Studies of protein complexes involved in the adenylation cascade of the nitrogen signalling pathway in Escherichia coli.

Adenylyl transferase (ATase), the glnE gene product from Escherichia coli, is a bifunctional enzyme that catalyses the opposing adenylylation and deadenylylation of glutamine synthetase (GS). The overall aim of this thesis was elucidation of the molecular mechanisms of the adenylylation cascade. A new central domain has been identified using ATase truncation constructs in activity assays and solubility trials. This new regulatory domain is flanked by two flexible Qlinkers, Q1 and Q2. Thus the topology of ATase can be represented as N-Q1-R-Q2-C. The N domain was PII-UMP independent in in vitro deadenylylation assays, and had 1000 fold less activity then entire ATase, suggesting PII-UMP binding impacts on the conformation of the deadenylylation active site. Monoclonal antibodies (mAbs) generated in this work against ATase were characterised using ATase mutants and the truncated proteins. Two mAbs, 5A7 (binds residues 502- 548) and 39G11 (binds residues 466-501) both binding in the R domain, blocked the binding of PII, GlnK, PII-UMP and GlnK-UMP to ATase. This is the first report that pinpoints the effector-protein binding sites to within the R domain of ATase. Both PII and ATase bound α-ketoglutarate (α-kg) in direct binding assays. Several lines of evidence suggested that PII contains the high affinity α-kg binding site and ATase the low affinity site. This study demonstrates for the first time that the two small effector-molecules α-kg and glutamine (gln) probably bind in the last 340 residues of ATase (Q2-C domain), possibly near the adenylylation active site. The demonstration that the ATase mutant W694G presented a gln independent phenotype suggests that the bulky side chain of Trp 694 must move out of the adenylylation active site, so that GS can dock and be modified. Surface plasmon resonance (SPR) data suggested the binding of gln within Q2-C is transmitted to the R domain as an allosteric inhibitor of PII-UMP binding, and consequently deadenylylation. A panel of mAbs was also produced against PII and characterised using a series of PII mutants. Two of the PII mAbs 19G4 (binds PII/GlnK) and 24H2 (binds PII) were used further to demonstrate that heterotrimers are formed between PII and GlnK in vivo in nitrogen starved cells. It is well documented that the T-loops of PII and GlnK are probably the principal regions used by these signalling proteins to bind to the various receptor-proteins such as ATase, UTase and NRII. This study suggests PII and GlnK also interact with GS. Using PII mutants carrying specific GlnK residues at positions 43, 52 and 54 in the T-loop this study demonstrated the Asp at position 54 was the critical determinant of PII T-loop binding to GS, whereas the interaction with UTase involved both Asp54 and Thr43.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

James Cook University

Year dc:date.issued

2004

Author dc:creator

• Clancy, Paula

Language dc:language

en