| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Institut für Mikrobiologie und Genetik, Technische Universität Darmstadt, D-64287 Darmstadt, Germany
Correspondence
Felicitas Pfeifer
pfeifer{at}bio.tu-darmstadt.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
DEex transformants that have incurred a deletion within the gvpD gene. A similar reduction was observed in Dex+Eex transformants harbouring both reading frames under fdx promoter control on two different plasmids. GvpD wild-type and also GvpD mutants were tested, and a significant reduction in the amount of GvpE was obtained in the case of GvpD wild-type and the super-repressor mutant GvpD3-AAA. In contrast, transformants harbouring GvpD mutants with alterations in the p-loop motif or the bR1 region still contained GvpE. Since the amount of gvpE transcript was not reduced, the reduction occurred at the protein level. These results underlined that a functional p-loop and the arginine-rich region bR1 of GvpD were required for the GvpD-mediated reduction in the amount of GvpE.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
; DasSarma et al., 1994). Hbt. salinarum PHH1 contains two different gvp gene clusters (the p-vac region located on plasmid pHH1, and the c-vac region on the chromosome), but forms spindle-shaped gas vesicles only by the expression of p-vac (Offner et al., 1996). The entire c-vac region is active in p-vac deletion variants, but only during the stationary growth phase (Englert et al., 1992a). Both vac regions are related but not identical to the gvp1 (p-vac) and gvp2 (c-vac) gene clusters of Hbt. salinarum NRC-1 (Ng et al., 2000). A third vac region, mc-vac (mc, mediterranei chromosomal), is found in Hfx. mediterranei, leading to gas vesicles in the stationary growth phase only (Englert et al., 1990).
gvp gene expression depends on the two regulatory proteins, GvpD and GvpE. The 21 kDa GvpE is a transcriptional activator and the larger GvpD protein (54–60 kDa) is required for the repression of gas vesicle formation. GvpE activates the promoters PA and PD located in front of gvpA and gvpD (Gregor & Pfeifer, 2001; Zimmermann & Pfeifer, 2003; Hofacker et al., 2004). A scanning mutagenesis performed with the 35 bp region upstream of PmcA-BRE in the mc-vac region determined that the sequence TGAAACGG-n4-TGAACCAA adjacent to the BRE element is important for GvpE-mediated activation, and a related sequence element is found in the promoters PpA of p-vac and PcA of c-vac (Gregor & Pfeifer, 2005). The GvpE protein dimerizes in solution and resembles a basic leucine-zipper protein, as suggested by molecular modelling of the C-terminal portion and mutational analyses (Krüger et al., 1998; Plößer & Pfeifer, 2002).
GvpD has a function in the repression of gas vesicle formation as demonstrated by transformation experiments with Hfx. volcanii. This halophilic archaeon lacks gvp genes and offers a clean genetic background for these investigations. Transformants containing an mc-vac region with an in-frame deletion in the mc-gvpD gene (D transformants) are gas vesicle overproducers (Englert et al., 1992b). The amount of gas vesicles is reduced to the wild-type level in D+Dex transformants, containing in addition to D the mc-gvpD reading frame expressed under ferredoxin promoter (Pfdx) control in the expression vector pJAS35 (Pfeifer et al., 1994). A p-loop motif (putative nucleotide-binding motif) located near the N terminus of GvpD and two arginine-rich regions (basic regions bR1 and bR2) are essential for the GvpD repressive function (Pfeifer et al., 2001). Comparison of the amino acid sequences of the three different GvpD proteins derived from p-vac, c-vac and mc-vac indicates highly conserved C- and N-terminal domains that are separated by an unconserved region. The size of the unconserved region is 70 aa in mcGvpD of Hfx. mediterranei, 61 aa in pGvpD of p-vac and 17 aa in cGvpD derived from the c-vac region of Hbt. salinarum (Englert et al., 1992a). Since the basal activity of the PmcA promoter is not reduced in the presence of GvpD, GvpD appears to have an indirect effect on PmcA promoter activity (Zimmermann & Pfeifer, 2003).
The two regulatory proteins of Hfx. mediterranei, mcGvpD and mcGvpE, are able to interact in vitro as demonstrated by affinity chromatography (Zimmermann & Pfeifer, 2003). This ability may be part of the regulation of gvp expression in vivo. Hfx. volcanii mcDEex transformants harbouring both consecutive genes expressed under Pfdx control in pJAS35 contain barely detectable amounts of mcGvpE, whereas this protein is present in large amounts in mcEex and mcDEex transformants (containing the same in-frame deletion in mc-gvpD as found in D). It appears that the presence of mcGvpD leads to a reduction in the amount of mcGvpE in these transformants. It is possible that this GvpD-mediated reducing effect on mcGvpE is initiated by the GvpD–GvpE interaction.
