The Structure and the Characteristic DNA Binding Property of the C-terminal Domain of the RNA Polymerase α Subunit fromThermus thermophilus *

The C-terminal domain of the α subunit of the RNA polymerase (αCTD) from Escherichia coli(Ec) regulates transcription by interacting with many kinds of proteins and promoter upstream (UP) elements consisting of AT-rich sequences. However, it is unclear how this system is common in all eubacteria. We investigate the structure and properties of αCTD from an extremely thermophilic eubacterium, Thermus thermophilus(Tt). The solution structure of Tt αCTD (85 amino acids) was determined by NMR, and the interaction betweenTt αCTD and DNA with different sequences was investigated by means of chemical shift perturbation experiments. The tertiary structure of Tt αCTD is almost identical with that ofEc αCTD despite 32% sequence homology. However,Tt αCTD interacts with the upstream region sequence of the promoter in the Tt 16 S ribosomal protein operon rather than the Ec UP element DNA. The upstream region sequence ofTt is composed of 25 base pairs with 40% AT, unlike theEc UP element with 80% AT. The DNA binding site inTt αCTD is located on the surface composed of helix 4 and the loop preceding helix 4. The electric charges on this surface are not remarkably localized like those of Ec αCTD.

The C-terminal domain of the ␣ subunit of the RNA polymerase (␣CTD) from Escherichia coli (Ec) regulates transcription by interacting with many kinds of proteins and promoter upstream (UP) elements consisting of AT-rich sequences. However, it is unclear how this system is common in all eubacteria. We investigate the structure and properties of ␣CTD from an extremely thermophilic eubacterium, Thermus thermophilus (Tt). The solution structure of Tt ␣CTD (85 amino acids) was determined by NMR, and the interaction between Tt ␣CTD and DNA with different sequences was investigated by means of chemical shift perturbation experiments. The tertiary structure of Tt ␣CTD is almost identical with that of Ec ␣CTD despite 32% sequence homology. However, Tt ␣CTD interacts with the upstream region sequence of the promoter in the Tt 16 S ribosomal protein operon rather than the Ec UP element DNA. The upstream region sequence of Tt is composed of 25 base pairs with 40% AT, unlike the Ec UP element with 80% AT. The DNA binding site in Tt ␣CTD is located on the surface composed of helix 4 and the loop preceding helix 4. The electric charges on this surface are not remarkably localized like those of Ec ␣CTD.
A eubacterial RNA polymerase holoenzyme is composed of five subunits, ␣ 2 ␤␤Ј. As seen on investigation of Escherichia coli (Ec), 1 core enzyme ␣ 2 ␤␤Ј is fully active in the polymerization of RNA. The ␣ subunit of Ec is composed of two structural domains, the N-terminal domain and the C-terminal domain (␣CTD). The Ec N-terminal domain of the ␣ subunit of the RNA polymerase plays a key role in RNA polymerase assembly (1), and Ec ␣CTD is necessary for transcription regulation. Ec ␣CTD interacts with transcription activators, for example, cyclic AMP receptor protein (CRP) (1,2), OmpR (3), and a transcription repressor, GalR (4). Furthermore, Ec ␣CTD recognizes promoter upstream (UP) elements that consist of ATrich sequences and enhances transcription initiation (5). The solution structure of Ec ␣CTD was determined by NMR measurement (6), and crucial amino acid residues for interaction with factors were determined (6,7).
However, it is unclear how these transcription regulation systems are used in all eubacteria. For example, it is unlikely that such AT-rich sequences as the Ec UP element exist in the genome of an extremely thermophilic eubacterium, Thermus thermophilus (Tt), because the average A ϩ T content of the genome is 30% (8). We thus investigated whether Tt ␣CTD binds the Ec UP element sequence or the upstream region sequence of the promoter in the Tt 16 S ribosomal protein operon (TUP) by NMR measurements (9). For this purpose, we constructed an overexpression system of Tt ␣CTD in Ec cells. Using the recombinant protein, the higher thermal stability of Tt ␣CTD was shown by circular dichroic (CD) measurement, and the solution structure of Tt ␣CTD was determined by NMR. The interaction of Tt ␣CTD with a variety of DNAs at 37 and 50°C was investigated by the chemical shift perturbation of NMR signals.

