288 14 Sequestered: Design and Construction of Synthetic Organelles
have solved the structure of the CsoS1D protein from the α‐CB, which possesses
a tandem BMC domain and forms a homotrimeric pseudohexamer with a much
larger pore of 14 Å [67]. Although this protein is of low abundance and only
recently detected in purified CBs, it may play an important role in allowing larger
molecules passage into and out of the CB [61, 68]. Even more intriguingly,
CsoS1D also crystallized in alternate conformations with both an open and
closed pore. In follow‐up work, Kerfeld et al. solved the structure of the ortholo
gous tandem repeat protein from the β‐CB and again observed open and closed
forms [69]. More recently, EutL has been shown to have negative allosteric regu
lation for pore opening by ethanolamine, and disulfide bonding may play a role
in modulating the binding affinity toward ethanolamine [70]. These findings
raise the possibility of posttranslational regulation of BMC permeability.
Catalyzing redox reactions is a critical component of PDU and EUT activity
and recent results also highlight the role of the shell in these processes.
Overexpression of the Citrobacter freundii PDU and its components in E. coli
led to the surprising realization that the shell protein PduT contains an Fe–S
cluster on its major symmetry axis [71, 72]. This was confirmed via electron
paramagnetic resonance and X‐ray crystallography. The midpoint potential was
measured at +0.099 V, suggesting the cluster may help recycle NADH, produced
during the oxidation of propionaldehyde to propionyl‐CoA, back to NAD+.
Similarly, the shell protein GrpU of GRP microcompartment also coordinates
Fe–S cluster [73]. While further experimental validation is required to demon
strate that these shells can truly participate in a redox reaction, it does under
score the potential for catalytic flexibility among BMCs.14.3.1.3 Chemical Environment
A related property is chemical environment, including redox state, pH, interme
diate concentrations, and cofactor status. This results from the interplay of shell
permeability and enzymatic activity in the lumen, creating steady‐state concen
trations of molecular species different than what exists in the cytosol. For exam
ple, a recent mathematical model predicts that a relatively acidic carboxysome
will exhibit higher equilibrium CO 2 concentration and, in turn, a higher degree
of RuBisCO saturation [23]. This finding raises the possibility that the actual
carboxysome may similarly be acidic in order to achieve maximum catalytic
efficiency. Preliminary biochemical analyses of carbonic anhydrases from some
β‐CBs also suggest that the lumen environment may be oxidative, promote
disulfide bond formation, and be a means of controlling protein activity
[74, 75, 138]. Interestingly, CsoS2, the putative scaffolding protein of the α‐CB,
contains many cysteine residues, half of which are conserved across amino acid
repeats. The abundance of cysteines may imply CsoS2’s participation in disulfide‐
bonding network within the carboxysomal lumen, which, if true, would explain
the exceptional robustness of α‐CB.
The chemical environment is likely more extreme in the PDU and EUT. As
described earlier, one explanation for PDU/EUT function is to sequester the
buildup of toxic aldehyde intermediates away from the cytoplasm. S. enterica
mutants with disrupted PDU shells accumulate 10× higher levels of cytosolic
propionaldehyde (~15 mM), suggesting that luminal aldehyde concentrations are