Heat shock inducibility of an archaeal TATA-like promoter is controlled by adjacent sequence elements


Charles J. Daniels. E-mail daniels.7@osu.edu; Tel. (614) 292 4599; Fax (614) 292 8120.


The expression of a heat-inducible cct1 (chaperonin-containing Tcp-1) family member gene is regulated at the transcription level in the archaeon Haloferax volcanii. Transcriptional fusions of the cct1 promoter region with a yeast proline tRNA reporter gene were constructed to analyse the functional domains of this archaeal heat shock promoter. Both basal and heat-induced transcription of the reporter gene was directed by an archaeal consensus TATA element (5′-TTTATA-3′) centred 25 bp upstream of the transcription start site. Deletion mutagenesis indicated that the 5′ boundary of the cct1 regulatory region mapped to position − 37. Nucleotide alignment with the 5′ flanking regions of two additional cct-related genes identified in H. volcanii showed a high degree of sequence conservation between positions +1 and − 37, especially in and immediately surrounding the TATA element of the putative core promoter. Mutational analysis of conserved sequences demonstrated that basal and heat-induced transcription required sequence elements located upstream and downstream of the TATA-box. These findings indicate that the regulatory sequences involved in heat-induced transcription lie within the core promoter region and suggest that the mechanism controlling heat shock gene expression in H. volcanii differs from the bacterial and eukaryal strategies.


Archaea compose a domain that is phylogenetically distinct from Bacteria and Eukarya (Woese et al., 1990). Similarities in the core promoter architecture and basal transcription apparatus, however, suggest that Archaea and Eukarya are related at a fundamental level (reviewed by Langer et al., 1995; Keeling and Doolittle, 1995; Baumann et al., 1995; Thomm, 1996; Reeve et al., 1997). For example, the archaeal consensus promoter, 5′-(T/C) TTA(T/A)A-3′, markedly resembles eukaryal polymerase II promoters in nucleotide sequence, location and function (Reiter et al., 1988; Thomm and Wich, 1988; Hausner et al., 1991; Hain et al., 1992; Palmer and Daniels, 1995; Qureshi et al., 1995a; Danner and Soppa, 1996). Perhaps even more significantly, homologues of the eukaryal TATA-binding protein (TBP) and basal transcription factor TFIIB have been identified recently in a number of Archaea (Ouzounis and Sander, 1992; Creti et al., 1993; Marsh et al., 1994; Rowlands et al., 1994; Qureshi et al., 1995a,b; Bultet al., 1996; J. R. Palmer and C. J. Daniels, unpublished). Although progress has been made in elucidating the basal transcription complex nucleated at an archaeal promoter (Hausner et al., 1996; Kosa et al., 1997; Qureshi et al., 1995a; 1997; Wettach et al., 1995), knowledge of the molecular mechanisms controlling differential gene expression at the transcription level is still lacking. Because of its universal conservation, the heat shock response is an attractive model system for studying transcription regulation in the Archaea.

Bacterial and eukaryal heat shock proteins have been characterized extensively from both structural and regulatory perspectives. In these phylogenetic domains, the preferential overproduction of Hsps is controlled primarily at the level of transcription. For example, in Escherichia coli, in which most of the prokaryotic research on the heat shock response has focused, promoters dictating the expression of heat-inducible genes are recognized by RNA polymerase holoenzymes carrying the σ32 subunit instead of the vegetative σ70 subunit (for a recent review, see Mager and de Kruijff, 1995). In Bacillus subtilis and other Gram-positive organisms, the expression of groESL and dnaK genes is controlled by a negative regulatory scheme. These operons possess an inverted repeat (consensus TTAGCACTC-N9-GAGTGCTAA; Wetzstein et al., 1992), which has been proposed to function as a binding site for a repressor protein (Zuber and Schumann, 1994).

In the Eukarya, the transcriptional activation of heat shock genes is mediated by a trimeric heat shock factor (HSF) that binds to a highly conserved cis-acting element, termed the heat shock element (HSE; reviewed in Sorger, 1991; Fernandes et al., 1994a). The enhancer-like heat shock element, which is functional at variable distances from the TATA element, is composed of a varying number of 5 bp AGAAn modules arranged as contiguous inverted repeats (e.g. nGAAnnTTCn; Pelham, 1982; Perisic et al., 1989; Fernandes et al., 1994b).

