Microbial Protein Compartments

Microbial compartmentalization strategies

Biological cells employ compartmentalization to overcome many difficult metabolic and physiological challenges. Eukaryotes mainly use membrane-bound organelles to sequester and control the flow of metabolites, store genetic information and segregate protein processing and export. In contrast, the majority of prokaryotes do not possess intracytoplasmic membrane systems and instead rely on protein-based approaches to achieve spatial control. However, certain specialized membrane compartments have been identified in specific bacterial strains, including polyphosphate-storing acidocalcisomes, photosynthetic thylakoids in cyanobacteria, magnetosomes in magnetotactic bacteria and pirellulosomes and anammoxosomes in morphologically complex Planctomycetes (Figure 1). Important examples for protein-based organelles include bacterial microcompartments (BMCs) like the CO2-fixing carboxysome, ferritins involved in iron homeostasis and maintaining redox balance and functionally diverse encapsulin nanocompartments (Figure 1). Encapsulating enzymes or biosynthetic pathways in semi-permeable protein organelles can increase the local concentrations of metabolites and enzymes, prevent the loss of toxic or volatile intermediates and create unique microenvironments necessary for the proper functioning of specialized enzymes. In addition, encapsulation allows for incompatible reactions and processes to take place in a single cell at the same time.

Figure 1: Overview of protein- and membrane-based microbial compartments.

Microbial Protein Compartments – Discovery and Basic Science

Genome-mining of new microbial protein compartments – Encapsulins and beyond

Encapsulin nanocompartments are a relatively new class of microbial protein compartments that assemble into T=1 (60 subunits, 20-24 nm) or T=3 (180 subunits, 30-40 nm) icosahedral hollow capsids. Their key feature is that they specifically encapsulate cargo proteins, which are targeted to the capsid interior via small C-terminal peptides referred to as targeting peptides (TPs). We recently systematically explored the distribution and diversity of encapsulin gene clusters in bacterial and archaeal genomes using a combination of bioinformatics, phylogenetic analysis and biochemistry. We identified more than 900 putative encapsulin systems in bacterial and archaeal genomes (Figure 2). Encapsulins can be found in fifteen bacterial and two archaeal phyla. Our analysis revealed one new capsid type and nine previously unknown cargo proteins targeted to the interior of encapsulins. We experimentally characterized three newly identified encapsulin systems and illustrated their involvement in iron mineralization, oxidative and nitrosative stress resistance and anaerobic ammonium oxidation, a process responsible for 30% of the N lost from the oceans (Figure 3). We now continue to biochemically, structurally and functionally characterize newly identified encapsulin systems in both a heterologous and native context. We will study systems of the newly identified cargo type that might incorporate a gated pore in the encapsulin capsid and might be involved in lipid or isoprenoid metabolism, different unusual systems likely involved in iron metabolism as well as encapsulin-associated components that are not targeted for encapsulation but important for the proper function of a given encapsulin system. In addition, we are continuing our genome-mining for new protein-based compartments in microbes. We have already identified a number of other putative phage-based systems in diverse microbes and will investigate their function at a biochemical, structural and functional level in the near future.


Figure 2: Distribution of encapsulins in bacterial and archaeal genomes. The number of identified encapsulin systems per phylum is show below the phylogenetic tree.

Figure 3: Genome-mining for new encapsulins and characterization of newly identified systems.