EXPERIMENTAL PROCEDURES
Preparation of the Expression Vector of Tt ␣CTD-Two synthesized DNA oligomers (142 base pairs each) were hybridized and used for the elongation reaction at 4°C for 3 h with the Klenow fragment (Takara Shuzo, Kyoto, Japan). The produced DNA fragment encoded the amino acid sequence of Tt ␣CTD (85 amino acids including the starting Met). However, codons in this fragment were replaced with those frequently used for the gene expression in Ec cells (The Wisconsin Package, Genetics Computer Group, Madison, WI). The fragments were amplified by polymerase chain reaction and then digested with the restriction endonucleases, NcoI and BamHI (Takara). The digested DNA fragments (CTZ) were purified by agarose electrophoresis and then inserted to the pET-15b vector (Novagen). The nucleic acid sequence of CTZ in the vector (258 base pairs) was confirmed with a DNA autosequencer (Applied Biosystems) and given GenBank™ accession number AB036061.
Expression and Purification of Tt ␣CTD-Ec BL21 (DE3) cells were transformed with the CTZ-inserted vector (pET-CTZ) and then grown in LB medium. Isopropyl-1-thio-␤-D-galactopyranoside was added to a liquid culture to 1 mM at A 620 ϭ 0.8. After 3 h of induction the cells were harvested and stored at Ϫ80°C until use.
The following procedure, except for the heat treatment, was carried out at 4°C. Wet cells from 200 ml of culture were suspended in 50 ml of lysis buffer (50 mM Tris-HCl (pH 8.5) and 500 mM NaCl) and then sonicated on ice. The crude extract was then incubated at 60°C for 30 min and centrifuged. The supernatant was recovered and dialyzed against 50 mM Tris buffer (pH 8.5). It was then applied to a HiTrap Q column (2-ml bed volume) (Amersham Pharmacia Biotech) and eluted with a linear gradient of sodium chloride (0 -500 mM NaCl with 50 mM Tris (pH 8.5)). The fractions of Tt ␣CTD detected by SDS polyacrylamide gel electrophoresis were collected. They were applied to a Superdex 75 gel-filtration column (2.6 ϫ 60 cm) (Amersham Pharmacia Biotech) with the gel-filtration buffer (50 mM K 2 HPO 4 /KH 2 PO 4 (pH * This work was supported by Grant-in-aid 09480176 from the Ministry of Education, Science, Sports and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AB036061.
The  1 The abbreviations used are: Ec, Escherichia coli; ␣CTD, C-terminal domain of the ␣ subunit of the RNA polymerase; BAF, barrier to autointegration factor; TUP, the upstream region sequence of the promoter in the 16 S ribosomal protein operon of Thermus thermophilus; Tt, Thermus thermophilus; UP, promoter upstream; rsmd, root mean square deviation. 6.8)). Through these expression and purification procedures, 2.5 mg of Tt ␣CTD was obtained from 200 ml of LB medium culture. The purity of the protein solution was checked by SDS polyacrylamide gel electrophoresis, and a single band was shown. The fractions of Tt ␣CTD were dialyzed against deionized water and stored at 4°C after lyophilization.
UV Spectroscopy-The molar extinction coefficient, E M , of Tt ␣CTD determined by the reported method (10) was 2.5 ϫ 10 5 M Ϫ1 ⅐cm Ϫ1 at 205 nm and pH 6.0. The concentration of Tt ␣CTD was determined spectrometrically using this E M value. CD spectra of Tt and Ec ␣CTD, at 23 M protein concentration in a 0.1-cm cell, were measured with a Jasco J-720 spectropolarimeter. The buffer solution comprised 20 mM K 2 HPO 4 /KH 2 PO 4 (pH 6.0), 30 mM KCl, and 1 mM dithiothreitol. Thermal stability was monitored by measuring the temperature dependence of the molar ellipticity at 222 nm with a heating rate of 1.0°C/min. NMR Spectroscopy of Tt ␣CTD-Uniformly 15 N/ 13 C-labeled Tt ␣CTD was expressed in Ec BL21 (DE3) cells in M9 medium with [ 15 N]NH 4 Cl (0.5 g/liter) and [ 13 C 6 ]D-glucose (2.0 g/l) as the nitrogen and carbon sources. Fractionally 13 C-labeled Tt ␣CTD was expressed in M9 medium containing 10% [ 13 C 6 ]D-glucose and 90% unenriched D-glucose. The expression and purification procedures were as described above. The samples for NMR measurements comprised 1-2 mM protein in 90% H 2 O/10% D 2 O, or in 100% D 2 O, containing 20 mM phosphate buffer (pH 6.0) and 30 mM KCl.