Heat shock genes for the third phylogenetic domain, Archaea, have been reported in a number of organisms (reviewed recently by Trent, 1996). These genes include the dnaK and dnaJ homologues from Methanosarcina mazei (Macario et al., 1991; Clarens et al., 1995), the hsp70 homologue from Haloarcula marismortui (Gupta and Singh, 1992), the thermosome from Methanopyrus kandleri (Andra et al., 1996) and thermophilic factor 55 (TF55) α- and β-subunits from Sulfolobus shibatae (Trent et al., 1991; Kagawa et al., 1995). Of the heat shock genes that have been identified in the Archaea, the TF55 α- and β-subunits from S. shibatae (Kagawa et al., 1995) and the dnaK and dnaJ homologues from M. mazei (Clarens et al., 1995) have been shown to be regulated at the level of transcription. We have described previously the characterization of two heat-inducible genes (cct 1 and cct 2) from the archaeon Haloferax volcanii that share high amino acid sequence similarity to members of the chaperonin-60 (cpn60) family, particularly the TF55 α- and β-subunits and the t-complex polypeptide-1 (Tcp-1) from humans (Kuo et al., 1997). Cct mRNA levels were shown to be preferentially induced when H. volcanii cells were exposed to 60°C (Kuo et al., 1997).

The purpose of the work described here was to begin an examination of the molecular mechanisms governing transcription regulation in Archaea by using the heat-inducible cct 1 gene as a model. To achieve this, we used a plasmid-based transcription reporter system to analyse the regulatable cct 1 promoter in vivo. We show that sequences required for heat-induced transcription lie within the core promoter region, which includes the TATA element and adjacent sequences.


In vivo reporter assay for measuring heat-induced transcription

We have described previously the construction of a plasmid-based transcription reporter system for measuring the in vivo activity of H. volcanii promoters (Palmer and Daniels, 1994). The expression module contains a modified yeast proline tRNA reporter gene coupled to a T-rich yeast polymerase III termination element. Fusion of a promoter to the reporter gene on the H. volcaniiE. coli pWL201 shuttle plasmid results in the production of a primary tRNAProM transcript that is easily quantified by Northern analysis (Palmer et al., 1994). To identify sequences required for heat-activated gene expression, we placed transcription of the tRNAProM reporter gene under the direction of the H. volcanii cct1 promoter region (Fig. 1A).

Figure 1.

. Fusion of the H. volcanii cct1 promoter region to the tRNAProM reporter gene. A. Schematic representation of the promoter–reporter module carried on the H. volcaniiE. coli pWL201 shuttle plasmid. The cct1 TATA element and initiation site (arrow) are indicated. The antisense oligonucleotide PROEXI is specific for exon 1 of the tRNAProM RNA. Important restriction endonuclease sites are shown: H, HindIII; X, XbaI; E, EcoRI. B. Primer extension analysis of a promoter–reporter fusion. The expression construct used in transcript analysis contained the cct1 promoter region with 5′ flanking sequences up to position − 397 relative to the transcription start site. Total RNA (10 μg) isolated from pWL-bearing H. volcanii cells grown at 37°C (lane 2) and heat shocked at 60°C for 45 min (lane 1) served as templates for cDNA synthesis. PROEXI oligonucleotide was used for transcript analysis and DNA sequencing. Primer extension products (lanes 1 and 2) were electrophoresed alongside dideoxy-terminated sequencing reaction products (lanes A, G, C and T) on a 6% denaturing polyacrylamide gel. The mapped transcription start site is indicated with an arrow.

Similar to eukaryal RNA polymerase II transcribed genes and archaeal genes in general (Thomm and Hausner, 1993; Zillig et al., 1993; Palmer and Daniels, 1995), the H. volcanii cct1 promoter region contained a TATA element (see Fig. 3A). Preliminary studies indicated that transcription of this gene initiated at a guanine residue located 25 bp downstream from the centre of this element (Kuo, 1997). To ascertain whether the transcription initiation site was altered by coupling the reporter gene to the cct1 promoter, primer extension analysis was performed on cellular RNA isolated from H. volcanii cells harbouring a pWL201 expression plasmid containing cct1 5′ flanking sequences up to position − 397. The tRNAProM exon 1-specific oligonucleotide PROEXI (Fig. 1A) was used as the primer for both cDNA synthesis by reverse transcriptase and DNA sequencing. As shown in 1Fig. 1B, transcription from the fused promoter in the expression module initiated from the same site as in the native gene for both basal and heat-induced expression, indicating that transcription start signals were unaffected by fusing the cct 1 promoter region to the tRNAProM reporter gene. It is interesting to note that this gene lacks a ribosome-binding sequence, a feature common among halophile protein genes (Palmer and Daniels, 1995).

Figure 3.