In this report we investigated this hypothesis in further detail. At first we studied whether the two regulatory proteins of the c-vac region of Hbt. salinarum, cGvpD and cGvpE, and also the heterologous pairs mcGvpD–cGvpE and cGvpD–mcGvpE were able to interact in vitro. In addition, previously constructed mcGvpD mutants with alterations in the p-loop motif or in one of the two basic regions were studied with respect to their ability to interact with GvpE. Additional deletion mutants of GvpD were constructed in order to assign the region required for GvpE interaction more precisely. From these analyses it appeared that this region is located neither in the N-terminal portion, including the p-loop motif, nor in the C-terminal domain. We further studied the influence of the various GvpD mutants on the amount of GvpE in Hfx. volcanii transformants. Mutations in the p-loop motif and bR1 abolished the reducing effect of GvpD on GvpE, suggesting that both regions are important for this property.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Prior to transformation, each construct was passaged through the E. coli dam– strain GM1674 (Palmer & Marinus, 1994) to avoid a halobacterial restriction barrier (Holmes et al., 1991). Transformants were selected on agar plates containing 6 µg mevinolin ml–1 (for selection of pWL102) and 0.2 µg novobiocin ml–1 (for selection of pJAS35). Mevinolin is a derivative of Lovastatin and was obtained as a generous gift of MSD Sharp and Dohme GmbH. The presence of the desired constructs in each transformant was controlled by DNA isolation and restriction analyses.
Constructs used for transformation of Hfx. volcanii WFD11.
The construction of mc-gvpDMut1-, mc-gvpDMut6-, mc-gvpDAEAE-, mc-gvpD3-ADA- and mc-gvpD3-AAA-pJAS35 (Pfeifer et al., 2001), mc-gvpE-pJAS35 (mcEex) (Gregor & Pfeifer, 2001), and c-gvpE-pJAS35 (cEex) (Krüger et al., 1998) has been described previously. The sequences of the oligonucleotide primers used in PCR to construct the deletion derivatives are given in Table 1.
|
) served as template for the construction of the various mc-gvpD deletion mutants. The fragment mc-gvpD-Cterm356 (or Cterm233 or Cterm119; the numbers state the number of amino acids derived from GvpD) was amplified by PCR using the oligonucleotides D-Cterm356 (or D-Cterm233 or D-Cterm119) and M13 Forward. The resulting fragments were inserted as NcoI-Acc65I fragments into pJAS35. The fragment mc-gvpD-Nterm318 (or Nterm362 or Nterm447) was obtained by PCR amplification using the oligonucleotides D-Nterm318 (or D-Nterm362 or D-Nterm447) and pJAS, and was inserted as an NcoI–Acc65I fragment into pJAS35. The p-loop fragment was produced by three consecutive PCRs. The first two PCRs were done using mcDex as template with the oligonucleotides D-ploop_r and pJAS (or D-ploop_f and M13 Forward) as primer pairs. The resulting fragments were fused in the third PCR using M13 Forward and pJAS. G–L was produced in an analogous manner using the primer pairs D-G32-L35_r and pJAS (or D-G32-L35_f and M13 Forward) for the first two PCRs.
The expression vector pWLfdx contains the Pfdx promoter (Pfeifer et al., 1993) as a 73 bp segment in plasmid pWL102 (Lam & Doolittle, 1989). For the construction of this vector, the Pfdx promoter was amplified as a 73 bp segment using the pJAS35 vector as template and the oligonucleotides fdxProm-BspHI and M13 Forward. The fragment was inserted in pWL102, resulting in the pWLfdx vector, carrying gvpA as an exchangeable fragment. ORFs can be inserted as NcoI–Acc65I fragments. The constructs mc-gvpD-pWLfdx, mc-gvpD3-AAA-pWLfdx, mc-gvpDMut6-pWLfdx and c-gvpE-pWLfdx were obtained using the respective mc-gvpD-, mc-gvpDMut6-, mc-gvpD3-AAA- and c-gvpE-pJAS35 constructs as DNA source to receive the respective NcoI–Acc65I fragments that were cloned into pWLfdx.
Expression and purification of Gvphis fusion proteins from E. coli M15.
The constructs c-gvpD-pQE8, c-gvpE-pQE8 (Krüger & Pfeifer, 1996) and mc-gvpD-pQE8, mc-gvpE-pQE8 (Zimmermann & Pfeifer, 2003) were used to produce the Gvphis fusion proteins. The expression of the respective gvphis reading frame in vector pQE8 (Qiagen) yields N-terminal hexa-histidine-tagged Gvp proteins. The procedure for the expression and isolation of the Gvphis proteins has been reported previously (Zimmermann & Pfeifer, 2003). For affinity chromatography, the purified Gvphis proteins were dialysed (since purification was carried out under denaturing conditions) against three buffers containing decreasing urea and increasing KCl concentrations in 10 mM Tris/HCl buffer, pH 7.2. Two different refolding procedures were used. (1) Buffer 1, 4 M urea, 0.8 M KCl; buffer 2, 2 M urea, 1.6 M KCl; buffer 3, 2.5 M KCl; used for mcGvpEhis, mcGvpDhis and cGvpEhis. (2) Buffer 1, 4 M urea, 0.8 M KCl, 10 % glycerol; buffer 2, 2 M urea, 1 M KCl, 10 % glycerol; buffer 3, 1 M KCl, 10 % glycerol; used for cGvpDhis. The latter buffers contained glycerol and KCl only up to 1 M to avoid precipitation of the cGvpDhis protein.
