Signal transduction to the Azotobacter vinelandii NIFL-NIFA regulatory system is influenced directly by interaction with 2-oxoglutarate and the PII regulatory protein
2000; Springer Nature; Volume: 19; Issue: 22 Linguagem: Inglês
10.1093/emboj/19.22.6041
ISSN1460-2075
Autores Tópico(s)Plant-Microbe Interactions and Immunity
ResumoArticle15 November 2000free access Signal transduction to the Azotobacter vinelandii NIFL–NIFA regulatory system is influenced directly by interaction with 2-oxoglutarate and the PII regulatory protein Richard Little Richard Little Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Francisca Reyes-Ramirez Francisca Reyes-Ramirez Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Yan Zhang Yan Zhang Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Wally C. van Heeswijk Wally C. van Heeswijk Department of Molecular Cell Physiology, Faculty of Biology, Free University, 1081 HV Amsterdam, The Netherlands Present address: Department of Plant and Microbial Biology, University of California at Berkeley, 111 Koshland Hall, Berkeley, CA, 94720-3102 USA Search for more papers by this author Ray Dixon Corresponding Author Ray Dixon Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Richard Little Richard Little Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Francisca Reyes-Ramirez Francisca Reyes-Ramirez Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Yan Zhang Yan Zhang Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Wally C. van Heeswijk Wally C. van Heeswijk Department of Molecular Cell Physiology, Faculty of Biology, Free University, 1081 HV Amsterdam, The Netherlands Present address: Department of Plant and Microbial Biology, University of California at Berkeley, 111 Koshland Hall, Berkeley, CA, 94720-3102 USA Search for more papers by this author Ray Dixon Corresponding Author Ray Dixon Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK Search for more papers by this author Author Information Richard Little1, Francisca Reyes-Ramirez1, Yan Zhang1, Wally C. van Heeswijk2,3 and Ray Dixon 1 1Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich, NR4 7UH UK 2Department of Molecular Cell Physiology, Faculty of Biology, Free University, 1081 HV Amsterdam, The Netherlands 3Present address: Department of Plant and Microbial Biology, University of California at Berkeley, 111 Koshland Hall, Berkeley, CA, 94720-3102 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (2000)19:6041-6050https://doi.org/10.1093/emboj/19.22.6041 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info PII-like signal transduction proteins, which respond to the nitrogen status via covalent modification and signal the carbon status through the binding of 2-oxoglutarate, have been implicated in the regulation of nitrogen fixation in several diazotrophs. The NIFL–NIFA two-component regulatory system, which integrates metabolic signals to fine-tune regulation of nitrogenase synthesis in Azotobacter vinelandii, is a potential target for PII-mediated signal transduction. Here we demonstrate that the inhibitory activity of the A.vinelandii NIFL protein is stimulated by interaction with the non-uridylylated form of PII-like regulatory proteins. We also observe that the NIFL–NIFA system is directly responsive to 2-oxoglutarate. We propose that the PII protein signals the nitrogen status by interaction with the NIFL–NIFA system under conditions of nitrogen excess, and that the inhibitory activity of NIFL is relieved by elevated levels of 2-oxoglutarate when PII is uridylylated under conditions of nitrogen limitation. Our observations suggest a model for signal transduction to the NIFL–NIFA system in response to carbon and nitrogen status which is clearly distinct from that suggested from studies on other diazotrophs. Introduction Transcriptional regulation of nitrogen fixation catalysed by molybdenum nitrogenase in Klebsiella pneumoniae and Azotobacter vinelandii is maintained by a regulatory protein complex comprising the σN-dependent transcriptional activator NIFA and the sensor protein NIFL (Dixon, 1998). Unlike conventional two-component systems, which communicate by a conserved phosphotransfer mechanism, NIFL inhibits the activity of NIFA in response to oxygen and fixed nitrogen through the formation of a stoichiometric protein–protein complex (Henderson et al., 1989; Govantes et al., 1996; Money et al., 1999). Whereas the redox sensing function of NIFL is relatively well understood (Hill et al., 1996; Schmitz, 1997; Macheroux et al., 1998), the mechanism whereby the NIFL–NIFA system responds to the nitrogen status to control nitrogen fixation is not well defined. The signal transduction protein PII, which plays a central role in global nitrogen regulation, is widely distributed in Bacteria, Archaea and plants (Ninfa and Atkinson, 2000). The mechanism of signal transduction is best understood in enteric bacteria and involves covalent modification of the PII protein, encoded by glnB, by a uridylyltranferase/uridylyl removing enzyme (UTase/UR) encoded by glnD (Merrick and Edwards, 1995). The UTase/UR transduces the nitrogen signal through uridylylation of PII under conditions of nitrogen limitation and via de-uridylylation of PII under conditions of nitrogen excess. Glutamine is the primary signal for the fixed nitrogen status, and modulates the uridylylation state of PII by acting as an effector of the UTase/UR (Jiang et al., 1998a). The PII protein interacts with three known receptors: UTase/UR; adenylyltransferase (ATase), which controls the activity of glutamine synthetase; and the sensor protein NtrB (NRII), which regulates the activity of the nitrogen regulatory protein NtrC (Jaggi et al., 1996, 1997; Jiang et al., 1997a, b, 1998a, b, c). These interactions are not influenced only by the uridylylation state of PII, but are also allosterically modulated through binding of the effector 2-oxoglutarate to the PII protein (Kamberov et al., 1995; Jiang et al., 1998c; Jiang and Ninfa, 1999). PII is thus able to coordinate the nitrogen signal, received by covalent modification, with the carbon status signalled by the binding of 2-oxoglutarate (Ninfa and Atkinson, 2000). Previous studies with K.pneumoniae suggested that neither glnB nor glnD was required for nitrogen sensing by NIFL, and it was postulated that an alternative nitrogen sensing pathway could be involved (Holtel and Merrick, 1989; Edwards and Merrick, 1995). Subsequently, an alternative PII protein encoded by the glnK gene was identified in Escherichia coli (van Heeswijk et al., 1995, 1996) and it is now apparent that two or more PII-like proteins are present in many bacteria (Ninfa and Atkinson, 2000). PII-like proteins are highly conserved in their amino acid sequence and have a very similar crystal structure as determined for PII and GlnK of E.coli (Cheah et al., 1994; Carr et al., 1996; Xu et al., 1998). NIFL inhibits NIFA activity irrespective of the nitrogen status in glnK mutants, implying that GlnK is required either directly or indirectly to relieve inhibition by K.pneumoniae NIFL under nitrogen-limiting conditions (He et al., 1998; Jack et al., 1999). Relief of NIFL inhibition is a function that is relatively specific to GlnK, although overexpression of glnB leads to some relief of inhibition (Arcondéguy et al., 1999). Genetic experiments suggest that uridylylation of GlnK is not essential for relief of inhibition by K.pneumoniae NifL (Edwards and Merrick, 1995; He et al., 1998) and it is not clear how the nitrogen signal is communicated via GlnK. Since expression of glnK is regulated by NtrC (van Heeswijk et al., 1996; He et al., 1997), there is probably an insufficient level of GlnK under conditions of nitrogen excess to relieve inhibition by NIFL, but this does not explain how the system responds rapidly to changes in nitrogen status (He et al., 1998; Arcondéguy et al., 1999). The aerobic diazotroph A.vinelandii contains homologues of enteric nitrogen regulatory genes (Toukdarian and Kennedy, 1986; Toukdarian et al., 1990; Contreras et al., 1991; Meletzus et al., 1998), but the interface between the global nitrogen regulatory system and regulation of nitrogen fixation may be different to that in enteric bacteria. First, mutations in glnD block the synthesis of nitrogenase in A.vinelandii. This phenotype can be suppressed by secondary insertion mutations that inactivate nifL, implying that, unlike in K.pneumoniae, uridylylation of a regulatory component may be required to prevent inhibition by NIFL (Contreras et al., 1991). Secondly, A.vinelandii apparently contains only a single PII-like protein encoded by a gene designated as glnK, which is located in an operon with amtB, which encodes a methyl ammonium transporter (Meletzus et al., 1998). Unlike in enteric bacteria, the glnK amtB operon is not apparently regulated in response to nitrogen status. Both A.vinelandii glnK and glnD appear to be essential for growth since stable null mutations have not been obtained in either gene (Rudnick et al., 1998; P.Rudnick and C.Kennedy, unpublished). Nevertheless, the A.vinelandii NIFL–NIFA regulatory system is responsive to nitrogen regulation in vivo when introduced into E.coli, indicating that the A.vinelandii proteins are responsive to enteric nitrogen regulatory components (Söderbäck et al., 1998). We have investigated the response of the A.vinelandii NIFL–NIFA proteins to signal transduction by PII-like regulatory proteins using a purified in vitro system. Surprisingly, we observed that the NIFL–NIFA complex is itself responsive to the presence of 2-oxoglutarate, which relieves inhibition by NIFL in the presence of adenosine nucleotides. Hence, the nitrogen fixation regulatory proteins are themselves exquisitely sensitive to the carbon status. Using the well characterized PII-like proteins from E.coli, we demonstrate that the non-uridylylated form of the PII protein (Ec PII), but not the GlnK protein (Ec GlnK), is competent to activate the inhibitory function of NIFL in the presence of 2-oxoglutarate and adenosine nucleotides. Furthermore, we show that the glnK-encoded PII-like regulatory protein from A.vinelandii (Av PII) also activates the inhibitory function of NIFL and this interaction is modulated by the uridylylation state of Av PII. These observations suggest a mechanism for signalling the nitrogen status in A.vinelandii through the interaction between PII-like proteins and the NIFL–NIFA system which is clearly different from that proposed for K.pneumoniae. Results Influence of adenosine nucleotides on NIFL activity in the absence of the redox response We have shown previously that the inhibitory activity of A.vinelandii NIFL on NIFA is stimulated by the presence of adenosine nucleotides (Eydmann et al., 1995) and that ADP increases the stability of the NIFL–NIFA protein complex (Money et al., 1999). Since ATP and 2-oxoglutarate are effectors of the Ec PII protein (Kamberov et al., 1995; Jiang et al., 1998c; Jiang and Ninfa, 1999) and its paralogue Ec GlnK (Atkinson and Ninfa, 1999), we first performed control experiments to determine whether these ligands influenced the activity of the NIFL–NIFA complex in the absence of PII-like regulatory proteins. NIFL also inhibits NIFA activity under oxidizing conditions as a consequence of oxidation of the FAD co-factor (Hill et al., 1996; Schmitz, 1997; Macheroux et al., 1998) located in the N-terminal PAS domain (Zhulin et al., 1997). In order to simplify the assays, we made use of a truncated version of NIFL, NIFL(147–519), which lacks the PAS domain and is not responsive to oxygen/redox control but remains responsive to adenosine nucleotides in vitro and to the nitrogen status in vivo (Söderbäck et al., 1998). NIFL(147–519) is competent to form stable complexes with NIFA in the presence of MgADP (Money et al., 1999). The ability of NIFL(147–519) to inhibit the positive control function of NIFA was determined by measuring the transcriptional activation by NIFA at the nifH promoter. These assays quantitate the formation of open promoter complexes by NIFA in the presence of σN RNA polymerase holoenzyme and integration host factor (Eydmann et al., 1995). Since nucleotide triphosphate hydrolysis is necessary for catalysis of open promoter complexes by NIFA, it is necessary to provide either ATP or GTP for the energy transduction step in open complex formation. We have shown previously that NIFL inhibits NIFA activity in the presence of ATP. At relatively low concentrations of adenosine nucleotides, this inhibition is probably due to the release of ADP during ATP hydrolysis by NIFA, since inhibition can be relieved by the addition of an ATP regenerating system (Eydmann et al., 1995). We first examined the effect of nucleotides on the truncated NIFL(147–519) protein (Table I). When GTP was used to promote open complex formation by NIFA, in the presence of a relatively low concentration of ADP (50 μM), NIFL(147–519) inhibited NIFA activity and this inhibition was relieved by the inclusion of an ATP generating system in the reactions (Table I, rows 4 and 5), as observed previously with the native full-length form of NIFL (Eydmann et al., 1995). However, when ATP was present at high concentration (3.5 mM), the regenerating system failed to relieve inhibition by NIFL(147–519), suggesting that ATP might also promote the formation of the inhibitory NIFL–NIFA complex at saturating concentrations (Table I, rows 6–9). This possibility was also suggested by the observation that non-hydrolysable analogues of ATP, ATPγS and AMPPNP also promoted inhibition by NIFL (Table I, rows 10–13), consistent with the observation that ATPγS promotes formation of the NIFL–NIFA complex (Money et al., 1999). Differential inhibition in response to ADP and ATP could reflect the higher affinity of NIFL for ADP (Kd = 13 μM) compared with ATP (Kd = 130 μM) (Söderbäck et al., 1998). Table 1. Influence of adenosine nucleotides on inhibition of NIFA activity by NIFL(147–519) in the presence of an ATP regenerating system No. Componentsa Radioactivity in open complex (%) Activity (%) 1 NIFA + GTP 41.3 100 2 NIFA + GTP + ADP 29.7 72 3 NIFA + GTP + ADP + CK/CP 40.6 98 4 NIFA + NIFL(147–519) + GTP + ADP 5.6 14 5 NIFA + NIFL(147–519) + GTP + ADP + CK/CP 41.6 100 6 NIFA +ATP 30.5 74 7 NIFA + ATP + CK/CP 36.5 88 8 NIFA + NIFL(147–519) + ATP 3.3 8 9 NIFA + NIFL(147–519) + ATP + CK/CP 2.6 6 10 NIFA + GTP + ATPγS 39.9 97 11 NIFA + NIFL(147–519) + GTP + ATPγS 5.8 14 12 NIFA + GTP + AMPPNP 47 114 13 NIFA + NIFL(147–519) + GTP + AMPPNP 4.8 12 a aFinal component concentrations were: NIFA, 100 nM; NIFL(147–519), 200 nM; GTP, 4 mM; ATP, 3.5 mM; ADP, 50 μM; ATPγS, 250 μM; AMPPNP, 500 μM. CK/CP indicates the presence of the ATP regenerating system as described in Materials and methods. Activity of the NIFL–NIFA regulatory system is responsive to 2-oxoglutarate Since 2-oxoglutarate is an allosteric effector of the PII-like signal transduction proteins (Kamberov et al., 1995; Jiang et al., 1998c; Atkinson and Ninfa, 1999; Jiang and Ninfa, 1999), control experiments were performed to determine whether this ligand influenced the activities of either NIFA or NIFL. When ATP was the donor for nucleotide hydrolysis, transcriptional activation by NIFA was not significantly influenced by the addition of 2-oxoglutarate in the absence of NIFL (Figure 1A, compare lanes 2–5). However, we were surprised to find that when NIFL(147–519) was also present, 2-oxoglutarate relieved the inhibition of NIFA activity seen in the presence of ATP (Figure 1A, compare lane 7 with 8, and lane 9 with 10). Quantitation of the data from the phosphoimager revealed that almost total relief was achieved in the presence of the ATP regenerating system, presumably as a consequence of the removal of ADP, which is a more potent effector of NIFL inhibition than ATP (Figure 1B). In the absence of the regenerating system, relief of inhibition was not total and required higher concentrations of 2-oxoglutarate, perhaps reflecting the competing activities of ADP and 2-oxoglutarate (Figure 1B). By including GTP as the donor for nucleotide hydrolysis by NIFA, it was possible to examine the effect of 2-oxoglutarate at different ADP concentrations. The activity of NIFA in the presence of NIFL(147–519) was highly responsive to 2-oxoglutarate at a low concentration of ADP (10 μM), but deactivation of the inhibitory function of NIFL required a high level of 2-oxoglutarate when the ADP concentration was presumably saturating with respect to binding to NIFL(147–519) (Figure 1C). More extensive titrations in the presence of the ATP regenerating system revealed a sigmoidal response to 2-oxoglutarate, which was effective in the range 0.01–2 mM with an apparent Kact of ∼150 μM (Figure 2). This is within the physiological range of the 2-oxoglutarate concentration in E.coli, which varies from ∼0.1 mM under conditions of nitrogen excess to ∼0.9 mM under conditions of nitrogen limitation (Senior, 1975). When GTP only was present in the assay, NIFL(147–519) did not inhibit NIFA and 2-oxoglutarate had no apparent influence on the formation of open promoter complexes (Figure 2). 2-oxoglutarate therefore appears to counteract inhibition by NIFL when adenosine nucleotides are present, rather than influencing NIFA activity per se. To determine the specificity of the interaction we checked the ability of similar compounds and tricarboxylic acid (TCA) cycle intermediates to relieve inhibition by NIFL(147–519). The following compounds were tested: L-asparagine, L-aspartate, citrate, fumarate, L-glutamate, L-glutamine, 2-ketobutyrate, 3-oxoglutarate, pyruvate and succinate. None of these was effective at the concentration tested (2 mM; data not shown). Relief of inhibition in response to 2-oxoglutarate is therefore specific, since neither 3-oxoglutarate nor 2-oxobutyrate was effective. Since 2-oxoglutarate is an effector of PII-like regulatory proteins we performed further checks to ensure that our protein preparations were not contaminated with Ec PII or Ec GlnK, which might modulate the activity of NIFL. Western blotting with an anti-Ec PII antibody (which also cross-reacts with Ec GlnK) revealed no contaminating cross-reacting material in our protein preparations and a contaminant of the appropriate molecular weight also was not apparent on silver-stained gels (data not shown). These observations suggest that the influence of 2-oxoglutarate is exerted directly either through NIFL(147–519) or NIFA, or potentially both proteins. Figure 1.Influence of 2-oxoglutarate on the activity of NIFL and NIFA as determined by the formation of open promoter complexes. Assays contained NIFA (100 nM), either 3.5 mM ATP (A and B) or 4 mM GTP (C) with the addition of NIFL(147–519) (200 nM), ADP (50 μM), creatine kinase/creatine phosphate (CK/CP) and 2-oxoglutarate as indicated. (A) Example of the original data and controls. Lane 1, free DNA; lane 2, NIFA and ADP; lane 3, NIFA, ADP and CK/CP; lane 4, NIFA, ADP and 2-oxoglutarate (1.8 mM); lane 5, NIFA, ADP, 2-oxoglutarate (1.8 mM) and CK/CP; lane 6, NIFA, NIFL(147–519), 2-oxoglutarate (1.8 mM) and CK/CP; lane 7, NIFA, NIFL(147–519) and ADP; lane 8, NIFA, NIFL(147–519), ADP and 2-oxoglutarate (1.8 mM); lane 9, NIFA, NIFL(147–519), ADP and CK/CP; lane 10, NIFA, NIFL(147–519), ADP, CK/CP and 2-oxoglutarate (1.8 mM). The arrow indicates the mobility of the heparin-resistant open promoter complexes. (B) Influence of 2-oxoglutarate concentration on NIFL and NIFA activity in the presence of ATP and ADP. Reactions contained NIFA and ADP (circles), NIFA, NIFL(147–519) and ADP (squares) or NIFA, NIFL(147–519), ADP and CK/CP (triangles) plus the indicated concentration of 2-oxoglutarate. The percentage relief of NIFL inhibition represents the activity relative to the extent of NIFA activity (open promoter complex formation) in the absence of NIFL. (C) Influence of 2-oxoglutarate and ADP concentrations on NIFL activity in the presence of GTP. Reactions contained NIFA and NIFL(147–519) without 2-oxoglutarate (open squares), with 50 μM 2-oxoglutarate (open triangles), with 200 μM 2-oxoglutarate (closed triangles) and with 2 mM 2-oxoglutarate (closed squares). Relative activity is related to the activity of NIFA in the absence of NIFL. Download figure Download PowerPoint Figure 2.Influence of GTP and ATP on the response of NIFL(147–519) and NIFA to 2-oxoglutarate. Reaction conditions were similar to those in Figure 1 with the exception that assays contained either ATP (3.5 mM) or GTP (4 mM) and ADP was omitted. The ATP regenerating system (CK/CP) and 2-oxoglutarate (various concentrations) were present as indicated. (A) Control experiments and examples of the primary data in the presence of ATP. Lane 1, free DNA; lane 2, NIFA; lane 3, NIFA and 2-oxoglutarate (2 mM); lane 4, NIFA and CK/CP; lane 5, NIFA, 2-oxoglutarate and CK/CP; lane 6, NIFA and NIFL(147–519); lane 7, NIFA, NIFL(147–519) and CK/CP; lanes 8–20, NIFA, NIFL(147–519), CK/CP and 2-oxoglutarate (0.1 μM–2 mM). (B) Quantitative analysis of the data in the presence of ATP and CK/CP (lanes 8–20 in A) (squares) or GTP (triangles). Relative activity is related to the activity of NIFA and CK/CP in ATP (27% of total radioactivity was in the open complex) or NIFA in the presence of GTP (54% of total radioactivity was in the open complex). Download figure Download PowerPoint Ec PII but not Ec GlnK stimulates inhibition by A.vinelandii NIFL Since the ligand binding and regulatory properties of the signal transduction proteins Ec PII and Ec GlnK are well characterized, we sought to determine whether these proteins could influence NIFL(147–519) or NIFA activity. Although this system is heterologous we have shown previously that A.vinelandii NIFL(147–519) and NIFA are responsive to nitrogen sensing in vivo in E.coli, suggesting that these proteins interact with enteric nitrogen signalling components (Söderbäck et al., 1998). Previous studies have also shown that the binding of 2-oxoglutarate and ATP strongly influences the regulatory functions of PII-like proteins and their interactions with effectors (Kamberov et al., 1995; Jiang et al., 1998c; Atkinson and Ninfa, 1999; Jiang and Ninfa, 1999). In the absence of NIFL(147–519), the addition of Ec PII and Ec GlnK had no influence on NIFA activity in the presence of 2-oxoglutarate and ATP (Figure 3A, compare lanes 2 and 3). However, when both NIFL(147–519) and NIFA were present, the inhibitory function of NIFL was activated by the addition of Ec PII in the presence of 2-oxoglutarate and ATP; complete inhibition was observed at apparent stoichiometric levels of Ec PII (200 nM trimer) and NIFL(147–519) (200 nM dimer) (Figure 3B). Therefore, the presence of Ec PII counteracts the relief of NIFL inhibition seen with 2-oxoglutarate. Control experiments showed that neither Ec PII, Ec PII-UMP nor Ec GlnK was competent to relieve the inhibitory activity of NIFL in the absence of 2-oxoglutarate (Figure 3A, lanes 4–7). This activation of the inhibitory function of NIFL(147–519) by Ec PII was not observed when ATP was replaced by GTP, indicating that adenosine nucleotides are required to activate inhibition by NIFL(147–519) (Figure 3B). Activation of the inhibitory function of NIFL(147–519) was apparently conferred by the non-modified form of Ec PII, since fully uridylylated Ec PII did not activate NIFL(147–519) (Figure 3B). The ability to activate inhibition by NIFL was specific to Ec PII since Ec GlnK had no apparent influence on the inhibitory function of NIFL(147–519) in the presence of 2-oxoglutarate and ATP. Figure 3.Ec PII, but not Ec GlnK or Ec PII-UMP increases the inhibitory activity of NIFL in the presence of 2-oxoglutarate and ATP. Reactions contained component concentrations as in Figure 1 with the exception that ADP was omitted and 2-oxoglutarate was present at 0.1 mM. The regenerating system (CK/CP) was present in reactions except when GTP was used to promote open complex formation. (A) Control experiments. Lane 1, free DNA; lane 2, NIFA; lane 3, NIFA, 2-oxoglutarate, Ec PII (1 μM) and Ec GlnK (1 μM); lane 4, NIFA and NIFL(147–519); lane 5, NIFA, NIFL(147–519) and Ec PII (1 μM); lane 6, NIFA, NIFL(147–519) and Ec PII-UMP (1 μM); lane 7, NIFA, NIFL(147–519) and Ec GlnK (1 μM); lane 8, NIFA, NIFL(147–519) and 2-oxoglutarate. Note that 2-oxoglutarate was absent in lanes 1, 2 and 4–7. (B) Influence of paralogues on inhibition by NIFL. Additions were Ec PII in the presence of ATP (closed squares); Ec PII in the presence of GTP (closed triangles); Ec PII-UMP in the presence of ATP (open squares); Ec GlnK in the presence of ATP (open triangles). One hundred per cent relative activity represents the extent of open promoter complex formation in the absence of PII-like proteins. Download figure Download PowerPoint Work by Ninfa and his colleagues has demonstrated that the binding of a single molecule of 2-oxoglutarate to an Ec PII trimer favours the interaction of PII with its receptor NRII (NtrB) (Kamberov et al., 1995) and ATase (Jiang et al., 1998c). However, upon binding of additional molecules of 2-oxoglutarate, Ec PII is presumed to adopt a different conformation in which these interactions are disfavoured. We observed that increasing the concentration of 2-oxoglutarate from 100 μM to 2 mM had little influence on the ability of PII to activate the inhibitory function of NIFL(147–519), although some decrease in inhibition was observed at 4 mM (Figure 4). We assume that the NIFL–NIFA system is saturated at 1 mM 2-oxoglutarate (Figure 2) and that at this concentration there is sufficient excess ligand to saturate PII (Jiang et al., 1998c). It would therefore appear that the interaction of PII with the NIFL–NIFA system may not be as responsive to 2-oxoglutarate concentration when compared with other PII–receptor interactions. Figure 4.Influence of Ec PII on the inhibitory activity of NIFL in response to the 2-oxoglutarate concentration. Reaction conditions were identical to those in Figure 3 with ATP used to promote open complex formation. 2-oxoglutarate concentrations were 100 μM (squares), 2 mM (triangles) and 4 mM (circles). Download figure Download PowerPoint The PII-like protein encoded by A.vinelandii glnK stimulates inhibition by NIFL(147–519) In A.vinelandii there is apparently only a single copy of a gene encoding a PII-like protein, designated glnK (Meletzus et al., 1998). In order to determine the response of NIFL and NIFA to the homologous PII protein, we overexpressed and purified the native A.vinelandii Av PII and examined the influence of this protein on NIFL(147–519) and NIFA activity. As in the case of E.coli PII, Av PII had no apparent effect on NIFA activity (Figure 5A, compare lanes 2 and 3), but in the presence of 2-oxoglutarate and ATP it stimulated the inhibitory activity of NIFL(147–519) (Figure 5B). Complete inhibition was observed with stoichiometric levels of Av PII and NIFL(147–519). This inhibitory activity was also dependent on the presence of adenosine nucleotides since inhibition by the Av PII protein was not observed when ATP was replaced with GTP (Figure 5B). Figure 5.Av PII increases the inhibitory activity of NIFL in the presence of adenosine nucleotides and 2-oxoglutarate. Reaction conditions and component concentrations were identical to Figure 3. (A) Control experiments. Lane 1, free DNA; lane 2, NIFA; lane 3, NIFA, 2-oxoglutarate and Av PII (1 μM); lane 4, NIFA and NIFL(147–519); lane 5, NIFA, NIFL(147–519) and Av PII (1 μM); lane 6, NIFA, NIFL(147–519) and 2-oxoglutarate. Note that 2-oxoglutarate was absent in lanes 2, 4 and 5. (B) Influence of Av PII concentration on inhibition by NIFL in the presence of ATP (squares) and GTP (triangles). Activity on the y-axis is shown relative to the activity of NIFA in the absence of NIFL and Av PII. One hundred per cent activity represents the extent of open promoter complex formation in the absence of Av PII. Download figure Download PowerPoint Uridylylation state of the A.vinelandii PII-like protein modulates NIFL(147–519) inhibition The activity of Av PII is likely to be modulated by uridylylation since A.vinelandii contains a glnD gene encoding a homologue of E.coli UTase/UR. Mutations in glnD give rise to a Nif− phenotype in A.vinelandii, indicating that the activity of the glnD product may play a role in the regulation of nitrogen fixation (Contreras et al., 1991). We overexpressed and purified the A.vinelandii UTase/UR from E.coli. When incubated under conditions similar to those used with the E.coli enzyme, A.vinelandii GlnD protein catalysed the modification of Av PII. The modification was associated with a shift in the mobility of Av PII on native gels and the incorporation of label from [α-32P]UTP (Figure 6A). The modification was confirmed as uridylylation since tryptic digestion of the modified Av PII followed by MALDI-TOF analysis indicated an increase in mass of 306.1 Da (expected mass increase 306.2 Da) on peptide 48–58, which contains the uridylylation site. By quenching the uridylylation reaction at various times we obtained a range of partially modified Av PII trimers. Extrapolation of the data shown in Figure 6B suggests that significant r
Referência(s)