. Comparison of archaeal cct-related gene promoter regions and mutagenesis of the H. volcanii cct 1 promoter. A. Nucleotide sequence alignment of the promoter regions of three heat-inducible cct genes from H. volcanii and promoter regions of related archaeal genes. Sequences conserved among the H. volcanii cct genes are indicated by the solid circles. A consensus (≥ 67%) based on a comparison of all archaeal cct-related promoter regions is presented below. The genes are: H. volcanii cct 1 (AF010470); H. volcanii cct 2 (AF010469); H. volcanii cct 3 (AF029873); Methanobacterium thermoautotrophicum cct (AE000666, Mt1817); Methanococcus janaschii chaperonin (U67542, MJ0999); Methanopyrus kandleri chaperonin (Z50745); Archaeoglobus fulgidus cct-related genes (http://www.tigr.com); Thermoplasma acidophilum thermosome alpha subunit (Z46649); Thermoplasma acidophilum thermosome beta subunit (Z46650); Thermococcus sp. chaperonin beta subunit (AB001083); Sulfolobus shibatae thermophilic factor 56 (tf56) alpha subunit (L34691); Sulfolobus shibatae thermophilic factor 55 (tf55) beta subunit (X63834); Sulfolobus sp (strain 7) chaperonin alpha subunit (AB001085); Sulfolobus sp (strain 7) chaperonin beta subunit (AB001086). B. Functional analysis of the conserved sequence domains of the H. volcanii cct1 promoter. The wild-type promoter region for cct1 is presented. Nucleotides conserved among all three H. volcanii cct 5′ regulatory regions are underlined. Base mutations and their positions are shown below the wild-type sequence. Basal transcription levels are expressed as a percentage of the wild-type activity (arbitrarily set at 100%), and heat shock induction folds were calculated as described in Experimental procedures.

Mapping the 5′ boundary of the cct1 promoter by deletion mutagenesis

To map the 5′ boundary of the cct 1 regulatory region, polymerase chain reaction (PCR) primers were used to create a series of sequential 5′ deletion mutants, each with a 3′ terminus corresponding to position +10 in the cct1 coding region (Fig. 2A). The DNA template used for PCR-based deletion mutagenesis was plasmid HS5, which contained the cct1 promoter region and the adjacent N-terminal coding region on a 2.3 kb MluI fragment in pUC19 (Fig. 2A). The ability of each 5′ deletion mutant to promote efficient heat-induced transcription in vivo was measured by comparing the level of tRNAProM transcript produced under growth at 37°C (0 min at 60°C) with the level produced under heat shock conditions (45 min at 60°C). A tRNALeu-specific probe was used as an internal control to evaluate RNA recoveries. To ensure that heat shock induction was caused by the fused cct1 promoter region and not by a cryptic promoter element within the pWL201 vector sequence, a construct containing the H. volcanii tRNALys promoter was used as a negative control (Fig. 2A).

Figure 2.

. Deletion analysis of the cct1 5′ flanking region. A. Schematic representations showing the amounts of 5′ flanking sequence fused to the yeast tRNAProM reporter gene in the pWL-based expression plasmid. 5′ Flank deletion mutants were generated by PCR amplification using HS5 plasmid DNA as the template. Induction folds were calculated by taking the ratio of normalized heat shock tRNAProM-derived signal (45 min at 60°C) to normalized non-shock tRNAProM-derived signal (set at 1). Positions of the cct1 TATA-like promoter element and initiation site (arrow) are indicated. The tRNALys promoter region coupled to the reporter gene served as a negative control for heat shock induction. H, HindIII; M, MluI; Ss, Sst I. B. Core promoter regions of 5′ deletion constructions and Northern blot hybridizations. Wild-type cct 1 sequences are represented by upper case letters, while pWL vector sequences (5′Δ-36) or cct 1 sequences in the reverse orientation (5′Δ-37) are represented by lower case letters. Those nucleotide sequences in common with 5′Δ-50 wild-type reference are indicated (closed circles). Northern hybridizations show the 5′ boundary for efficient heat-induced transcription. Hybridization signals corresponding to the reporter tRNAProM transcript and the internal control, tRNALeu transcript, are indicated.