Affinity chromatography with Gvphis-loaded Ni-NTA matrices and cell lysates of Hfx. volcanii transformants.
Affinity chromatography was carried out as described previously (Zimmermann & Pfeifer, 2003) with the following modifications. Ni-NTA agarose matrix (100 µl) was loaded with 400 µg His-tagged Gvp proteins for 1 h at 20 °C and washed twice for 10 min with 1 ml wash buffer 1 (2.5 M KCl, 10 mM Tris/HCl, pH 7.2). The Gvphis-loaded matrix was incubated for 30 min with 2 mg soluble proteins from Hfx. volcanii WFD11 transformants expressing the reading frame of the putative binding partner under Pfdx promoter control in pJAS35. Soluble proteins were isolated in the stationary growth phase. The matrix was washed four times for 10 min with 1 ml wash buffer 2 (wash buffer 1+50 mM imidazole). An unspecific binding of the putative binding partner to the Ni-NTA agarose could be avoided by the addition of imidazole to wash buffer 2. Elution of the Gvphis proteins together with the potential interacting partner was carried out by incubating the matrix with 100 µl elution buffer (0.5 M imidazole, 10 mM Tris/HCl, pH 7.2), two times for 20 min. In each case, 10 µl of the elution fractions were analysed by Western analysis.
Preparation of crude extracts from Hfx. volcanii transformants and Western analysis.
To determine the stability of GvpE in Dex+Eex transformants, samples of Hfx. volcanii transformants were taken during the exponential and stationary growth phase, centrifuged at 12 000 g and resuspended in 400 µl TE buffer containing 1 µg DNase I ml–1. The suspension was dialysed against 10 mM Tris/HCl, pH 7.2, at 4 °C overnight and centrifuged for 20 min at 12 000 g for membrane removal. The protein concentration was determined by the Bradford assay (Ausubel et al., 1988) using BSA as standard. The soluble proteins (5–20 µg) were separated on a 12 % Tricine SDS-polyacrylamide gel (Schägger & von Jagow, 1987). Western analyses were performed as described previously (Pfeifer et al., 2001). The GvpD and GvpE proteins were detected by the respective antisera (Krüger & Pfeifer, 1996; Zimmermann & Pfeifer, 2003) at a dilution of 1 : 1000 in blocking buffer, and the reacting antibodies were detected using the ECL detection system (Amersham Life Science).
For the isolation of soluble proteins from Hfx. volcanii transformants for affinity chromatography, cells were harvested at OD600 1.2–1.5, resuspended in 0.1 vol. lysis buffer (2.5 M KCl, 50 mM MgCl2, 1 mM EDTA, 5 % glycerol and 50 mM Tris/HCl, pH 8.0) and disrupted by sonication. Insoluble fragments and membranes were removed by centrifugation.
Isolation of RNA and transcript analysis.
RNA was isolated from Hfx. volcanii transformants according to the single step method of Chomczynski & Sacchi (1987). Northern analysis involved electrophoresis of 5 µg RNA on denaturing, formaldehyde-containing 1 % (w/v) agarose gels, followed by transfer to nylon membranes (Ausubel et al., 1988).
Strand-specific RNA probes were synthesized using a 240 bp Acc65I fragment derived from mc-gvpE cloned in pJAS35 or a 367 bp XcmI–EcoRI fragment derived from mc-gvpA cloned in pBluescript as template for the T3/T7 polymerase system. The RNA was labelled using the DIG RNA Labelling Kit from Roche (Germany). Strand-specific DNA probes were synthesized using the c-gvpE reading frame cloned in pBluescript-II KS+ (Krüger et al., 1998) as template for a PCR with the oligonucleotides E-Ex-HindIII and E-Mut-ZIP S (Plößer & Pfeifer, 2002) as primers resulting in a 152 bp fragment which was labelled using the DIG DNA Labelling Mix (Roche). Northern hybridization was carried out as described by Ausubel et al. (1988), but the hybridization solution contained 10 % (w/v) dextran sulfate (Sigma), 1 % (w/v) SDS and 0.5 % (w/v) skim milk powder.
DNA sequence analysis.
The sequences of all constructs were confirmed by DNA sequencing according to the Sanger method using the Sequi-Therm EXCEL II Long-Read DNA Sequencing Kit-LC protocol (Biozym). The fragments were separated and determined by a Licor DNA sequencer.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). In the case of cGvpD, the concentration of KCl was only increased up to 1 M in 10 % glycerol to avoid protein precipitations. Halophilic archaea use the salt-in strategy to adapt to their high salt environment, and most proteins require high KCl concentrations for a proper function. A successful refolding of halophilic proteins under comparable conditions has been reported for a TBP protein of Hbt. salinarum and the glucose dehydrogenase of Hfx. mediterranei (Soppa & Link, 1997; Pire et al., 2001).