NMR measurements were performed with Bruker DMX500 and DMX600 spectrometers equipped with pulsed-field gradient units and triple resonance probes. All measurements, except for in the hydrogendeuterium exchange experiment (described below), were carried out at 37°C. NMR data were processed using NMRpipe/PIPP software (11,12).
The resonance assignments were obtained through the following three-dimensional experiments: HCAN, HNCA, HN(CO)CA, HNCACB, and CBCA(CO)NH for the backbone nuclei, and 15 N-edited total correlation spectroscopy and HCCH total correlation spectroscopy for the side chain nuclei (13,14). The methyl groups of leucine and valine residues were assigned stereospecifically by means of the 1 H-13 C constant time HSQC spectrum of the fractionally 13 C-labeled protein (15). Aromatic ring proton resonances were assigned by two-dimensional nuclear Overhauser enhancement spectroscopy (200-ms mixing period). Nuclear Overhauser enhancements were collected through three-dimensional 15 N-edited nuclear Overhauser enhancement spectroscopy and three-dimensional 13 C-edited nuclear Overhauser enhancement spectroscopy experiments (200-ms mixing period). An HMQC-J experiment was performed to determine the three-bond HN-H␣ coupling constants (16).
The structure calculations were performed with XPLOR 3.1 (17), and the simulated annealing was started from an extended polypeptide conformation. After 100 structure calculations, 20 structures that had no distance violations of larger than 0.3 Å, no dihedral angle violations of larger than 5°, and a low target energy were obtained. An energyminimized structure was obtained from the averaged structure by minimization of the target function. The structural statistics are summarized in Table I. NMR Titration for the Tt ␣CTD-DNA Interaction-Five kinds of DNA fragments were prepared to examine their affinities as to Tt ␣CTD (Fig.  1). The sequence (25-mer) of fragment UP was that of the UP element from Ec (5). The sequence (25-mer) of fragment TUP is located directly upstream of the promoter of 16 S rRNA from Tt (9). Fragment SUB was the control where the AT-rich sequence of the UP element was substituted by the average sequence of upstream promoters and was often used as the control for the interaction of the 25-mer UP elements (5). The sequence (14-mer) of fragment Ϫ46 has the smaller unit (AAATTT) to bind Ec ␣CTD extracted from the Ec UP element 2 . Fragment NON (14-mer) is a control of fragment Ϫ46 of the same length without any AT nucleotides. Each of the DNA fragments was added to a solution of uniformly 15 N-labeled Tt ␣CTD (0.2 mM final protein concentration, and buffer conditions were the same as above). Perturbed signals in 1 H-15 N HSQC spectra at 37 and 50°C were observed for each Tt ␣CTD and DNA mixture.

RESULTS
CD Measurements-The UV CD patterns around 220 nm of Tt ␣CTD and Ec ␣CTD at 37°C ( Fig. 2A) show that Tt ␣CTD mainly consists of ␣-helical structures and that the ␣-helix content is roughly identical with that of Ec ␣CTD. The thermal stability of Tt ␣CTD was characterized by a change in ellipticity at 222 nm from 20 to 90°C (Fig. 2B). The [] 222 values at 90°C of Tt ␣CTD and Ec ␣CTD were around Ϫ0.8 ϫ 10 4 (degree cm 2 dmol Ϫ1 ). Because the [] 222 value of a typical random structure is about 0.3 ϫ 10 4 (degree cm 2 dmol Ϫ1 ) (18), both ␣CTDs were not denatured completely at 90°C. However, the partial denaturation at 70°C of Ec ␣CTD was obvious compared with that of Tt ␣CTD.  (5). The sequence of fragment TUP is located directly upstream of the promoter of 16 S rRNA from Tt (9). Fragment SUB was the control for these DNA fragments (5). The sequence of fragment Ϫ46 has the smaller unit (AAATTT) to bind Ec ␣CTD, extracted from the Ec UP element. 2 Fragment NON is a reference of fragment Ϫ46.

FIG. 2.