2Figure 2A shows that promoter constructs with 5′ deletions to position − 397, − 233, −125, − 80 and − 50 exhibited efficient heat-induced transcription. When 5′ flanking sequences were deleted to position − 36, however, the heat shock inducibility of the fused promoter decreased to twofold. This result suggested that the 5′ boundary of the functional cct1 promoter region mapped to a site between positions − 50 and − 36 relative to the transcription start. To evaluate the role of sequences between positions − 50 and − 36, we examined tRNAProM transcripts from cells carrying a construct (5′Δ-37) in which the HindIII–HindIII fragment of the cct 1 promoter fusion 5′Δ-233 was cloned into the expression vector pWL201 in the reverse orientation. As the HindIII site lies immediately upstream of the TATA region, the 5′Δ-37 construct retained the wild-type sequence from positions − 37 to +10, and showed no significant similarity to the wild-type sequence between positions − 37 and − 50 (Fig. 2B). This construct directs heat shock transcription at a level similar to the 5′Δ-50 deletion mutant, indicating that the 5′ boundary of the functional cct1 promoter extends only to position − 37.

Sequence conservation among three H. volcanii cct promoter regions

To define further the cct1 promoter, we compared the 5′ flanking regions of two other cct family member genes discovered in H. volcanii (Kuo et al., 1997; this study). Like the initially characterized cct1 gene, the monocistronic cct 2 and cct 3 genes encode an approximately 1.7 kb RNA that is preferentially induced during heat shock at 60°C (results not shown). The nucleotide sequence alignment of the three cct promoter regions is presented in 3Fig. 3A. Most notable is the high degree of sequence conservation in and immediately flanking the TATA element. All three cct genes possess a perfectly conserved 9 bp TATA element, 5′-TTTATAGAA-3′, centred 24 ± 2 nucleotides upstream from the transcription start site (Fig. 3A). The initiation sites for all three H. volcanii cct genes mapped to the guanine residue in the conserved TG dinucleotide initiation element under normal growth conditions and when H. volcanii cells were challenged with heat stress (Kuo et al., 1997; D. K. Thompson and C. J. Daniels, unpublished).

The cct 5′ flanking regions also possess a conserved tetranucleotide sequence element (5′-CGAA-3′) beginning at position − 37 (cct 1 and cct 2) or − 39 (cct 3), and located precisely 5 bp upstream from the TATA element. Sequences 5′ of the CGAA motif noticeably diverge, again suggesting that position − 37, or position − 39 in the case of cct3, constitutes the 5′ limit of the functional cct promoter. Each promoter also has a less conserved pentanucleotide sequence block (5′-CAAaC-3′) positioned 6 bp downstream from the TATA element. Finally, we did not detect the consensus sequence for the eukaryal heat shock element (nGAAnnTTCn) in the 5′ sequences flanking the cct coding regions.

Other sequence elements with a potential regulatory role are conserved between cct1 and cct2 but absent in cct3. Two direct repeat motifs — one with the half-site sequence 5′-AATCA-3′ and the other with the sequence 5′-CAA-3′— were located between the TATA and initiator core promoter units. In contrast, the cct3 promoter region contains a different direct repeat of two 5′-GTTGA-3′ half-sites separated by 5 nucleotides.

Mutational analysis of conserved promoter domains

Based on the comparative sequence analysis, H. volcanii cct promoters appear to possess three separate functional domains: a sequence element 5′ proximal to the TATA element, an AT-rich core promoter element with a central TATA motif and a sequence element 3′ proximal to the TATA box that does not include the transcription initiation site. To establish whether these conserved domains do indeed play a role in the transcriptional control of cct1 gene expression, PCR-based mutagenesis was used to alter targeted nucleotides, and then the in vivo transcriptional activities of the mutant promoters were measured using the tRNAProM reporter gene assay. As the critical role of TATA-like sequences for transcription efficiency has been well established both in vitro (Reiter et al., 1990; Hausner et al., 1991; Hain et al., 1992) and in vivo (Palmer and Daniels, 1995; Danner and Soppa, 1996), these sequences were not examined, and emphasis was placed on the conserved sequences located 5′ and 3′ to the TATA element. The basal transcription activity of each promoter mutant was compared with the expression of the 5′Δ-50 construct.

As predicted, the 5′-CGAA-3′ motif, located precisely 5 nucleotides upstream from the TATA promoter element, was essential for efficient basal and induced transcription of the tRNAProM reporter gene. When the wild-type tetranucleotide sequence was replaced with the arbitrary sequence 5′-GTCG-3′, the relative basal transcription activity of the mutant cct1 promoter region was reduced to 9% of the wild-type activity (Fig. 3B). Moreover, the ability of this mutant promoter to stimulate transcription in response to heat was completely abolished. The same results were observed when the 5′-CGAA-3′ element was shifted 5 nucleotides upstream by a 5 bp insertion. Basal transcription for this construct decreased to 4% of the wild-type activity, and heat shock induction was eliminated (data not shown), indicating that the motif is conserved in terms of position as well.

