292 14 Sequestered: Design and Construction of Synthetic Organelles
threefold axis pores – they are positively charged in the Thermotoga encapsulin
while uncharged in many other classes. The explanation for this divergence is
yet unknown, although it may reflect the nature of the small molecules that must
traverse the shell.
This study also led to the identification of a putative signal sequence for encap
sulins. Bioinformatic analysis has revealed that two classes of enzymes, peroxi
dases and ferritin‐like proteins, preferentially cluster in minimal operons
adjacent to the shell‐forming encapsulin gene. Serendipitously, there was addi
tional electron density in the Thermotoga structure abutting the inner face of the
encapsulin shell. This density was of sufficient signal to identify a primary pep
tide sequence, which matched the C‐terminus of the adjacent ferritin‐like gene
in the operon, establishing the link between the gene cluster and protein struc
ture. Deletion of the C‐terminal region also disrupted targeting of the enzyme to
the lumen, confirming this sequence is essential for targeting [83]. By employing
this targeting sequence, many studies have reported successful targeting of
heterologous cargo into encapsulins [90–92].
Encapsulins therefore have many advantages as potential synthetic organelles.
They are in many ways a minimal version of BMCs. They assemble from a single
shell protein into a compartment possessing about 1/100 the volume. This
genetic simplicity likely ensures porting structures between organisms will be
easier than for BMCs (Figure 14.5a). Preliminary experiments agree with this
hypothesis – encapsulins from many organisms including B. linens, T. maritima,
M. xanthus, and M. tuberculosis can be expressed heterologously in E. coli [87].
As an additional advantage, encapsulins commonly display exceptional resist
ance to temperature, pH, denaturant, proteases, and mechanical compression
[83, 90, 91, 93, 94]. Therefore, they may serve as appealing alternatives for appli
cations that demand extreme conditions incompatible with other biological
compartments. However, it appears the “addressability” – the number of proteins
that can be targeted to its lumen – will be limited to one or two. For example,
Snijder and colleagues employed native mass spectroscopy to show that one
B. linens microcompartment precisely packages one hexamer of peroxidases,
suggesting that the limited capacity is likely a valid concern [94]. In this vein, we
imagine that very short pathways, that is, two steps with a single toxic intermedi
ate, would make excellent candidates for encapsulation. Future work will also be
required to understand and engineer shell permeability.14.4 Lipid-Based Organelles
The alternative to protein‐based complexes is to leverage the natural organiza
tion of metabolism found in eukaryotes – membranous organelles. This makes
practical sense as many key pathways of catabolism and anabolism are segregated
at the organelle level, as discussed in the following text. In addition, much of our
biological understanding of these processes comes from the yeast Saccharomyces
cerevisiae, which is arguably the most important organism for metabolic engi
neering (Figure 14.6). Thus, there is already a working understanding of organelle
targeting, permeability, chemical environment, and biogenesis.