A, UV CD spectra of Tt ␣CTD and Ec ␣CTD at 37°C. The former is shown as a solid line and the latter as a broken line. Ec ␣CTD was prepared as described (6). B, temperature dependence of the ellipticity at 222 nm for Tt ␣CTD and Ec ␣CTD, indicated as a solid line and a broken line, respectively.  NMR Analysis of Tt ␣CTD-All chemical shifts of NH, N, C␣, and H␣ in the backbone were assigned, except for those of Met at the N terminus, Ala 232 and Val 233 . Their signals were not observed in the 1 H-15 N HSQC spectrum (Fig. 3). The Met at the N terminus may be degraded in Ec cells, for example, by methionine aminopeptidases (19,20). In this case, the signal of the amide proton of Ala 232 would not be observed, because the rate of exchange of the amino proton at the terminus is high. That of Val 233 is also weak, because the exchange rate becomes higher with the charge of the N terminus. The resonances of the amide protons of 11 residues (Leu 269 , Lys 270 , Glu 272 , Leu 280 , Leu 281 , and Ile 302 -Glu 307 ) remained at 37°C for longer than 10 min after dissolution in D 2 O. Twenty-one resonances remained at 20°C after 60 min. Among them, the amide protons of 19 residues (Leu 259 , Leu 269 -Glu 272 , Ile 274 , Leu 280 -Lys 283 , Ile 291 , Leu 299 , Ile 302 -Glu 307 , and Lys 309 ) were considered to form hydrogen bonds and were used as distance restraints of hydrogen bonds for calculation of the structures of Tt ␣CTD. Other slowly exchanged amide protons were not assigned to specific hydrogen bonds.
Because there were no long range nuclear Overhauser en-hancement signals in the Ala 234 -Pro 246 region, the distance constraints in the Glu 247 -Glu 315 region of Tt ␣CTD were analyzed for structure calculation. A summary of the structural statistics for the final set of structures of Tt ␣CTD (Glu 247 -Glu 315 ) is given in Table I. The overlaid 20 backbone structures are shown in Fig. 4A. The averaged and energy-minimized structure was analyzed by means of the PROCHECK NMR program (21). The results showed Tt ␣CTD comprises Leu 256 -Leu 259 (helical turn), Thr 263 -Gly 273 (helix 1), Val 277 -Ala 282 (helix 2), Leu 285 -Leu 288 (helix 3), and Glu 296 -Lys 309 (helix 4) (Fig.  5). From the deviations of 13 C␣ chemical shifts from a random coil, the secondary structure of Tt ␣CTD has been estimated (22). This structure was almost identical with the current result. The rmsd for the backbone heavy atoms in the Leu 256 -Lys 309 region is 0.61 Å (Table I). Helices 1, 2, and 3 are almost perpendicular to each other. Although helix 3 is rather short according to the results with the PROCHECK NMR program, this region was determined to be a helix on comparison of the secondary structures of Tt ␣CTD and Ec ␣CTD (Fig. 5). Helix 3 is roughly antiparallel to helix 4 (Fig. 4B).  Fig. 3. They drifted either over 0.7 ppm along the axis of HN or over 2 ppm along the axis of N in the presence of 0.2 mM TUP. It is likely that these three amino acid residues directly interact with DNA. Although amino acid residues Thr 263 , Arg 264 , Val 265 , His 267 , Leu269, Ser 268 , Glu 296 , and Arg 297 in Tt ␣CTD correspond to the important residues for the DNA interaction in Ec ␣CTD (Fig. 5), their degrees of perturbation were below 0.3 ppm along the axis of HN and below 1 ppm along the axis of N in the presence of 0.2 mM TUP.
The perturbed chemical shifts of Gly 293 , Gly 295 , and Ser 298 were aligned in the order of NON Ͻ SUB and UP Ͻ Ϫ46 Ͻ TUP (0.2 to 0.4 mM) on straight lines (Fig. 3). The mode of specific binding is probably the same as that of nonspecific binding, because their shifts were aligned on the same linear lines. When the concentration of TUP was increased with a constant concentration (0.2 mM) of Tt ␣CTD, the observed shifts of Gly 293 and Ser 298 were saturated with 0.3 mM TUP (Fig. 3), indicating that most Tt ␣CTD formed the Tt ␣CTD⅐TUP complex under these conditions. From this titration of the chemical shifts, we could estimate the dissociation constant between Tt ␣CTD, and TUP was 10 Ϫ4 M assuming a 1:1 complex. This value is 10 times smaller than for the ordinary nonspecific interaction, as observed for Ec ␣CTD⅐SUB. 2 On the other hand, no obvious difference between those of Tt ␣CTD⅐UP and Tt ␣CTD⅐SUB was observed. Tt ␣CTD does not exhibit meaningful affinity to UP. Because the affinity of Tt ␣CTD for fragment NON, which is composed of G and C, was the lowest among those of the complexes investigated in this study, Tt ␣CTD does not prefer just GC-rich sequences. Fragment Ϫ46 has the internal sequence of the UP element to which Ec ␣CTD binds with as high affinity as in the case of UP. 2 Although the affinity of Tt ␣CTD for Ϫ46 was larger than that not only for NON but also that for UP, it was less than that for TUP.