We also investigated the functional significance of each nucleotide in the 5′-CGAA-3′ motif. Base substitutions at each position had deleterious effects on heat shock induction, with the most severe reduction occurring when nucleotide G at position − 36 was altered (Fig. 3B). Interestingly, the A→C transversion at position − 35 had essentially no effect on basal transcription but significantly decreased heat-induced transcription, while other changes affected both basal and heat-induced transcription (Fig. 3B). Disruption of the six-nucleotide thymine tract by the substitution of a C residue at position − 28 appeared to have a more serious effect on basal transcription than on activated transcription, with the basal level reduced to 39% of the wild-type promoter activity (Fig. 3B). Interestingly, mutagenesis of the G (− 22), GA (− 22, − 21) and GAA (− 22 to − 20) residues to their nucleotide complements led to a significant decrease in heat-induced transcription (Fig. 3B). In addition, when the CA dinucleotide at position −16 and −15 in the conserved 5′-CAAaC-3′ motif was changed to TG, the heat shock inducibility of this mutant promoter was reduced to less than twofold, while basal transcription remained essentially at the wild-type level. Other mutations had a more or less neutral effect on transcription.


Using an in vivo transcription reporter system, we have confirmed an earlier observation that the cct1 (chaperonin-containing Tcp-1) gene from H. volcanii is regulated positively at the level of transcription initiation when cells are exposed to heat shock at 60°C. The promoter region of this gene contains a consensus TATA element. Transcription of the tRNAProM reporter gene initiates at the same site under normal growth and heat stress, indicating that the same TATA element is used under both conditions (Fig. 1B). Deletion mapping demonstrated that the 5′ boundary of the functional cct1 promoter region lies between − 36 and − 50 relative to the transcription start site, and two additional observations strongly suggest that position − 37, in fact, constitutes the 5′ limit of the inducible cct1 promoter region. First, comparative analysis of sequences upstream of three cct-related coding regions revealed a striking degree of conservation in and immediately surrounding an archaeal consensus TATA element (5′-TTTATA-3′), whereas sequences 5′ of position − 37 were found to diverge significantly (Fig. 3A). Secondly, sequences upstream of position − 37 could be completely altered without adversely affecting the heat shock inducibility of the cct1 promoter as long as the conserved TATA-proximal 5′-CGAA-3′ motif remained intact (Fig. 3B). We conclude, therefore, that the regulatory sequences controlling heat-induced transcription are in close proximity to the TATA element (Fig. 4).

Figure 4.

. Regulatory elements in the H. volcanii cct1 promoter region. Sequence elements responsible for heat-inducible transcription lie within the basal promoter region (solid lines) and are conserved in H. volcanii cct promoters (closed circles). These sequences are also within the DNA regions bound by TATA-binding protein and TFIIB.

We also examined the effects of point mutations and multiple nucleotide replacements on the heat shock responsiveness of the H. volcanii cct 1 promoter. Two general classes of mutants were observed. The first class exhibited significant decreases in both basal and heat-induced transcription. These included changes at positions − 34 to − 37, at position − 22 and, to a lesser extent, at positions −19 and −18 (Fig. 3B). This confirmed earlier observations that sequences surrounding the TATA element are essential for basal transcription in the Archaea (Reiter et al., 1990; Palmer and Daniels, 1995). The second class of mutants exhibited normal or slightly higher than normal basal transcription but failed to give the expected increase in transcription when cells were challenged with heat shock. These included changes at position − 35 and in the sequence regions − 22 to − 20 and −16 to −15 (Fig. 3B). In each case, the decrease in heat-induced transcription was significant but did not totally abolish induction. The pattern of sensitivity indicated that the sequence regions needed for heat-induced transcription were located immediately adjacent to the TATA element, and these sequences were in some instances coincident with those regions needed for basal transcription (Fig. 4). This latter case is illustrated in the sequence region of − 37 to − 34 (5′-CGAA-3′; Fig. 3B), where changes at positions − 37, − 36 and − 34 affected both basal and heat-induced transcription. In contrast, an A to C change at position − 35 specifically affected heat-induced transcription (Fig. 3B). Whether the three sequence elements shown to be involved in the heat shock response act in concert or independently remains to be determined.