Protein compartments in anammox bacteria

In our genome-mining study, one of the most surprising and interesting findings was the presence of a conserved encapsulin operon in anammox bacteria of the phylum Planctomycetes. These chemolithotrophic bacteria can be found in the oxygen minimum zones of the oceans (OMZs) and are named for their unique energy metabolism based on the anaerobic oxidation of ammonium, responsible for ca. 30% of the nitrogen loss from the oceans globally (Figure 4). Nitrogen is of paramount importance in the oceans because primary production is generally limited by the availability of fixed (biologically available) nitrogen.  The nitrogen cycle of the ocean is predominantly microbially mediated – microbes control the total nitrogen inventory and the pathways by which fixed nitrogen is lost and gained.  Biological nitrogen fixation is the main input process, and it occurs mostly in the sunlit surface layers.  Denitrification and anaerobic ammonium oxidation (anammox) are the two processes by which nitrogen is lost. There are five candidate genera of anammox bacteria, which occur in environments ranging from wastewater treatment plants to the oligotrophic open ocean.  Mainly Scalindua-like sequences have been retrieved from the open ocean (OMZs).  Pyrosequencing of the nirS gene in samples from the Eastern Tropical South Pacific and the Arabian Sea OMZs detected one very abundant OTU that was 92% identical to the published Scalindua sequence.  16S rRNA sequences also indicate that a very homogeneous clade of close relatives of Scalindua are the predominant anammox bacteria found in the oceanic OMZs.  These data indicate that marine anammox bacteria are most similar to Scalindua but that they are not identical to Scalindua, especially at the level of functional gene sequences. The metabolism of anammox bacteria is primarily based around the oxidation of ammonium coupled to the reduction of nitrite with concomitant dinitrogen formation, using carbon dioxide as the sole carbon source. Studies have shown that nitric oxide (NO) and hydrazine are key intermediates in the anammox process. The genomes of all known anammox bacteria contain a large number (up to 11) of paralogous proteins (annotated as hydroxylamine oxidoreductases, Hao) proposed to be involved in the anammox process. Some of those enzymes have recently been characterized and shown to convert nitrogen containing compounds between different oxidation states leading to hydrazine, NO and dinitrogen. The exact function of the different Hao paralogs and how they contribute to the overall anammox process remains to be discovered.

Figure 4: The biological nitrogen cycle in the oceans.


We identified a conserved operon in anammox genomes containing an unusual encapsulin gene as well as a gene coding for a fusion protein containing nitrite reductase and hydroxylamine oxidoreductase domains (Nir-Hao) (Figure 5). Initial experiments indicate that the Nir-Hao is indeed localized to the interior of the nanocompartment. The encapsulin capsid protein is itself fused to a diheme-containing domain that lines the inner surface of the compartment. This newly identified encapsulation system likely plays a role in the anammox process. Depending on the redox-chemistry of Nir-Hao, it could be involved in intermediate recycling, the scavenging of toxic intermediates (e.g. NO, hydrazine, hydroxylamine), the formation of an anammox intermediate or the evolution of dinitrogen itself. Considering that encapsulin nanocompartments possess a selectively permeable protein shell, the encapsulation of Nir-Hao would likely lead to the concentration of reactants, prevent harmful intermediates from escaping the compartment and exclude competing side reactions. Recent studies in various encapsulin systems indicate that in addition to cargo proteins localized to the same genetic locus as the capsid protein, secondary cargo and non-targeted components dispersed throughout the genome might play important roles in encapsulin function. Understanding this surprisingly complex multi-component nanocompartment system newly identified in anammox bacteria will lead to new molecular insights into the anammox process and shed more light on one of the major processes involved in the biological nitrogen cycle. This project is currently being funded by the Gordon and Betty Moore Foundation (Figure 6) and is carried out in collaboration with the lab of Bess Ward (Department of Geosciences, Princeton University, https://www.princeton.edu/nitrogen/). Together we plan to decipher the role of encapsulins in the anammox bacteria and how wide-spread this intriguing encapsulation system is in the OMZs of the oceans.

Figure 5: Newly identified encapsulin operons in anammox bacteria.


Figure 6: Moore Foundation Grant logo (Grant number: GBMF#5506).


Microbial Protein Compartments – Synthetic Biology and Applications

Encapsulins as a general protein-based compartmentalization strategy

Compared with many natural and synthetic lipid-, polymer- and inorganics-based compartments, protein-based organelles exhibit a number of superior properties. They self-assemble in a precise fashion from one or a small number of capsid proteins leading to defined nanoarchitectures of uniform size and shape. Their biological origin makes them inherently biocompatible and allows facile genetic functionalization. In addition, the defined shell-like structure of many protein compartments enables the creation of multifunctional nanodevices by rationally modifying both their inner and outer surfaces.