When the perturbed signals in 1 H-15 N HSQC spectra were observed at 50°C, the degrees of the perturbation were roughly the same as those at 37°C. Those of Gly 293 , Gly 295 , and Ser 298 were larger than those of others. The affinity of Tt ␣CTD for TUP was also recognized at 50°C.

Comparison of the Solution Structures of Tt ␣CTD and Ec
␣CTD-The secondary structure of Tt ␣CTD is almost identical with that of Ec ␣CTD (Fig. 5). However, Ec ␣CTD has two helical turn regions, Pro 251 -Leu 254 and Val 257 -Asp 259 , whereas Tt ␣CTD has a single helical turn, Leu 256 -Leu 259 . Besides, Ec ␣CTD has the C-terminal loop (Met 316 -Glu 329 ), whereas Tt ␣CTD does not have such a loop at the C terminus. The Ala 234 -Pro 246 region of Tt ␣CTD is expected to be flexible, because there were no long-range nuclear Overhauser enhancement signals in the region, and the corresponding region of Ec ␣CTD (Arg 235 -Glu 248 ) is also flexible (23).
The backbone folding of Tt ␣CTD (Glu 247 -Glu 315 ) is quite similar to that of Ec ␣CTD (Glu 248 -Glu 329 ) (Fig. 6A). Comparison of the Leu 256 -Lys 308 region of Tt ␣CTD with the Val 257 -Ser 309 region of Ec ␣CTD gives an rmsd value of 1.70 Å for the backbone atoms despite the 32% identity of their amino acid That of Ec ␣CTD was as described previously (23). The regions in blue letters form an ␣ helix (the Leu 256 -Leu 259 region comprises one turn of a helix called a helical turn in the text). Asterisks and dots indicate identical and similar amino acid residues, respectively. The residues are numbered from the N terminus of a full-length ␣ subunit. Green arrows indicate the important amino acid residues for the interaction with DNA shown by the perturbation experiments. These amino acid residues are Arg 265 , Ser 266 , Asn 268 , Leu 270 , Asn 294 , and Ser 299 in Ec ␣CTD (6), and Gly 293 , Gly 295 , and Ser 298 in Tt ␣CTD (this study). Yellow arrows indicate Arg 265 , Asn 268 , Cys 269 , and Lys 297 in Ec ␣CTD, which are important for the interaction with the UP element shown by the Ala scan experiment (7). sequences. Helix 2 of Ec ␣CTD is considered to be essential for the formation of the hydrophobic core (Phe 249 preceding helical turn 1 and Trp 321 and Ile 326 in the C-terminal loop, shown in Fig. 6A) to stabilize the protein structure; the hydrogen-deuterium exchange rates of the amide protons in this helix were very slow (6). In Tt ␣CTD, however, the exchange rates in helix 2 are not slow compared with those in helix 4. This is probably due to the fact that Tt ␣CTD lacking the C-terminal loop allows helix 2 to expose for solvent. However, the dependence of the ellipticities of Tt and Ec ␣CTDs on temperature showed that the thermal stability of Tt ␣CTD is higher than that of Ec ␣CTD. Thus, the C-terminal loop of ␣CTD is not essential for stabilizing the protein folding. Another role of the C-terminal loop in Ec ␣CTD could be interaction with activator proteins, for example, OmpR (24).
When the tertiary structure of Tt ␣CTD was analyzed with DALI (version 2.0) (25), the similarity was indicated with that of the human barrier to autointegration factor (BAF) (26) (Z score, 3.1) and that of Ec ␣CTD (Z score, 7.6). BAF is an 89-residue protein, and it is composed of five helices. The result with DALI showed that the structures of helix 1-loop-helix 2 and that of helix 3-loop-helix 4 in Tt ␣CTD correspond roughly to that of helix 2-loop-helix 3 and that of helix 4-loop-helix 5 in BAF, respectively. The arrangement of the helix-loop-helix regions of Tt ␣CTD also corresponds roughly to that of BAF (Fig.  6B). BAF is the only protein whose folding is similar to that of eubacterial ␣CTD.
