INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

It is becoming increasingly apparent that the transcriptional apparatus of archaea (archaebacteria) resembles the system in eucaryotes more than that in eubacteria. RNA polymerases from archaea structurally and functionally resemble those of eucaryotes, especially RNA polymerase II (23). Moreover, a TATA box-like element with consensus sequence 5'-TTA(T/A)ATA-3' (32) is typically found ca. 25 bp upstream of archaeal transcription starts. No evidence of eubacterium-like sigma factors or promoters in archaea has been found. Instead, homologs of the TATA-binding protein (TBP) (27, 30, 33) and transcription factors TFIIB (10, 31) and TFIIS (24) have been found in archaea. It has been shown that TBP and TFIIB homologs (along with RNA polymerase and promoter DNA) are sufficient for transcription initiation in a cell-free transcription system (37) and that the Methanococcus jannaschii genome contains open reading frames encoding TBP, TFIIB, and TFIIS homologs (3), but no homologs of TFIIA have been found. There have also been demonstrations of functional interactions between archaeal and eucaryotic transcription apparatus components (30, 31, 33, 38).

Thus far the only DNA sequence element other than the TATA box and the initiation site that has been found to promote transcription is a purine-rich region usually (but not always) present directly upstream of archaeal TATA box-like elements (29). Despite these advances, much remains to be learned about transcription and its initiation and regulation in archaea. The size and makeup of the archaeal transcription initiation complex is not known nor is the potential role of upstream activators.

Evidence for regulation of the expression of some archaeal genes at the level of transcription has accrued. Examples include carbon monoxide dehydrogenase genes in Methanosarcina thermophila (36), heat shock genes in Haloferax volcanii (22) and Methanosarcina mazei (9, 26), and nitrogenase genes in methanogens (6, 35). An interesting system in Halobacterium halobium regulates the synthesis of the light-driven proton transporter bacteriorhodopsin (14). In that system, the bat gene product is believed to be a trans-acting inducer which responds to oxygen. Interestingly, a 124-amino-acid stretch of this protein shows significant similarity (30% identity, 56% similarity) to the N terminus of the Klebsiella NifL protein, which is an oxygen-sensing protein that inhibits NifA protein function in eubacteria (17). In eubacteria, the sigma subunit of RNA polymerase is essential for promoter recognition and transcriptional initiation, and several alternative sigma factors have been reported under different growth conditions (16). In contrast, no factors involved in promoter selection, and thus transcription specificity, have so far been identified in archaea.

Our previous results have shown that ammonium-grown cells of Methanosarcina barkeri do not reduce acetylene to ethylene (the standard assay for nitrogenase activity) or show bands cross-reacting with the antibody to eubacterial component 2 in immunoblots (25). Dot blot (6) and Northern blot (7) hybridizations suggest that expression of the nitrogenase structural genes was repressed by ammonium at the level of transcription. Thus, expression of genes involved in nitrogen fixation, an energetically costly process, is highly regulated. In this study, we examined the binding of proteins to the TATA box (TATAAATA) promoter region of nifHDK2 genes found upstream from the transcription start site mapped by primer extension (see Fig. 1). This promoter region contains a 7-bp purine-rich sequence upstream of the TATA box. We also report here the presence of a substance in extracts of ammonium-grown cells (ammonium-grown extracts) that inhibits binding at the TATA box promoter region by the DNA-binding proteins.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains and growth conditions. M. barkeri 227 (ATCC 43241, DSM 1538, and OCM 35) was obtained from our own culture collection. M. barkeri Fusaro (OCM 83, ATCC 29787, and DSM 804) was purchased from the Oregon Collection of Methanogens. Cells were grown under nitrogen-fixing conditions and in the presence of ammonia, as described previously (6).

Preparation of crude extracts and protein samples from M. barkeri 227 for gel mobility shift assays. Extracts of N₂-grown cells (N₂-grown extracts) and ammonium-grown extracts were prepared by loading the cells into a French pressure cell (SLM Aminco, Urbana, Ill.), as described previously (25). The cells were broken at 20,000 lb/in², and the extract was collected in an N₂-flushed vial. Cell debris was removed by centrifugation at 25,000 × g in 50-ml polypropylene tubes with screw caps containing silicone rings (Nalge Co.) loaded inside the anaerobic chamber. Crude extract was stored at 20°C.

