Using comparative genomics to uncover new kinds of protein-based metabolic organelles in bacteria

Abstract

Bacterial microcompartment (MCP) organelles are cytosolic, polyhedral structures consisting of a thin protein shell and a series of encapsulated, sequentially acting enzymes. To date, different microcompartments carrying out three distinct types of metabolic processes have been characterized experimentally in various bacteria. In the present work, we use comparative genomics to explore the existence of yet uncharacterized microcompartments encapsulating a broader set of metabolic pathways. A clustering approach was used to group together enzymes that show a strong tendency to be encoded in chromosomal proximity to each other while also being near genes for microcompartment shell proteins. The results uncover new types of putative microcompartments, including one that appears to encapsulate B(12) -independent, glycyl radical-based degradation of 1,2-propanediol, and another potentially involved in amino alcohol metabolism in mycobacteria. Preliminary experiments show that an unusual shell protein encoded within the glycyl radical-based microcompartment binds an iron-sulfur cluster, hinting at complex mechanisms in this uncharacterized system. In addition, an examination of the computed microcompartment clusters suggests the existence of specific functional variations within certain types of MCPs, including the alpha carboxysome and the glycyl radical-based microcompartment. The findings lead to a deeper understanding of bacterial microcompartments and the pathways they sequester.

Figure 1

Morphology of MCPs and models of their sequestered pathways. (A) Thin-section EM of a dividing cell of the cyanobacterium Synechocystis sp. PCC6803 (left) along with an enlargement of a single carboxysome (right, courtesy of Wim Vermaas). (B) Electron micrographs of a thin-section of Salmonella enterica serovar Tyhpimurium LT2 (left; Reproduced from Ref., with permission from Thomas Bobik) and purified Pdu microcompartments (right; Reproduced from Ref., with permission from Thomas Bobik). (C) Models for CO₂ fixation and 1,2-propanediol and ethanolamine metabolism in the carboxysome, Pdu and Eut microcompartments, respectively.

Figure 2

Schematic for identifying pairs of proteins or enzyme families that tend to co-occur in the context of microcompartment (MCP) operons. As step 1, BMC-proximal genes and their encoded proteins are collected following our operational definition of a MCP operon (see Methods). In step 2, to gain clarity and statistical power, the BMC-proximal genes are assigned to Protein Functional Groups, grouping together similar protein sequences where possible. In step 3, the co-occurrence tendency is evaluated by a pairwise correlation coefficient (PCC) for every pair of Functional Groups. In the last step, after applying statistical confidence tests, strongly linked Protein Functional Groups are clustered together. Different clusters identify MCPs with distinct metabolic functions. The scheme shown is a simplification; application to real genomic data leads to more MCP types and more proteins per cluster.

Figure 3

Clusters of proteins and enzymes predicted by a computational approach to constitute distinct kinds of microcompartments. Clusters have been numbered from 1 to 10. Each Protein Functional Group is represented as a node (plain white for strong nodes, light gray for weak nodes) and a significant correlation between two nodes is represented as an edge. Protein Functional Groups are labeled with their corresponding gene name when consistent annotations are available across the different species. Clusters 1, 2, 3, and 5 are related to the canonical microcompartments (beta and alpha carboxysomes, and Pdu and Eut microcompartments, respectively). Cluster 4 is related to the cob operon, which upon closer inspection is seen to relate to the Pdu MCP. Clusters 6–9 relate to a presumptive MCP for glycyl radical-based propanediol utilization (which we name Grp), along with variations under which it appears in different species. Cluster 10 identifies a potential MCP in mycobacteria that could involve amino alcohol metabolism.

Figure 4

Phylogenetic profile of the BMC-proximal proteins and enzymes from the presumptive glycyl radical-based propanediol utilization (Grp) microcompartment operon across 23 bacterial species. (A) The 23 microorganisms are extracted from the analysis of cluster 6 (Fig. 3). Each colored block refers to the presence of a given protein (on the left) in the operon featured by a given organism (on the top). Cluster 6, corresponding to the core enzymes of the pathway, is depicted in light blue while supplemental enzymes from clusters 7, 8, and 9 are shown in red, purple and green respectively. A black block next to a protein name indicates the presence of a presumptive N-terminal targeting extension in this protein compared to homologues not involved in microcompartments, as analyzed following previously described methods. The last row represents the profile of a divergent BMC shell protein, apparently specific to the Grp MCP, which was subjected to initial experimental characterization (see text). Next to each organism name, the number of tandem BMC proteins present in the operon is enclosed in red brackets. (B) Examples of gene organization in species featuring the Grp operon. BMC shell genes are colored in light gray, while the genes clustered by our approach are colored consistently with their corresponding clusters depicted in panel A.

Figure 5

Proposed model for metabolism in the Grp microcompartment. Similar to the Pdu microcompartment, the expected final products include propanol and propionyl-phosphate.

Figure 6

An updated classification of MCPs including presumptive types identified computationally. Microcompartments are divided here into seven main classes according to the core enzymes and pathways confined in their lumen; the two carboxysome subtypes are separated here on the basis of their partially distinct compositions. For some of the MCPs, their core functions appear to be augmented by the presence of additional groups of proteins or enzymes; some of these may be directly involved with MCP function while others could be more peripheral (e.g., regulatory). These extensions suggest the existence of more complex or diverse MCP variants.