The Ni-NTA matrices were either loaded with cGvpEhis, mcGvpEhis (as positive control) or used as pure matrix (negative control) and incubated with cell lysates of the mcDex transformant carrying the mc-gvpD reading frame expressed under Pfdx control in pJAS35. The three matrices were washed extensively and the GvpEhis proteins plus putative binding partner were eluted with a buffer containing 0.5 M imidazole. After separation of the proteins by SDS-PAGE, the samples were analysed by Western analysis using an antiserum raised against mcGvpD to determine whether mcGvpD bound to the GvpEhis columns. The pure Ni-NTA matrix showed a minor hybridization band due to unspecific binding of GvpD (Fig. 1a). Larger amounts of the mcGvpD protein were found in samples derived from the mcEhis matrix, underlining the previously reported results (Zimmermann & Pfeifer, 2003). Also the cEhis matrix retained mcGvpD in similar amounts to the mcEhis matrix, demonstrating that the heterologous cGvpE protein bound mcGvpD (Fig. 1a). Similar experiments done with a lysate of a cDex transformant (containing c-gvpD derived from c-vac in pJAS35) indicated that also cGvpD bound to the mcEhis and cEhis matrices (data not shown). The reciprocal experiment using mcDhis and cDhis matrices and cell lysates of cEex (or mcEex) transformants containing cGvpE (or mcGvpE) resulted in the binding of cGvpE (or mcGvpE) to both matrices (Fig. 1b and data not shown). The cGvpE monomer as well as the cGvpE dimer (Plößer & Pfeifer, 2002) could be visualized by Western analysis. A third band appeared in the sample of mcDhis+cEex, possibly representing an unspecific binding. In contrast, the pure Ni-NTA matrix did not retain GvpE (Fig. 1b). The results of these experiments suggested that the two regulatory proteins cGvpD and cGvpE derived from the c-vac region of Hbt. salinarum were also able to interact. In addition, this was true for the heterologous pairs of the regulators (mcGvpD–cGvpE and cGvpD–mcGvpE) derived from c-vac of Hbt. salinarum or the mc-vac region of Hfx. mediterranei. Similar amounts of GvpD protein were recovered from these matrices when homologous or heterologous interaction partners were used.
|
), and two additional mcGvpD mutants harbouring either a deletion of the p-loop motif (p-loop) or a deletion of the four adjacent amino acids (G–L=G32–L35) (see Fig. 2) were used to investigate whether the p-loop and its surroundings were involved in the GvpD–GvpE interaction. GvpDMut1 and GvpDMut6 are unable to repress the gas vesicle overproduction in Hfx. volcanii D transformants (Pfeifer et al., 2001), and the two novel GvpD mutants (Dp-loop and DG–L) were also tested for this feature. The respective genes cloned in pJAS35 were used to complement D transformants.
|
D) form pinkish opaque colonies (Fig. 3a). D+Dex transformants contained significantly reduced amounts of gas vesicles compared to D transformants as seen by the red transparent colonies (D+Dex, Fig. 3a) (Pfeifer et al., 1994, and this report). The respective D+Dp-loop and D+DG–L transformants formed pinkish colonies, indicating that both mcGvpD deletion variants (p-loop and G–L) were unable to reduce the amount of gas vesicles (Fig. 3a). Northern analysis indicated large amounts of mc-gvpA mRNA (encoding the major gas vesicle structural protein GvpA) in D transformants, especially in stationary growth phase (Fig. 3b). The amount of this transcript was significantly reduced in the D+Dex transformant, demonstrating that the downregulation occurred at the level of PmcA activity. Both D+Dp-loop and D+DG–L transformants yielded large amounts of mc-gvpA mRNA, underlining their defect in the repression of PmcA (Fig. 3b). Methylene-blue stained 23S rRNA indicated that similar amounts of RNA were applied to each slot (Fig. 3b, bottom). Western analysis confirmed the presence of similar amounts of GvpD in all D+Dex transformants, whereas the protein was absent in D transformants as expected (Fig. 3c). Thus, the lack of the repressive activity in D+Dp-loop and D+DG–L was indeed due to the respective p-loop deletions and not to an absence of GvpD. These results underlined the importance of the p-loop motif and of the surrounding amino acid sequences for the GvpD repressive function.
|
p-loop and GvpDG–L, were able to bind mcEhis (Fig. 3d
and data not shown), demonstrating that the region of the p-loop motif in GvpD was not involved in the GvpD–GvpE interaction. A summary of the GvpD mutants and their properties is given in Table 2.
|
; Pfeifer et al., 2001). The bR1 mutant GvpDAEAE and the two bR2 mutants GvpD3-ADA and GvpD3-AAA were analysed for their ability to bind to an mcEhis matrix. As demonstrated previously, mutant GvpD3-AAA acts in vivo as a super-repressor of gas vesicle formation, whereas the other two mcGvpD mutant proteins are unable to reduce overexpression in D and cannot repress GvpE-mediated activation of the mc-gvpA promoter PmcA (Pfeifer et al., 2001). GvpD mutant proteins were produced in similar amounts in the respective transformants as determined by Western analysis using the mcGvpD antiserum (Fig. 4a). Binding studies were performed using mcEhis-Ni-NTA matrices. All mcGvpD mutant proteins were recovered in similar amounts as the mcGvpD wild-type protein or in slightly enhanced amounts in the case of the super-repressor GvpD3-AAA, whereas the pure Ni-NTA agarose bound very minor amounts of mcGvpD (Fig. 4b). These results suggested that mutations in the bR1 and bR2 regions did not abolish GvpE binding.