Preparation of DNA for gel mobility shift assays. DNA fragments were amplified from M. barkeri chromosomal DNA by PCR with two gene-specific primers. One of these primers was radioactively labeled with [³²P]dATP and T4 polynucleotide kinase (New England Biolabs). PCR fragments were purified by the Promega Wizard PCR purification kit, according to the manufacturer's instructions. Oligonucleotides Oligo 2 (5'-GCTCGCGAACAATGTCA-3') and Oligo nifH2 (TCCAATTCCCACCCTTTTCCG) were used to amplify a 165-bp DNA fragment (Fig. 1), while Oligo 1 (ATACATAGTACAACGGTTACCGGC) and Oligo nifH2 were used to amplify a 103-bp fragment. A DNA fragment (270 bp) containing the promoter region for methyl coenzyme M reductase was prepared by PCR from M. barkeri Fusaro chromosomal DNA by using oligonucleotides Oligo MR1 (5'-TTTCGATCGATACGGTT-3') and Oligo MR2 (5'-TGTCGTCGTAGATGTCT-3'). A nifH2 DNA fragment (165 bp) containing the promoter region was amplified from M. barkeri 227 by oligonucleotides Oligo 2 and Oligo nifH2 (Fig. 1). These DNA fragments were used to compete with the end-labeled probe prepared as described previously.

View larger version (18K): [in this window] [in a new window]   FIG. 1.   The 5' end and upstream region of nifH2 in M. barkeri. *, transcription start site; Box A, archaeal TATA box promoter; Pur Box, upstream purine-rich region; RBS, ribosome binding site. Oligonucleotides used for PCR amplification of the template DNAs are indicated by arrows.

Gel mobility shift assays. Protein-DNA complexes were resolved on low-ionic-strength polyacrylamide gels (12, 13). Protein samples were incubated with approximately 5,000 cpm (0.3 ng) of end-labeled double-stranded DNA fragments in the presence of 0.5 µg of poly(dI-dC) · poly(dI-dC) (Pharmacia) in a final volume of 20 to 40 µl. In early experiments, 0.5 µg of calf thymus DNA (Sigma) was used instead of poly(dI-dC) · poly(dI-dC), but better results were obtained with the latter as nonspecific DNA. Incubations were carried out on ice for 30 to 60 min in a solution of 10 mM Tris-HCl (pH 7.6), 50 mM NaCl, 5% (vol/vol) glycerol, 1 mM EDTA, and 1 mM dithiothreitol. Samples were loaded onto low-ionic-strength 5% (wt/vol) polyacrylamide gels (acrylamide-to-bisacrylamide weight ratio of 37.5:1) which were preelectrophoresed for 1 h at 170 V in 1× TBE buffer consisting of 8.9 mM Tris-HCl (pH 8.0), 8.9 mM boric acid, and 2 mM EDTA. Gels were electrophoresed at 170 V at room temperature until the bromophenol blue had run to the bottom of the gel. They were then laid on top of Whatman 3MM paper, dried, and autoradiographed.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Detection of proteins that bind to the promoter region of nifHDK2. PCR was used to amplify the region upstream from nifH2 (Fig. 1) with oligonucleotides Oligo 2 and Oligo nifH2, producing a fragment 165 bp long. When Oligo 1 was used in place of Oligo 2, a fragment 103 bp long and lacking the promoter region was produced. These DNA fragments were then used as substrates for band shift assays using crude extracts of M. barkeri cells grown with N₂. As shown in Fig. 2A, extracts of N₂-grown cells contained factors binding to the fragment containing the promoter region, thereby causing retardation of a fraction of the DNA, while no shifted bands were detected from the DNA fragment lacking the promoter region.

View larger version (53K): [in this window] [in a new window]   FIG. 2.   Detection, by gel mobility shift assay, of binding to the promoter region of nifH2 by proteins in a crude extract from N2-grown M. barkeri. (A) Requirement of the TATA box region for binding. Probe A (lanes 1 and 2) is a 165-bp fragment amplified by PCR with oligonucleotides Oligo 2 and Oligo nifH2 (Fig. 1) that contains the promoter region. Probe B (lanes 3 and 4) is the 103-bp product of PCR amplification of oligonucleotides Oligo 1 and Oligo nifH2 and lacks the TATA box. Lanes 2 and 4 received 20 µg of crude protein extract. In each assay, 0.5 µg of calf thymus DNA was added, and other conditions are as described in Materials and Methods. (B) Effect of protein concentration on shifting. Probe A was incubated with 0.5 µg of poly(dI-dC) · poly(dI-dC) DNA and with 0, 2, 4, 8, 16, or 32 µg of protein extract per assay (lanes 1 through 6, respectively).

Specificity of proteins binding to the nifH2 promoter. Binding occurred in the presence of nonspecific competitors such as calf thymus DNA (Fig. 2A), poly(dI-dC) · poly(dI-dC) (Fig. 2B), and pBluescript (data not shown). To examine the specificity of DNA binding to the nifH promoter further, we also examined binding to the promoter region for genes encoding the methyl coenzyme M reductase enzyme complex (mcr), which carries out the final step in methanogenesis and is highly expressed and presumably constitutive whether the cells are fixing N₂ or not. Unfortunately, the mcr genes from strain 227 have not been cloned so we used closely related M. barkeri Fusaro, which has been sequenced (2) and for which the upstream sequence TTTAAGTA has been proposed to be the promoter.

