a diameter of around 100 nm but comprising only two different shell proteins and having the ability to encapsulate Halothiobacillus neapolitanus Rubisco.

Throughout the description and claims of this specification, the words “comprise” and “contain” and variations of them mean “including but not limited to”, and they are not intended to (and do not) exclude other moieties, additives, components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

Features, integers, characteristics, compounds, chemical moieties or groups described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and/or steps are mutually exclusive. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed. The reader's attention is directed to all papers and documents which are filed concurrently with or previous to this specification in connection with this application and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference.