14.3 roteinnBased Organelles 287work is required to understand the true role of CsoS2 and to develop a strategy
to target foreign cargo to α‐CBs.
Despite these advances, several outstanding questions in carboxysome target
ing remain. Firstly, the stoichiometry of cargo proteins can vary over 10‐fold, that
is, there are roughly 100 protomers of carbonic anhydrase and 2000 protomers of
RuBisCO, yet there is no mechanistic explanation for how this can be pro
grammed through protein–protein interactions alone. This will be critical to
understand as future engineers attempt to balance flux through multistep enzy
matic pathways. Secondly, little is known about protein targeting in the α‐CB.
The α‐CB from H. neapolitanus is a structurally robust BMC that can assemble
without cargo and be transgenically expressed in Escherichia coli, making it an
intriguing chassis for synthetic biological purposes [60, 61]. Making this a reality,
however, will ultimately require a complete biophysical understanding of the tar
geting motifs and mechanisms.
14.3.1.2 Permeability
The structure of the shell proteins is thought to control permeability of the BMC
(Figure 14.3c). X‐ray crystal structures of various hexameric and pentameric
shell proteins have revealed pores along the major axis of symmetry that, in prin
ciple, would facilitate passive diffusion of substrates and products. The pores are
generally small (4–6 Å in diameter), implying specificity [62]. In the case of the
CB, positively charged residues are found at the narrowest area of the pore, sug
gesting a mechanism for screening for negatively charged molecules, such as the
substrates/products bicarbonate, ribulose 1,5‐bisphosphate, and 3‐phosphoglyc
erate, and against molecules without a dipole such as O 2. Although there are few
permeability measurements to support these hypotheses, physiological data
clearly implies that there is minimal photorespiration (fixation of O 2 ) when
RuBisCO is inside the CB, suggesting O 2 exclusion may be one effect of encapsu
lating RuBisCO [63, 64]. In addition, csoS4‐disrupted H. neapolitanus have a
HCR phenotype, and their CBs leak CO 2 as interpreted by kinetic experiments
[65]. Thus, a tight BMC shell appears to act as a gas barrier to exclude O 2 and
sequester CO 2. This theme is seen in other BMCs, as well. For instance,
Salmonella enterica mutants that cannot produce PDU accumulate 10‐fold
increased levels of propanal in the cytosol [11]. Aldehydes, as nonspecific cross‐
linkers, damage DNA, and the 10× increase in propanal levels proved to be highly
mutagenic. Similarly, alteration of pore‐lining residues in PduA resulted in pro
panal leakage, reduced 1,2‐propanediol influx, and increased glycerol influx, fur
ther substantiating the role of PDU shell as a selective diffusion barrier [66]. In
the case of the EUT, shell mutants also leak their intermediate, acetaldehyde, but
here, physiological data supports the hypothesis that the sequestration acts to
stop the loss of a volatile intermediate out of the pathway. Thus, BMC shells can
achieve many catalytic goals – enhancing pathway specificity and yield while
reducing toxicity – by tuning their permeability.
Recent X‐ray structures highlight an expanded toolkit for altering shell perme
ability. The relatively small size of most pores (4–6 Å) is at odds with the required
permeability for larger substrates such as ribulose 1,5‐bisphosphate or the cofac
tors coenzyme A and NAD used in the PDU and EUT. Kerfeld and colleagues