View larger version (54K): [in this window] [in a new window]   FIG. 3.   Band shift assays of labeled methyl coenzyme M reductase (mcr) DNA and labeled nifH2 DNA by M. barkeri N2-grown and ammonium-grown extracts. The binding reaction mixtures were as described in the legend to Fig. 2B except that 0.3 ng of labeled nifH2 DNA was used in lanes 1 through 4 and 0.3 ng of labeled mcr DNA was used in lanes 5 through 8. (A) Lane 1, no protein (control); lane 2, 5 µg of N2-grown extract; lane 3, 15 µg of N2-grown extract; lane 4, 40 µg of N2-grown extract; lane 5, no protein (control); lane 6, 5 µg of N2-grown extract; lane 7, 15 µg of N2-grown extract; lane 8, 40 µg of N2-grown extract. (B) Lane 1, no protein (control); lane 2, 40 µg of ammonium-grown extract; lane 3, no protein (control); lane 4, 40 µg of ammonium-grown extract. Lanes 1 and 2 received 0.3 ng of labeled nifH2 DNA, while lanes 3 and 4 received 0.3 ng of labeled mcr extract.

View larger version (74K): [in this window] [in a new window]   FIG. 4.   Competition between the nifH2 promoter region and the mcr promoter region from M. barkeri for protein binding to the nifH2 promoter region. Various amounts of unlabeled DNA fragments were added to a reaction mixture containing 0.3 ng of labeled nifH2 promoter region DNA before the extract was added. All binding reaction mixtures contained 60 µg of crude extract (except lane 1) and 0.5 µg of calf thymus DNA. Reaction mixtures contained labeled nifH2 DNA plus either no protein (NP) and unlabeled DNA (lane 1) or 0 (lane 2), 3 (lane 3), 10 (lane 4), 20 (lane 5), or 35 (lane 6) ng of an unlabeled fragment containing the nifH2 promoter. Alternatively, reaction mixtures contained 0 (lane 7), 3 (lane 8), 10 (lane 9), 20 (lane 10), 35 (lane 11) ng of an unlabeled fragment containing the methyl coenzyme M reductase (mcr) promoter.

Detection of an inhibitor of binding of proteins to the nifH2 promoter region in extracts of ammonium-grown culture. As described previously, up to 40 µg of extracts from ammonium-grown cells showed no band shifting of DNA containing the nifH2 promoter. This lack of binding may be due to any number of reasons, including inactive extract. To investigate this phenomenon further, we examined the effect of ammonium-grown extracts on binding by N₂-grown extracts. Adding as little as 1 µg of protein from extracts from ammonium-grown cells to 20 µg of protein from N₂-grown cells led to complete inhibition of band shifting (Fig. 5). Therefore, there appears to be a substance present in ammonium-grown extracts that inhibits binding of proteins to this DNA but not to mcr DNA. It is unlikely to be competing DNA, since one would not expect inhibition at 1/20 of the concentration used in that case (Fig. 4).

View larger version (50K): [in this window] [in a new window]   FIG. 5.   Inhibition of protein binding to the nifH2 promoter region by extracts from ammonium-grown cells of M. barkeri. Binding reaction mixtures were as described in Materials and Methods. The labeled nifH2 DNA fragment was incubated with 20 µg of N2-grown extract and 0 (lane 1), 1 (lane 2), and 5 (lane 3) µg of ammonium-grown extract.

View larger version (40K): [in this window] [in a new window]   FIG. 6.   Heparin-Sepharose fractionation of extracts from ammonium-grown cells of M. barkeri. (A) Fractionation scheme of ammonium-grown extracts. The band-shifting and inhibitory activities of the fractions are indicated by + or . Binding reaction mixtures were as described in Materials and Methods. (B and C) DNA binding (B) and inhibition of DNA binding (C) by N2 extracts by different fractions of ammonium-grown extracts. Each reaction mixture contained 40 µg of a heparin-Sepharose fraction; in addition, 60 µg of N2 extract was used in panel C. Lanes: 1, 0 M NaCl; 2, 0.25 M NaCl; 3, 0.5 M NaCl; 4, 0.75 M NaCl.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Using band shift mobility assays, we demonstrated the presence of proteins in extracts of N₂-grown M. barkeri that bind specifically to the promoter region of nifH2 (Fig. 1). The specificity of this interaction is supported by several lines of evidence: (i) binding to the DNA was not inhibited by various nonspecific sources of DNA, such as poly(dI-dC) · poly(dI-dC) or calf thymus DNA; (ii) binding to the labeled DNA was inhibited by unlabeled DNA containing this promoter region; (iii) there was no binding to the DNA fragment produced by PCR amplification with oligonucleotides Oligo 1 and Oligo nifH2, which lacked the promoter region; and (iv) binding was not detected with extracts from ammonium-grown cells, while there was binding by these extracts to the mcr promoter region.