A comparison of the halobacterial cct promoter regions with those of related archaeal heat shock proteins indicates a high degree of conservation of a 17 bp region containing the TATA element and the upstream 5′-GAA-3′ sequence (Fig. 3A). The upstream 5′-GAA-3′ sequence includes the A residue found to be needed for heat-induced transcription in H. volcanii. The 5′-GAA-3′ sequence does not appear to be unique to heat shock promoters, as a short purine-rich sequence located 10 bp upstream of the TATA element is found in many archaeal promoters (Hain et al., 1992; Palmer and Daniels, 1995). The second sequence element, 5′-CAA-3′, was not conserved among archaeal heat shock promoters, suggesting that its function may not be conserved. The overall pattern of conservation suggests that residues important for the regulation of non-halophile heat shock promoters will also lie within those regions needed for basal transcription. The results of this study and the absence of proteins related to the eukaryal heat shock transcription factor in the archaeal genome sequence databases suggest that the mechanism of archaeal heat shock gene regulation differs from that in eukaryal cells.

The recent solution of the crystal structure of the Pyrococcus woesei TATA DNA–TATA-binding protein–TFIIB complex (Kosa et al., 1997) has provided the first detailed structural view of an archaeal preinitiation complex. Unexpectedly, it was found that the TATA-binding protein was bound to the TATA–DNA sequence in the opposite polarity to that observed for the related eukaryal TATA-binding protein (Nikolov et al., 1995). Despite the difference in polarity of binding, the highly symmetrical archaeal TATA-binding protein retained most of the protein–DNA contacts found in the eukaryal complex. The binding of TFIIB to the carboxyl-terminal repeat of the TATA-binding protein was also retained. This causes the archaeal TFIIB to be positioned with its amino-terminus pointing towards sequences upstream of TATA instead of towards the transcription start site as is the case for eukaryal TFIIB. Although the truncated carboxy-terminal form of the archaeal TFIIB interacts with the phosphodiester backbone on either side of the TATA element, the precise disposition of the amino-terminus has not been determined. In vitro DNaseI footprinting studies with the Pyrococcus TBP and TFIIB proteins indicate that the addition of TFIIB to a TBP–TATA–DNA complex extends the footprint from − 20 and − 34 to −19 and − 42 (Hausner et al., 1996). This is in contrast to the eukaryal complex, in which TFIIB addition to a TBP–TATA–DNA complex extends the footprint towards the initiation site (Lee and Hahn, 1995). It is interesting to speculate that the archaeal TFIIB protein may interact with sequences located upstream of the TATA element. This could explain the essential role for sequences located 5′ of the TATA element in basal transcription. Interaction of specific regulatory factors in this region, either directly with the DNA or through the TATA-binding protein–TFIIB complex, may influence both basal and regulated transcription. Related to this, we have recently found that H. volcanii possesses multiple TATA-binding proteins and TFIIBs (J. R. Palmer, D. K. Thompson and C. J. Daniels, unpublished). One of these TFIIB-encoding genes is induced in response to heat shock, suggesting that, in the halophiles, specific pairings of the basal transcription factors may play a role in controlling heat-induced transcription.

Experimental procedures

Bacterial strains, plasmids and culture conditions

E. coli ED8767 harbouring heat-responsive cosmid clones of an H. volcanii DS2 genomic library (Charlebois et al., 1991) were provided by R. L. Charlebois and grown in Luria–Bertani (LB) medium (Sambrook et al., 1989) supplemented with kanamycin (30 μg ml−1). E. coli strains DH5α and JM110 were routinely cultured in LB or circlegrow medium (Bio101) with ampicillin (100 μg ml−1) to select for pUC19 plasmids or pWL-based plasmids (Lam and Doolittle, 1989). Construction of the yeast tRNAProM reporter module in the H. volcaniiE. coli shuttle vector pWL201 has been described previously (Palmer and Daniels, 1994). H. volcanii WFD11 (Charlebois et al., 1987) was grown aerobically in complex medium (Nieuwlandt and Daniels, 1990) at either 42°C (solid media) or 37°C (liquid media) and supplemented with 20 μM mevinolin (a gift from Merck and Co.) when cells carried pWL-based expression plasmids. Plasmid HS5 (obtained from K.-P. Kuo) is a pUC19 clone containing the cct1 5′ flanking region and the adjacent N-terminal coding region on a 2.3 kb MluI fragment (Fig. 2A). Plasmid pUC19.1, which contains a unique MluI restriction site in the multiple cloning region, was provided by J. R. Palmer. H. volcanii transformations were carried out as described previously (Palmer and Daniels, 1995).

DNA manipulation and sequencing

PCR amplification was used to generate deletion and site-specific mutants of the cct1 5′ flanking region using plasmids HS5 and 5′Δ-50 (this study) as DNA templates. Amplification reactions were carried out using 2 U of Vent DNA polymerase (New England Biolabs), 100–200 ng of template DNA, 1 μM each primer, 200 μM dNTP and 10 μl of 10 ×  Vent buffer (New England Biolabs) in a total volume of 100 μl. PCR reactions were programmed for 30 sequential cycles of denaturation (94°C for 1 min), primer annealing (50°C for 1 min) and extension (72°C for 1–5 min depending on the size of the target DNA) followed by a final extension reaction (72°C for 5 min).

For the deletion mutagenesis, forward PCR primers were complementary to upstream sequences that had 5′ termini corresponding to positions − 397, − 233, −125, − 80 and − 50 relative to the mapped cct1 transcription start. The reverse primer was complementary to a DNA region that included the transcription initiation site and the first two amino acid codons (Met-Ser) of the cct1 ORF. Mutagenic primers carrying the desired nucleotide change were used to create site-specific mutations. The sequences of the primers used in PCR-based mutagenesis are shown in Table 1. For cloning purposes, PCR primers were designed to append synthetic HindIII and XbaI endonuclease restriction sites to the 5′ and 3′ termini, respectively, of the amplified DNAs.

Table 1. . Primers used in PCR-based mutagenesis. Underlined nucleotides indicate base mutations.Thumbnail image of

Promoter regions were cloned into the HindIII and/or XbaI sites of the pWL-based expression module (Fig. 1A) according to standard protocols (Sambrook et al., 1989). Deletion mutant 5′Δ-36 was constructed by cleaving the internal HindIII site located immediately 5′ of the cct1 TATA element and then ligating the 40 bp HindIII–XbaI promoter region into the equivalent sites of the pWL-based expression plasmid using T4 DNA ligase (Gibco BRL). Additional upstream sequences were ligated into the 5′Δ-36 construct as HindIII–HindIII fragments. The cct 1 construct 5′Δ-37, which contained wild-type sequences from positions − 37 to +10, was created by cloning the HindIII–HindIII fragment of deletion mutant 5′Δ-233 in the reverse orientation in the pWL201 expression vector. Promoter constructs were identified by colony blot hybridizations and sequenced by the dideoxy chain termination method (Sanger et al., 1977) using Sequenase T7 polymerase (USB) and the 7-deaza-dGTP sequencing kit (USB) for verification.

In a preliminary study, Trieselmann and Charlebois (1992) identified seven heat-responsive loci (A199, H11, 10D2, 268, 452, 456 and 531) on the H. volcanii chromosome by hybridizing cDNAs derived from heat shock RNAs to a minimal set of overlapping cosmid clones (Charlebois et al., 1991) that covered 96% of the 4.1 Mbp H. volcanii genome. We obtained a complete set of the H. volcanii cosmid library from R. L. Charlebois. To identify other H. volcanii heat shock promoters, heat-responsive cosmid clones were digested with MluI and separated by 0.8% agarose gel electrophoresis. Restricted nucleic acids were transferred by alkaline (0.4 M NaOH) capillary action to a Zeta-Probe nylon membrane (Bio-Rad) as specified by the manufacturer. Approximately 200 ng of a pWL clone containing the complete cct1 gene was radiolabelled with [α-32P]-dATP (20 μCi μl−1, 3000 Ci mmol−1) using the High Prime DNA labelling kit (Boehringer Mannheim), and this was used to probe the Southern blot. In agreement with Trieselmann and Charlebois (1992), we obtained two strong hybridization signals that corresponded to a 3.2 kb fragment from cosmid 268 and a 3.6 kb fragment from cosmid 452 (data not shown). For further analysis, the 3.2 kb and 3.6 kb MluI DNA fragments were cloned into the equivalent site of a modified pUC19 plasmid, which contained an MluI site inserted between the BamHI and XbaI sites in the multiple cloning region of pUC19 (J. R. Palmer and C. J. Daniels, unpublished). The cct gene regions were localized by first sequencing the fragment ends using M13 universal primers and then searching the GenBank protein database for homology to Sulfolobus shibatae TF55 and eukaryal Tcp-1 amino acid sequences. Once the cct ORFs were located, priority was given to identifying the promoter regions of these genes. The heat-responsive gene on cosmid 268 was later sequenced to completion (Kuo et al., 1997).