|
). The mcGvpD derivatives lacked different parts of the N- or C-terminal domains: Nterm318 consisted of the first 318 aa (36 kDa protein), Nterm362 contained the first 362 aa up to the unconserved region of mcGvpD (41 kDa) and Nterm447 possessed the first 447 aa of mcGvpD (50 kDa). Additional mcGvpD derivatives lacked N-terminal portions: Cterm356 (39 kDa) possessing the last 356 aa after a deletion of the first 189 aa, Cterm233 (26 kDa) consisted of the last 233 aa and Cterm119 (13 kDa) consisted of the conserved C terminus of mcGvpD, including bR2 (Fig. 2). The presence of the different mcGvpD deletion derivatives was checked by Western analysis in lysates of Hfx. volcanii transformants containing the respective reading frames in the expression vector pJAS35. Except for the two smallest variants, Cterm233 and Cterm119, each mcGvpD deletion variant was present in sufficient amounts (data not shown). The two smallest C-terminal variants were almost not detectable by Western analysis, suggesting that both proteins were produced in minor amounts, were unstable in the transformants or were not detected very well by the antiserum (data not shown). Therefore, these variants were not useful for our analyses.
None of these four GvpD deletion derivatives (Nterm318, Nterm362, Nterm447 and Cterm356) was able to reduce gas vesicle formation in D transformants (data not shown). Binding assays performed with the mcEhis matrix indicated that the mcGvpD variants Nterm318 and Nterm362 were not retained by mcEhis, implying that these portions of mcGvpD were not sufficient for GvpE binding (Fig. 5a, b). The deletion variant Nterm447 (lacking the C-terminal domain) bound mcGvpEhis (Fig. 5c), as did the Cterm356 protein, suggesting that each of these proteins harboured the interaction site (Fig. 5d). These results implied that neither the immediate N-terminal region (lacking Cterm356), nor the C-terminal domain (lacking Nterm447) were required for GvpE binding, but rather a central portion close to the unconserved region was involved. These results are summarized in Table 2.
|
DEex or Eex transformants (Zimmermann & Pfeifer, 2003). These data suggest that the presence of GvpD causes a strong reduction in the amount of GvpE in DEex transformants.
To investigate this effect of GvpD on the amount of GvpE in further detail, we tested the various mcGvpD mutants for their ability to cause a reduction in the amount or even a lack of the GvpE protein in transformants. To facilitate these analyses, we first investigated whether a strong reduction in the amount of GvpE is also seen when GvpD is produced from a different plasmid (i.e. Dex+Eex versus DEex transformants). For this purpose, a second expression vector (pWLfdx) was constructed harbouring the ferredoxin promoter Pfdx as a 73 bp segment in plasmid pWL102 (Lam & Doolittle, 1989) (see Methods). The use of the Pfdx promoter to express both gvp genes should lead to similar amounts of Gvp proteins in the transformants, since both expression vectors (pJAS35 and pWLfdx) are present in comparable amounts in the cell.
In the first series of experiments the mcEex construct (mc-gvpE reading frame expressed in pJAS35) was used to produce mcGvpE, and the pWLfdx vector was used to produce mcGvpD or the respective mcGvpD mutant proteins (DpWL-fdx constructs). The transformants DpWL-fdx (=mcD), DpWL-fdx+pJAS (=mcD+pJAS) and mcEex+DpWL-fdx (=mcEex+mcD) were analysed for the presence of mcGvpD by Western analysis (Fig. 6a). In all three transformants, GvpD was detected in similar amounts in both exponential and stationary growth phases. With respect to GvpE, the presence of this protein was analysed in mcEex and mcEex+pWLfdx (as controls), as well as in mcEex+mcD transformants. Similar amounts of mcGvpE were found in the two control transformants mcEex and mcEex+pWLfdx, whereas only minor amounts of mcGvpE were detected in the mcEex+mcD transformants (Fig. 6b). The latter result demonstrated that the amount of mcGvpE was indeed reduced in the presence of mcGvpD and that both genes could be supplied on different vector plasmids to test for this property. The mutants mcGvpDMut6 and mcGvpD3-AAA were analysed in similar experiments. Both contained GvpD, but the GvpDMut6 amount was slightly reduced (Fig. 6a). The transformant mcEex+DMut6 yielded a slight reduction in the amount of mcGvpE, whereas GvpE was almost undetectable in the mcEex+D3-AAA transformants (Fig. 6b). The latter result suggested that the presence of the super-repressor protein mcGvpD3-AAA led to a significant reduction in the amount of mcGvpE. The slightly reduced amount of mcGvpE observed in the p-loop defective mcGvpDMut6 could be due to the reduced amount of mcGvpDMut6 observed in this transformant (Fig. 6a). However, similar experiments with the p-loop deletion mutants yielded no reduction in the amount of GvpE (see below), implying that a functional p-loop in GvpD was required to reduce the amount of GvpE in mcEex+mcD transformants.