The appearance of slower-migrating bands with increasing extract concentrations (Fig. 2) in band shift assays using a DNA fragment containing the promoter region indicates successive binding to the site by more than one protein molecule per DNA molecule. There can be more than one protein binding at each step, and the proteins could be the same or different from each other. Likely candidates for part of the DNA-binding protein complex in M. barkeri are TBP and TFIIB homologs, both of which have been found in archaea (10, 31). The TBP typically binds as a monomer, as it contains two nearly identical domains (21, 27). We recently cloned the gene encoding a TBP homolog from M. barkeri (5) and are presently working to overexpress an active form for use in further experiments.

We detected no proteins binding to the nifH promoter region in ammonium-grown extracts. That these ammonium-grown extracts are able to shift the mcr promoter region demonstrates that this lack of binding was specific and not just an artifact of a low-potency extract. The lack of binding to the nifH promoters by extracts of ammonium-grown cells could be due to the lack of specific binding proteins or the presence of a binding inhibitor. Our evidence supports the latter possibility, since as little as 1 µg of ammonium-grown extract inhibited binding by 20 µg of N₂-grown extract. Moreover, when fractionated on a heparin-Sepharose column, most of the inhibitory activity was obtained in the pass-through fraction while a fraction eluting at higher salt concentrations showed the ability to bind to the nifH promoter region, indicating that the ability to bind that promoter is present in the ammonium-grown cells but is inhibited under those conditions. That several small molecules (ammonia, glutamate, or glutamine) often used as cellular signals of nitrogen sufficiency (19) did not inhibit DNA binding by N₂-grown extracts, and that this inhibitory activity was retained by a 10,000 molecular-weight-cutoff filter, suggests that the inhibitor is a protein, but more characterization is needed.

To determine whether the nif promoter region has binding proteins in common with other promoters, we performed competition assays, using the promoter region from a closely related M. barkeri strain, of the methyl coenzyme M reductase (mcr) genes; these genes encode the central enzyme in methanogenesis, which would presumably be constitutively and highly expressed in the presence of ammonia or under nitrogen-fixing conditions. A DNA fragment harboring this promoter region was readily shifted with extracts from strain 227 (Fig. 3A). Thus, the mcr promoter region readily competed with the nifH promoter for some common part of the transcription initiation machinery (Fig. 4), most likely including the TBP and TFIIB homologs. While the mcr and nifH2 promoter regions compete for DNA-binding proteins, there are clearly differences between them since, in contrast to the nifH2 promoter region, binding to the mcr region occurred in ammonium-grown cells. This phenomenon is difficult to explain as simple binding to a TBP-TFIIB initiation complex and suggests the presence of a protein providing nifH2 specificity, which would potentially interact with the proposed ammonium-growth-specific inhibitor. Such an inhibitor also might bind directly to the TBP or TFIIB.

We had anticipated binding by a repressor in ammonium-grown cells to be the mechanism of ammonia repression, but we failed to find evidence for this (Fig. 4). Indeed, Cohen-Kupec et al. (8) have demonstrated that ammonia repression of nitrogen fixation in the archaeon Methanococcus maripaludis involves repressor binding to a palindromic site associated with the transcription start of the structural nif genes from that organism. We have not found a similar site associated with the nifH2 transcription start, or elsewhere downstream of the promoter area, in M. barkeri. It is not surprising to find that these two methanogens have different mechanisms for regulation of transcription of their nif genes, since the evolutionary distance between them is comparable to that between Klebsiella pneumoniae and Clostridium pasteurianum (39), which also have different mechanisms of regulation (4).

The proposed mechanism of transcription inhibition in ammonium-grown cells does have precedents. For example, the proto-oncogene product p53 (34) in eucaryotes can inhibit TBP function by binding to it, although its binding appears to be at the amino terminus of the TBP, which is not involved in DNA binding (18). Other examples of negative regulators in eucaryotes have also been reported (1, 11, 20, 28). Dr1 binds to TBP and thereby inhibits both basal and activated transcription by preventing interaction of TFIIA and TFIIB with TBP (20); Dr2 interacts with TBP and represses basal transcription (28); and Mot1 (AD1), a global repressor of RNA polymerase II transcription, inhibits TBP binding to DNA by an ATP-dependent mechanism (1). To prove or disprove aspects of this model, we are presently attempting to purify and characterize the DNA-binding proteins and the putative inhibitor protein.