Induction of heat shock response and isolation of RNA

H. volcanii cells containing appropriate pWL-based expression plasmids were harvested for heat shock when the cell density reached mid-log phase (typically, A550 = 0.5–0.7). Cultures were transferred from 37°C to a 60°C shaking water bath to induce heat shock, and aliquots of 1.5 ml were removed before stress challenge (0 min control) and 45 min after incubation at 60°C. Total RNA was isolated immediately using the TRIzol reagent (Gibco BRL) as specified by the manufacturer, with the exception that centrifugations were performed at room temperature. RNA pellets were stored in 75% ethanol at −70°C until ready for analysis.

Northern blot hybridization and transcript quantitation

The ability of each promoter–reporter fusion to affect heat-induced transcription was measured by Northern blot analysis. RNA pellets stored in 75% ethanol were centrifuged (7500 × g for 5 min at room temperature), dried briefly under vacuum and dissolved in 20 μl of RNA loading buffer [7 M urea, 10% (w/v) glycerol, 0.05% xylene cyanol and 0.05% bromophenol blue]. RNA species were resolved by electrophoresis through a 6% denaturing (8.3 M urea) polyacrylamide gel and transferred to Zeta-Probe nylon membrane (Bio-Rad) using an electrophoretic blotter (Idea Scientific). RNA was fixed to the membrane by baking under vacuum (80°C for 30 min) or by ultraviolet irradiation (120 mJ cm− 2) using the GS Gene Linker (Bio-Rad). Exon 1-specific oligonucleotide PROEXI (5′-CCCAAAGCGAGAATCATACCAC-3′) was used to detect tRNAProM transcript levels. The oligonucleotide LEU3E (5′-GGGGACGAGATTCGAACTCGCGAACCCCTACG-3′), which is specific for chromosomally encoded leucine tRNA (UUA), was used as an internal control. Each oligonucleotide (5–10 pmol) was end-labelled at its 5′ terminus using [γ-32P]-ATP (160 μCi μl−1, 7000 Ci mmol−1) and 5 U of T4 polynucleotide kinase (Gibco BRL). The 32P end-labelled oligonucleotides were added to hybridization solution [0.25 M Na2HPO4 (pH 7.2) and 7% SDS] and allowed to anneal to their complementary RNA sequences for approximately 15–20 h at 50°C. Membranes were washed twice in 2 × SSC (1 × SSC contains 150 mM NaCl and 15 mM trisodium citrate) and 0.5% SDS for 15 min each at room temperature.

Hybridization signals were quantitated using the Packard InstantImager 2024. The values for the PROEXI-derived signals were normalized by calculating the ratio of total counts for tRNAProM over the total counts for tRNALeu. Heat shock induction folds were determined by taking the ratio of normalized heat shock tRNAProM signal to normalized non-shock tRNAProM signal (arbitrarily set at 1). For the site-specific mutagenesis study, the basal transcription level exhibited by the promoter fusion 5′Δ-50 served as the wild-type reference. Experimental values for basal transcription were expressed as a percentage of mean activity relative to the wild-type activity (arbitrarily set at 100). In eight independent analyses of the 5′Δ-50 construct, the basal level of tRNAProM hybridization relative to that of the tRNALeu internal control was observed to vary by as much as 53%. The observed variation for heat shock induction was 33%. Therefore, only those cct1 promoter mutations that gave basal transcription levels 53% above or below the wild-type promoter level and heat shock induction folds below five or above nine were considered significant. Each value presented is the average of at least two independent determinations showing a variation of ≤ 33%.

Primer extension analysis

Total RNA (10 μg) isolated from the appropriate pWL-bearing (5′Δ-233) H. volcanii cells under non-shock (37°C) and heat shock (60°C for 45 min) conditions served as the template for cDNA synthesis. The tRNAProM exon 1-specific oligonucleotide PROEXI was used as the primer for both cDNA synthesis by Superscript II RNase H reverse transcriptase (Gibco BRL) and DNA sequencing. PROEXI was allowed to anneal to its complementary RNA sequence in 1 × hybridization buffer [0.3 M NaCl, 10 mM Tris-HCl (pH 7.5), 2 mM EDTA (pH 8.0)] for 1 h at 50°C. The extension reaction was catalysed by adding 200 U of Superscript II RNase H reverse transcriptase to a reaction mixture containing 1 × First Strand buffer (Gibco BRL), 10 mM DTT, 1 mM dNTP and the PROEXI primer–RNA hybrid template. Dideoxy-terminated DNA sequencing reaction products and primer extension reaction products were separated by electrophoresis on a 6% denaturing (8.3 M urea) polyacrylamide gel.


We thank Rob Charlebois for the H. volcanii cosmids. This work was supported by a grant from the Department of Energy, DE-FG02-91ER20041, to C.J.D. C.J.D. is an associate of the Canadian Institute for Advanced Research.