Because of their unique properties, encapulins represent an attractive platform for applications ranging from biomedicine (delivery and imaging devices) and catalysis (nanoreactors and scaffolds) to materials science (biological synthesis of inorganics). The fact that encapsulins are able to naturally internalize cargo proteins and that cargo proteins carry short C-terminal targeting peptides (TPs) predisposes those systems for applications where non-native proteins, peptides or functionalizable small molecules need to be encapsulated. This fact, in combination with chemical conjugation technologies and the accessibility of both N- and C-terminus for further genetic modification makes encapsulins excellent platforms for the integration of multiple functionalities at the nanoscale. Moreover, encapsulins have been shown to be unusually stable over a large range of pH and temperature.

TPs are extraordinarily efficient and are able to function in both prokaryotic and eukaryotic heterologous hosts as well as in vitro. Thousands of cargo proteins carrying TPs have been identified in our lab and can be divided into at least 3 distinct classes. One of the main goals of artificial compartments is to encapsulate multiple components in a defined stoichiometry which would allow the precise engineering and tuning of particular functionalities. We have discovered natural encapsulin systems consisting of up to four targeted components and have shown that multiple non-native components can be efficiently encapsulated in heterologous hosts. Through both screening of naturally occurring TPs and mutagenesis we aim to generate a library of TPs with distinct targeting strengths. This toolbox will then be used to rationally assemble artificial compartmentalization systems imbued with defined functions.

A second important goal of engineering compartments is controlling the transport of metabolites across the semi-permeable compartment barrier. Encapsulins possess 1-3 nm pores at the points of 5- and 6-fold symmetry which are thought to allow specific components related to a particular function to selectively access the compartment interior. We are working on using directed evolution and rational engineering approaches to engineer encapsulin pores in terms of size, charge and hydrophobicity. This will enable us to control pore selectivity, allowing certain metabolites to access the compartment while excluding unwanted molecules. A number of naturally occurring encapsulin systems have been proposed to assemble into compartments containing gated pores. We are also working on characterizing those systems with the goal of incorporating them in artificial engineered compartmentalization systems leading to compartments whose permeability to specific metabolites could be controlled via the metabolic state of the host or specific small molecule triggers.

A minimal CO2-fixing organelle for increased plant photosynthetic efficiency

In recent decades, population growth has outpaced improvements in agricultural productivity with conservative projections estimating a world population of 9.7 billion in 2050. Continuously rising CO2-levels in the atmosphere leading to accelerated anthropogenic climate change will result in increased flooding, loss of agricultural land and droughts. Thus, there is an urgent need to improve agricultural productivity beyond yield potentials achievable by traditional methods. In plants, photosynthetic efficiency in terms of light energy converted to biomass is only ca. 1% and has been identified as one of the most promising targets for improving agricultural productivity.

Compared with plant ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO), the key CO2-fixing enzyme, cyanobacteria have evolved faster enzymes that are more sensitive to O2. What makes cyanobacteria nonetheless thrive in high oxygen environments is their sophisticated CCM. A core component of this CCM is a bacterial microcompartment called the carboxysome. RuBisCO and carbonic anhydrase (CA) are encapsulated inside these protein organelles leading to a high local concentration of CO2 around the RuBisCO active site. The selectively permeable protein shell of the carboxysome minimizes internal O2 concentrations thus preventing wasteful photorespiration.

It has been estimated that transplanting a functional cyanobacterial CCM to C3 plants would lead to an increase in CO2-assimilation yield of 36-60%. The main challenge with transferring functional cyanobacterial carboxysomes to plant chloroplasts is the fact that between 10 and 15 genes would need to be targeted to chloroplasts to assemble functional carboxysome compartments. In addition, carboxysome assembly and activity is known to be highly dependent on the correct relative expression levels of various matrix and shell components.

Our strategy to circumvent these major challenges is to construct an artificial minimal CO2-fixing protein organelle based on encapsulin nanocompartments (Figure 7). The ability to specifically target non-native proteins in heterologous hosts using TPs would be used to encapsulate RuBisCO and CA in T=3 encapsulin capsids. The relative stoichiometry of the two targeted components could be controlled by using TPs with differential targeting strengths and by controlling relative expression levels. Directed evolution and targeted mutagenesis of encapsulin pore residues would be used to optimize O2 exclusion and CO2 retention, minimizing photorespiration and maximizing RuBisCO catalytic efficiency, respectively. The same approaches would also be used to guarantee optimal access of ribulose-1,5-bisphosphate, the substrate of RuBisCO, to the compartment interior. By using fusions of the RuBisCO small and large subunits which were shown to maintain near wild-type levels of catalytic activity, CA and a single encapsulin capsid protein, only three genes would be enough to assemble a functional carboxysome-mimicking CO2-fixing organelle in plant chloroplasts. To complete a functional cyanobacterial CCM, two additional issues will have to be addressed. Bicarbonate transporters need to be installed in the inner chloroplast membrane and CA activity has to be minimized in the chloroplast lumen. Combined with an encapsulin-based system to decrease photorespiration and increase RuBisCO catalytic activity, these changes would result in an up to 60% increase in CO2-assimilation yield of important C3 crops like rice, wheat, barley and soybean.

In addition, due to the simple 3-gene architecture of the proposed system, non-autotrophic organisms could be engineered to directly utilize CO2 as a supplementary carbon source. This engineered artificial autotrophy might positively benefit the fermentative production of many compounds. The amount of carbon source (sugar) could be reduced while at the same time consuming a potent greenhouse gas (CO2).

We have already established the feasibility of packaging RuBisCO and CA into encapsulins in vivo and have shown that RuBisCO maintains CO2-fixing activity. We are now working on investigating how co-localization of RuBisCO and CA influences carbon fixation rates, how the kinetic parameters differ between encapsulated vs non-encapsulated enzymes, on gaining fine-control over encapsulated protein stoichiometries and on engineering encapsulin pores for improved metabolic flux across the compartment shell.

Figure 7: Our approach to construct a minimal synthetic CO2-fixing organelle based on encapsulins.  

Engineering protein-based compartmentalization and spatial control in yeast

Different types of scaffolds and localization strategies have been used to introduce spatial control to metabolic engineering. The vast majority of work has been focused on bacteria with little work done in eukaryotes, even though the yeast Saccharomyces cerevisiae is one of the most widely used industrially relevant production hosts. Initial results indicate that encapsulin nanocompartments can be successfully assembled in S. cerevisiae and used to encapsulate non-native cargo proteins. Based on their unique properties and suitability to engineer protein-based compartmentalization, we will use encapsulins as a metabolic engineering tool in yeast. Their application will be aimed at co-localizing particular steps of a pathway of interest to the interior of the encapsulin compartment. This would lead to an increase in the effective local molarity of both enzymes and reaction intermediates leading to improved pathway flux. In addition, encapsulation would protect the rest of the cell from toxic intermediates.

This approach will prove useful in improving the yield of a number of important engineered pathways including the antimalarial drug artemisinin and medically-relevant alkaloids. Artemisinin biosynthesis is known to proceed via a toxic aldehyde intermediate which could be selectively confined inside protein compartments by targeting the respective pathway step to the inside of encapsulins. Likewise, the biosynthesis of many engineered alkaloids is thought to form reactive nitrogen-based heterocyclic intermediates. This problem could also be approached using engineered compartmentalization resulting in reduced host toxicity and thus higher yields.

Encapsulins as a platform for the biological synthesis of organic-inorganic hybrid materials

Engineered organisms are used with increasing success in diagnostics, therapeutics and for the production of high-value or commodity organics. Initial steps have been taken to expand the scope of synthetic biology towards the synthesis of functional inorganic nanomaterials. We propose to expand the use of encapsulin nanocompartments as systems for the in vivo and in vitro synthesis of organic-inorganic hybrid materials.

Using biogenic protein-based nanostructures like encapsulins as templates and catalysts for the production of inorganic nanomaterials has a number of inherent advantages: The size-range accessible via encapsulins stretches from the nano- (single isolated encapsulin) to meso-scale (higher order encapsulin assemblies) allowing the synthesis of many different nanomaterials displaying specific properties only accessible at these scales. The defined size and shape of encapsulin compartments allows for the size-constrained synthesis of inorganics in their interior, leading to a homogeneous and monodisperse population of nanosized objects. The possibility to use both the interior and exterior surfaces of shell-like nanostructures like encapsulins allows for the potential synthesis of nanomaterials that integrate multiple functionalities. And lastly, using biology for the synthesis of inorganics will decrease the amount of harmful chemicals and solvents used in inorganic syntheses as well as reduce the energy input needed.

In nanobiotechnology, inorganic nanomaterials of a defined size and shape are synthesized by combining a compartment with precursor metal salts. Targeted mineralization is initiated by either directing precursor ions to the interior of a nanostructure via charge complementarity followed by the addition of a reducing agent, or by using genetic fusions of metal-binding/precipitating proteins and peptides.

Using encapsulins will allow for the simultaneous presentation of different metal-precipitating peptides on the compartment inside and outside based on capsid protein and TP fusions. This will enable us to generate new composite materials and nano-alloys in vivo and in vitro not found in nature that could be used as inorganic catalysts, electrodes, photovoltaic and magnetic materials and photonic nano-devices. Genetically encoding the synthesis of functional nanomaterials would allow advanced applications as non-invasive biological sensing and control systems. Using biology would also enable us to combinatorially integrate mutational screens of capsid proteins and metal-binding/precipitating peptides and media compositions resulting in the high-throughput synthesis of inorganics, an approach not feasible using classical inorganic chemistry due to the extremes in temperature and pressure involved. Another distinguishing feature compared to inorganic chemistry is the ability of biological systems to generate non-equilibrium structures and materials with unique properties. Using TPs, it would also be possible to specifically embed one or multiple enzymes inside an inorganic nanoparticle of defined size leading to porous organic-inorganic hybrid-materials.

In a first study, we were able to employ encapsulins for the size-constrained synthesis of antimicrobial silver nanoparticles (Figure 8).

Figure 8: Using engineered encapsulins for the biological synthesis of inorganic nanomaterials (here: silver nanoparticles).



Giessen, TW, Silver, PA. (2017). Engineering carbon fixation with artificial protein organelles. Curr. Opin. Biotechnol. (accepted).

Giessen, TW, Silver, PA. (2017). Widespread distribution of encapsulin nanocompartments reveals functional diversity. Nat. Microbiol., 2017 (under review).

Liu, X, Lopez, PA, Giessen, TW, Giles, M, Way, JC, Silver, PA. (2016). Engineering genetically-encoded mineralization and magnetism via directed evolution. Sci. Rep. 6:38019.

Giessen, TW, Silver, PA. (2016). A catalytic nanoreactor based on in vivo encapsulation of multiple enzymes in an engineered protein nanocompartment. ChemBioChem 17(20), 1931-1935.

Giessen, TW, Silver, PA. (2016). Converting a natural protein compartment into a nanofactor for the size-constrained synthesis of antimicrobial silver nanoparticles. ACS Synth. Biol. 5(12), 1497-1504.

Giessen, TW. (2016). Encapsulins: microbial nanocompartments with applications in biomedicine, nanobiotechnology and materials science. Curr. Opin. Chem. Biol. 34, 1-10.

Giessen, TW, Silver, PA. (2015). Encapsulation as a strategy for the design of biological compartmentalization. J. Mol. Biol. 428(5 Pt B), 916-927.