Interaction between Tt ␣CTD and DNA-The consensus UP element sequence of Ec was determined using an in vitro selection procedure (27), it being found that A tracts in the UP element are important for increasing the activity of the Ec rrnB P1 promoter (28). A tracts in DNA intrinsically cause bending of the DNA molecule (29). The UP element in Ec is AT-rich and also has a tendency to be bent (30). This bending causes narrowing of the minor groove that contributes to the affinity to Ec ␣CTD, as observed by NMR measurements. 2 Furthermore, it was suggested that the activation mechanism by the interaction between ␣CTD and A-tract DNA should be applicable throughout eubacteria (27). If this mechanism is used in Tt cells, A tracts must be remarkably characteristic of the sequence of the genome, of which the average G ϩ C content is 70% (8). However, Tt ␣CTD does not bind to Ec UP element DNA with meaningful affinity. At least in Tt, it is thought that transcriptional regulation by the interaction between ␣CTD and A-tract DNA does not occur.
Therefore, in what points is the mode of DNA binding of Tt ␣CTD different from that of Ec ␣CTD? In Ec ␣CTD, helix 1, helix 4, and the loop preceding helix 4 directly interact with DNA (6). In Tt ␣CTD, however, only helix 4 and the loop preceding helix 4 are involved in the interaction with DNA, and the Gly 293 residue enables the loop to form a ␤-turn structurally (Fig. 4B). Thus, the DNA binding site in Tt ␣CTD is expected to form a typical phosphate-binding helix-turn-helix module that binds to DNA without sequence specificity (31).
The DNA binding surface in Ec ␣CTD has a positive charge because of residue Arg 265 on helix 1, and residues Lys 297 and Lys 298 on helix 4 (Fig. 7A). These residues are involved in the DNA recognition, as indicated by the NMR experiments (6) and the Ala replacement experiments (7). In Tt ␣CTD, however, the three residues, Gly 293 , Gly 295 , and Ser 298 , were indicated to be important for the interaction in this study, and they do not have charged side chains. The amide protons in these residues, particularly Gly 293 and Ser 298 , are expected to form hydrogen bonds with the phosphate backbone in DNA. Although the surface for these residues on Tt ␣CTD corresponds to the DNA binding surface on Ec ␣CTD, the deviation of the electric charge in this region is less because Lys 297 (charged positive) in Ec structurally corresponds to Glu 296 (charged negative) in Tt (Fig. 7B). Not only Glu 296 in Tt but also Arg 264 in Tt corresponding to Arg 265 in Ec did not show a large shift. Thus, the interaction between Tt ␣CTD and DNA should not be due to the localization of the electric charge on the surface. However, it is difficult to explain why Tt ␣CTD has higher affinity to TUP than UP and SUB, because the nucleotide sequence of TUP is not palindromic nor AT-rich like UP.
Perturbation experiments should be carried out at 75°C, which is the optimal growth temperature of Tt (8). However, at 60°C, some signals of the amide protons disappeared because of their fast exchange. Besides, it is difficult to perform experiments at 75°C, because the DNA oligomers easily melt. However, the interaction between Tt ␣CTD and TUP at 50°C should have biological meaning, because the minimum growth temperature of Tt is 47°C (8).
Some Tt promoters have upstream regions (up to 50 base pairs from the Ϫ35 region of the promoter) whose A ϩ T contents are 42-46%, and their transcription activities in vivo FIG. 7. Molecular surfaces of (A) Ec ␣CTD and (B) Tt ␣CTD. The molecular surfaces are colored according to the electrostatic potential; blue corresponds to a positive potential and red to a negative potential. Residues whose amide resonances were perturbed by the presence of DNA are connected by green lines (Ref. 6 and this study). Residues in Ec ␣CTD connected by yellow lines are important for the interaction with the UP element in Ec ␣CTD indicated by the Ala scan experiment (7). Residues Arg 265 and Lys 297 in Ec ␣CTD correspond to Arg 264 and Glu 296 in Tt ␣CTD, respectively, whose positions are shown in B. Their surfaces are viewed from the left side of the view shown in Fig.  4. These figures were calculated using the GRASP program (33).
are higher than those of other Tt promoters (32). We, however, found that the affinity of Tt ␣CTD for TUP is higher than that for SUB, though the former A ϩ T content is also 40%, and the same applies to the latter. Thus, 40% A ϩ T content of a promoter upstream region should not be sufficient for transcription activation. Although SUB was used as the control for the Ec UP element (5), TUP locates in the upstream region of the promoter of the Tt 16 S rRNA operon and corresponds to the Ec UP element, judging from the genome structure. If we can assume the biological function of TUP is similar to that of the Ec UP element, TUP might have the ability to activate the transcription of the 16 S ribosomal RNA operon by interacting with the ␣ subunit of the RNA polymerase in Thermus thermophilus.