|
). Similar amounts of mc-gvpE transcripts (in each case more in the stationary growth phase) were seen in mcEex and mcEex+pWLfdx transformants, and slightly reduced but comparable amounts of transcript were seen in the mcEex+DpWL-fdx, mcEex+DMut6 and mcEex+D3-AAA transformants. The vector pJAS35 lacks a transcriptional termination signal, leading to transcripts that are even longer than the 600 bp mc-gvpE transcript. No hybridization signal was detectable in the RNA of the mcE-negative control transformant (Fig. 6c). Since the mcEex+DpWL-fdx, mcEex+DMut6 and mcEex+D3-AAA transformants yielded different amounts of GvpE protein but similar amounts of mc-gvpE transcript, these results implied that the reducing effect of GvpD did not occur at the level of transcription, but rather at the protein level.
All other mutant mc-gvpD genes were only available as pJAS35 constructs (mcDex). Because of this we analysed the ability of these mutants to reduce the amount of GvpE in a second series of transformants containing the c-gvpE reading frame on vector pWLfdx (mcDex+cEpWL-fdx) (Fig. 7a, b). The reducing effect of mcGvpD wild-type on the amount of cGvpE was clearly seen compared to the cEfdx+pJAS transformant containing instead of mcDex the empty pJAS35 vector (Fig. 7b). All transformants harbouring the mutated mc-gvpD reading frames (mcDMut1, mcDAEAE, mcD3-ADA, mcDp-loop, mcDG–L, Cterm356 and Nterm362) produced the respective mcGvpD protein in sufficient amounts (Fig. 7a). With respect to cGvpE, none of these mcGvpD mutants caused a reduction in the amount of cGvpE; this protein was detectable in samples derived from the exponential and stationary growth phases (Fig. 7b). The mcGvpD mutants Nterm318 and Nterm447 did not cause a reduction in the amount of cGvpE either (data not shown). Northern analysis to determine the amount of c-gvpE mRNA in each of these transformants yielded comparable amounts of the c-gvpE transcripts in all cases (Fig. 7c), again suggesting that the reduction in the amount of GvpE in the presence of GvpD wild-type was caused at the protein level. A summary of these results is given in Table 2.
|
p-loop, mcGvpDG–L or the mutant mcGvpDAEAE (alterations in the arginine-rich region bR1) contained cGvpE (Fig. 7b), these GvpD mutants were unable to initiate the lack of cGvpE. These results demonstrated that an intact p-loop motif and an intact bR1 were required for this property. In the case of the bR2 region, the mcGvpD3-ADA mutant did not affect the amount of cGvpE in vivo (Fig. 7b). This was in contrast to the result obtained with the super-repressor protein mcGvpD3-AAA described above, where GvpE was not detectable (Fig. 6b). Both mcGvpD mutants differ only in the alteration of R495 (R495D versus R495A) in bR2, suggesting that a negatively charged amino acid in this position destroyed the ability of GvpD to induce a reduction in the amount of GvpE. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Using affinity chromatography we could demonstrate that cGvpD and cGvpE of Hbt. salinarum also interact, as do the heterologous pairs mcGvpD–cGvpE and cGvpD–mcGvpE. The GvpD proteins of Hbt. salinarum and Hfx. mediterranei exhibit an identity of 71 %, and cGvpE and mcGvpE are 41 % identical to each other (Horne et al., 1991). Because of the observed heterologous interaction one could assume that the region of GvpD involved in the interaction is located in one of the conserved portions and not in the unconserved region that separates the N- and C-terminal domains of GvpD.
Several mcGvpD mutants with alterations in the p-loop motif or the two arginine-rich basic regions bR1 and bR2 (Pfeifer et al., 2001) were also tested in vitro for their ability to bind GvpE, but none of them lost the ability to interact with mcGvpEhis. The region of the p-loop motif was not required for the interaction since the deletion mutants mcDploop and mcDG–L were still able to bind GvpE. The same result was obtained with the mcGvpD deletion variant Cterm356 which had a deletion of the N-terminal 189 aa and still bound mcGvpE, whereas the GvpD mutants Nterm318 and Nterm362 (consisting of the first 318 or 362 aa, including the p-loop motif) were unable to bind mcGvpE. Thus, the N-terminal portion of GvpD, including the p-loop motif, was not required for GvpE binding. The hypothesis that the conserved C-terminal domain of mcGvpD might be involved in the interaction was ruled out, since the deletion variant Nterm447 (lacking 98 aa of the C terminus) still bound mcGvpEhis. These results rather suggested that the interaction site was located in the central portion of mcGvpD. The unconserved region is certainly not involved since cGvpD lacks almost all of this region.
