286 14 Sequestered: Design and Construction of Synthetic Organelles
signal sequence can interact with the hexameric shell proteins (PduA, PduJ, and
PduK), and inclusion of this sequence allows encapsulation of foreign cargo
[47, 48]. In addition, subsequent studies also reveal that the short N‐terminal
extension of the medium subunit (PduD) of adenosylcobalamin diol dehydratase
(PduCDE) and an unknown protein PduV can target their respective cargo to
PDU [49, 50]. Besides PDU, other catabolic BMCs, including EUT and a glycyl
radical‐based propanediol utilization (GRP) microcompartment, use signal pep
tides to encapsulate their respective cargo [51–53]. Strikingly, these targeting
peptides have been shown to enable targeting of green fluorescent protein (GFP)
to PDU, suggesting that the targeting specificity for these BMCs is not stringent
and may be determined by the composition rather than the sequence of the tar
geting peptides [53]. This relaxed specificity allowed for the de novo construction
of synthetic signal peptide for PDU targeting. With the growing repertoire of
natural and synthetic signaling peptides, it may soon be possible to encapsulate
multiple enzymes in a BMC to constitute a longer metabolic pathway. While the
mechanism of encapsulation via signaling peptide is not completely understood,
interesting applications have already emerged, including the construction of an
ethanol nanoreactor by encapsulating pyruvate decarboxylase and alcohol dehy
drogenase in PDU [47] and compartmentalization of polyphosphate kinase
(PPK1) in PDU to enhance the conversion of biological phosphates to cellular
polyphosphate [54].
In contrast to catabolic BMCs, the targeting strategy used by carboxysomes is
less well characterized, and most of the understanding came from β‐CBs. Pull‐
down and yeast two hybrid experiments probing components of the cyanobacte
rial β‐CB revealed that RuBisCO is anchored to the shell via specific interactions
with the protein CcmM, which acts as an intermediate bridge between enzy
matic cargo and the shell. CcmM also possesses a nonfunctional carbonic anhy
drase‐like domain and recruits the functional carbonic anhydrase, CcaA, forming
a functional carbon‐fixing complex [37, 38]. More recently, a CB protein CcmN
was found to be essential for the shell recruitment during carboxysome assem
bly, and its deletion resulted in a large shell‐less RuBisCO aggregate [55]. A C‐
terminal extension of this protein appears to interact with the major shell
hexamer CcmK2 and is sufficient for targeting the GFP into CBs [56]. Therefore,
CcmN may be the actual mediator between the shell and the CcmM/CcaA/
RuBisCO complex discussed earlier. In α‐CBs, homologs of CcmM and CcmN
are not present, but a poorly characterized protein called CsoS2 may perform an
analogous function. CsoS2 is an intrinsically disordered protein with many
amino acid repeats [57, 58]. These properties are often associated with proteins
that function as “assembly coordinators” for large complexes, thus providing an
important clue about the function of CsoS2 [59]. As evidence of the necessity of
CsoS2 to α‐CB assembly, deletion of CsoS2 in Halothiobacillus neapolitanus
abolishes carboxysome formation and renders the organism high‐CO 2 requiring
(HCR) [57]. Interestingly, it was shown that one csoS2 coding sequence produces
two CsoS2 isoforms via a co‐translational mechanism [58], reminiscent of CcmM
and CcmN in β‐CBs. If the CsoS2 isoforms are indeed functionally analogous to
CcmM and CcmN, understanding the way they interact with other carboxyso
mal proteins may shed light on how cargo in α‐CB is encapsulated. Additional