**5.3. PII regulation of arginine metabolism**

The PII signal transduction proteins are widely distributed in prokaryotes and chloroplasts, where they play a coordinating role in the regulation of nitrogen assimilatory processes [71– 73]. For this purpose, PII senses the energy status of the cell by binding ATP or ADP in a competitive way [74]. Binding of ATP and synergistic binding of 2-oxoglutarate (2-OG) allows PII to sense the current carbon/nitrogen status of the cell [75]. 2-OG is the carbon skeleton for the GS/GOGAT reactions and thereby links the carbon and nitrogen metabolism in all domains of life [76, 77]. The pool size of 2-OG reacts quickly to changes in nitrogen availability, wherefore 2-OG is an indicator of the carbon/nitrogen balance [78, 79]. Depending on the nitrogen supply, PII may be phosphorylated at the apex of the T-loop at position Ser49 [80, 81]. Binding of the effector molecules ATP, ADP and 2-OG as well as phosphorylation leads to conformational rearrangements of the large surface-exposed T-loop, PII's major protein-interaction structure [82]. These conformational states direct the interaction of PII with its various interaction partners and thereby regulate the cellular C/N balance [83].

In cyanobacteria, PII regulates the global nitrogen control transcriptional factor NtcA, through binding to the NtcA co-activator PipX [84]. In common with other bacteria, cyanobacterial PII proteins can interact with the biotin carboxyl carrier protein (BCCP) of acetyl-CoA carboxylase (ACC) and thereby control the acetyl-CoA levels [85]. Furthermore, PII controls arginine biosynthesis via regulation of NAGK [68, 69, 86].

PII proteins form a cylindrical-shaped homotrimer with 12–13 kDa per subunits. The T-loop, a large and surface-exposed loop, protrudes from each subunit. The effector binding sites are positioned in the three inter-subunit clefts [87, 88]. If sufficient energy and nitrogen are available, indicated by a high ATP and low 2-OG level, non-phosphorylated PII forms an activating complex with NAGK.

The crystal structure of the PII-NAGK complex from *Synechococcus elongatus* strain PCC 7942 revealed two PII trimers sandwiching a NAGK homohexamer (trimer of dimers) [88]. Each PII subunit contacts one NAGK subunit [88]. Two parts of PII are involved in interaction with NAGK.The first structure, called B-loop, is located on the PII body and interacts with the C-domain of NAGK subunit, involving residue Glu85. The interaction of the B-loop is the first step in complex formation. Second, the T-loop must adopt a bent conformation and insert into the interdomain cleft of NAGK [89]. This enhances the catalytic efficiency of NAGK, with the Vmax increasing fourfold and the Km for N-acetylglutamate decreasing by a factor of 10 [86]. Furthermore, feedback inhibition of NAGK by arginine is strongly decreased in the presence of PII [86].

During PII mutagenesis, a PII variant was identified that binds constitutively NAGK in vitro. This PII variant exhibits a single amino acid replacement, Ile86 to Asn86, hereafter referred as PII(I86N) [89]. The crystal structure of PII(I86N) has been solved, showing an almost identical backbone than wild-type PII. However, the T-loop adopts a compact conformation, which is a structural mimic of PII in the NAGK complex [89, 90]. Addition of 2-OG in the presence of ATP normally leads to a dissociation of the PII-NAGK complex, however PII(I86N) no longer responds to 2-OG [90].

The PII(I86N) variant enables a novel approach of metabolic pathway engineering by using custom-tailored PII signaling proteins. By replacing the wild-type PII with a PII carrying the mutation for I86N in *Synechocystis* sp. PCC 6803, it was possible to engineer the first cyanobacterial CGP overproducer strain. Strain BW86, containing the PII(I86N) version, shows an increase of NAGK activity, which causes a more than 10-fold higher arginine content than the wild-type [10]. Under balanced growth conditions with nitrate as nitrogen source, strain BW86 accumulates up to 15.6 ± 5.4% CGP relative to the CDM, i.e., on average almost sixfold more than the wild type. Appropriate starvation conditions can further increase the CGP content of strain BW86 up to 47.4 ± 2.3% per CDM under phosphate starvation and 57.3 ± 11.1% per CDM under potassium starvation, without addition of arginine to the medium [10]. Furthermore, the CGP, which is produced by strain BW86, shows a high polydispersity ranging from 25 to 100 kDa, similar to the polydispersity of cyanobacterial wild-type CGP, which contrasts CGP from recombinant producer strains using heterologous expression systems with heterotrophic bacteria, yeasts or plants [10]. CGP isolated from those strains have a size ranging of 25–45 kDa [27, 31, 32].