The different mcGvpD mutants were also used to test a second feature of GvpD in vivo, namely that the mcGvpE protein is almost undetectable in the presence of mcGvpD in mcDEex transformants (Zimmermann & Pfeifer, 2003), implying that mcGvpD is able to cause a reduction in the amount of mcGvpE in transformants. To test whether the p-loop motif, bR1 and bR2 are required for this behaviour, transformants were produced containing gvpD and gvpE on two different plasmids. The mcDpWL-fdx+mcEex transformant, carrying both reading frames on two separate vector plasmids and expressed under Pfdx control, contained a significantly lower amount of mcGvpE, whereas mcGvpE was present in large amounts in pWLfdx+mcEex, i.e. in the absence of mcGvpD. Also, the presence of the super-repressor protein mcGvpD3-AAA resulted in barely detectable amounts of mcGvpE in mcD3-AAA+mcEex transformants. We further tested the effect of mcGvpD on the heterologous cGvpE in mcDex+cEpWL-fdx transformants, and the amount of cGvpE was significantly reduced as well. In contrast, the transformants harbouring cEpWL-fdx plus the various mcGvpD mutants DMut1, DMut6, Dp-loop, DG–L, DAEAE, DCterm356 or D3-ADA still contained cGvpE in detectable amounts, implying that each of these mutants lost the ability to induce a reduction in the amount of GvpE. Since the amount of gvpE transcripts was similar in all of the Eex+Dex and Eex+
transformants, the reducing effect of GvpD on GvpE must have occurred at the protein level.
In summary, these results suggest that an intact p-loop motif and an intact arginine-rich region bR1 are important for the ability of GvpD to induce a breakdown of GvpE in vivo. Interestingly, all mcGvpD mutants unable to reduce the amount of mcGvpE in transformants were also unable to repress the GvpE-mediated activation of the gvpA promoter in D transformants. The lack of the GvpE-reducing function in these ‘inactive’ GvpD proteins is presumably the reason for the overproduction of gas vesicles in these transformants. These results suggest that GvpD is not a gene repressor, but acts at the level of the activator inducing the degradation of this protein.
The exact mode of the GvpD–GvpE interaction is not yet known. It is possible that the interaction results in a phosphorylated or otherwise structurally altered GvpE protein that is degraded faster. However, phosphorylation of GvpE could not be detected. Interestingly, the N-terminal portion of GvpD exhibits a significant similarity to proteins like RadA or KaiC belonging to the RecA superfamily of ATPases (Leipe et al., 2000; Iyer et al., 2004), especially in the amino acid sequence and spacing of the p-loop motif and the arginine-rich region bR1. The loss of the GvpD function in mutants of the p-loop and bR1 motifs demonstrated that these two regions are important features for the function of this protein. Members of the RecA superfamily of ATPases form hexameric ring structures (or other oligomeric ring structures) and use the binding of ATP in the p-loop region to stabilize this structure. Whether GvpD forms a multimer or hexamer is currently under investigation. Another interesting point is that a subfamily of the AAA+ superfamily, the AAA ‘unfoldases’, associate with proteasomes and catalyse the energy-dependent degradation of proteins (Maupin-Furlow et al., 2004). Future experiments will help to unravel the outcome of the interaction of this unusual pair of regulatory proteins.
|
ACKNOWLEDGEMENTS |
|---|
Edited by: J. van der Oost
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Chomczynski, P. & Sacchi, N. (1987). Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162, 156–159.[Medline]
DasSarma, S., Arora, P., Lin, F., Molinari, E. & Yin, L. (1994). Wild-type gas vesicle formation requires at least 10 genes in the gvp gene-cluster of Halobacterium halobium plasmid pNRC100. J Bacteriol 176, 7646–7652.
Englert, C., Horne, M. & Pfeifer, F. (1990). Expression of the major gas vesicle protein gene in the halophilic archaebacterium Haloferax mediterranei is modulated by salt. Mol Gen Genet 222, 225–232.[CrossRef][Medline]
Englert, C., Krüger, K., Offner, S. & Pfeifer, F. (1992a). Three different but related gene clusters encoding gas vesicles in halophilic archaea. J Mol Biol 227, 586–592.[CrossRef][Medline]
Englert, C., Wanner, G. & Pfeifer, F. (1992b). Functional analysis of the gas vesicle gene cluster of the halophilic archaeon Haloferax mediterranei defines the vac-region boundary and suggests a regulatory role for the gvpD gene or its product. Mol Microbiol 6, 3543–3550.[Medline]
Gregor, D. & Pfeifer, F. (2001). The use of a halobacterial bgaH reporter gene to analyse the regulation of gene expression in halophilic archaea. Microbiology 147, 1745–1754.
Gregor, D. & Pfeifer, F. (2005). In vivo analyses of constitutive and regulated promoters in halophilic archaea. Microbiology 151, 25–33.
Hofacker, A., Schmitz, K.-M., Cichonczyk, A., Sartorius-Neef, S. & Pfeifer, F. (2004). GvpE- and GvpD-mediated transcription regulation of the p-gvp genes encoding gas vesicles in Halobacterium salinarum. Microbiology 150, 1829–1838.
Holmes, M. L., Nuttall, S. D. & Dyall-Smith, M. (1991). Construction and use of halobacterial shuttle vectors and further studies on Haloferax DNA gyrase. J Bacteriol 12, 3807–3813.
Horne, M., Englert, C., Wimmer, C. & Pfeifer, F. (1991). A DNA region of 9 kbp contains all genes necessary for gas vesicle synthesis in halophilic archaebacteria. Mol Microbiol 5, 1159–1174.[CrossRef][Medline]
Iyer, L. M., Leipe, D. D., Koonin, E. V. & Aravind, L. (2004). Evolutionary history and higher order classification of AAA+ ATPases. J Struct Biol 146, 11–31.[CrossRef][Medline]
Krüger, K. & Pfeifer, F. (1996). Transcript analysis of the c-vac region, and differential synthesis of the two regulatory gas-vesicle proteins GvpD and GvpE in Halobacterium salinarium PHH4. J Bacteriol 178, 4012–4019.
Krüger, K., Hermann, T., Armbruster, V. & Pfeifer, F. (1998). The transcriptional activator GvpE for the halobacterial gas vesicle genes resembles a basic region leucine-zipper regulatory protein. J Mol Biol 279, 761–771.[CrossRef][Medline]
Lam, W. L. & Doolittle, W. F. (1989). Shuttle vectors for the archaebacterium Halobacterium volcanii. Proc Natl Acad Sci U S A 86, 5478–5482.
Leipe, D. D., Aravind, L., Grishin, N. V. & Koonin, E. V. (2000). The bacterial replicative helicase DnaB evolved from a RecA duplication. Genome Res 10, 5–16.
Maupin-Furlow, J. A., Gil, M. A., Karadzic, I. M., Kirkland, P. A. & Reuter, C. J. (2004). Proteasomes: perspectives from the archaea. Front Biosci 9, 1743–1758.[Medline]
Ng, W. L., Kennedy, S. P., Mahairas, G., Berquist, B., Pan, M., Shukla, H., Lasky, S., Baliga, N., Thorsson, V. & other authors (2000). Genome sequence of Halobacterium species NRC-1. Proc Natl Acad Sci U S A 97, 12176–12181.
Offner, S., Wanner, G. & Pfeifer, F. (1996). Functional studies of the gvpACNO operon of Halobacterium salinarium reveal that the GvpC protein shapes gas vesicles. J Bacteriol 178, 2071–2078.
Palmer, B. & Marinus, M. (1994). The dam and dcm strains of Escherichia coli – a review. Gene 143, 1–12.[CrossRef][Medline]
Pfeifer, F. & Ghahraman, P. (1993). Plasmid pHH1 of Halobacterium salinarum: characterization of the replicon region, the gas vesicle gene cluster and insertion elements. Mol Gen Genet 238, 193–200.[Medline]
Pfeifer, F., Griffig, J. & Oesterhelt, D. (1993). The fdx gene encoding the 2Fe–2S ferredoxin of Halobacterium salinarium (H. halobium). Mol Gen Genet 239, 66–71.[Medline]
Pfeifer, F., Offner, S., Krüger, K., Ghahraman, P. & Englert, C. (1994). Transformation of halophilic archaea and investigation of gas vesicle synthesis. Syst Appl Microbiol 16, 569–577.
Pfeifer, F., Zotzel, J., Kurenbach, B., Röder, R. & Zimmermann, P. (2001). A p-loop motif and two basic regions in the regulatory protein GvpD are important for the repression of gas vesicle formation in the archaeon Haloferax mediterranei. Microbiology 147, 63–73.
Pire, C., Esclapez, J., Ferrer, J. & Bonete, M.-J. (2001). Heterologous overexpression of glucose dehydrogenase from the halophilic archaeon Haloferax mediterranei, an enzyme of the medium chain dehydrogenase/reductase family. FEMS Let 200, 221–227.
Plößer, P. & Pfeifer, F. (2002). A bZIP protein from halophilic archaea: structural features and dimer formation of cGvpE from Halobacterium salinarum. Mol Microbiol 45, 511–520.[CrossRef][Medline]
Schägger, H. & von Jagow, G. (1987). Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 166, 368–379.[CrossRef][Medline]
Soppa, J. & Link, T. A. (1997). The TATA-box-binding protein (TBP) of Halobacterium salinarum. Cloning of the tbp gene, heterologous production of TBP and folding of TBP into a native conformation. Eur J Biochem 249, 318–324.[Medline]
Zimmermann, P. & Pfeifer, F. (2003). Regulation of gas vesicle formation in Haloferax mediterranei: the two regulatory proteins GvpD and GvpE interact. Mol Microbiol 49, 783–794.[CrossRef][Medline]
Received 20 November 2006;
revised 5 January 2007;
accepted 9 January 2007.
This article has been cited by other articles:
|
|
 |
|
  M. Bauer, L. Marschaus, M. Reuff, V. Besche, S. Sartorius-Neef, and F. Pfeifer Overlapping activator sequences determined for two oppositely oriented promoters in halophilic Archaea Nucleic Acids Res., February 2, 2008; 36(2): 598 - 606. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
