Transcriptional Regulation of the Cytochrome bhG2-o Complex in Escherichia coli GENE EXPRESSION AND MOLECULAR CHARACTERIZATION OF THE PROMOTER*

Transcriptional regulation of the cyo operon coding for a terminal oxidase, the cytochrome bGG2-o complex, of the aerobic respiratory chain of Escherichia coli was studied using a chromosomal operon fusion tech- nique with the 1acZ gene. Expression of the cyo gene was found to be subject to catabolite repression and also to control by the oxygen concentration. Informa- tion on the mechanisms of these regulations was ob- tained by nucleotide sequencing of the regulatory region, which corresponds to the 5’-flanking region of the cyoA gene. A typical promoter sequence and two noteworthy composite structural features, that is, two potential catabolite gene activator protein-binding and a region of hyphenated dyad symmetry were found. The transcription start point was identified by primer extension analysis, the in the nucleotide The homology of the nucleotide in the region of hyphenated dyad nucleotide

The aerobic respiratory chain of Escherichia coli branches at a ubiquinol site into two terminal oxidases, cytochrome bSG2-o and cytochrome bsss&& complexes, which both catalyze four-electron transfer to molecular oxygen and couple this redox reaction with the generation of a proton motive force across the membrane (1). Mutants lacking either of the oxidases can grow normally, but a mutant lacking both enzymes cannot grow aerobically on nonfermentable substrates such as succinate (2).
These two terminal oxidases have been studied extensively both biochemically and at a molecular level (3). The cytochrome bS6.-o complex has been purified and shown to consist of two or four polypeptides (4, 5). The two major subunits (subunit I and II) seem to be functionally essential for ubiquinol oxidation and generation of a membrane potential (4). The genes encoding these subunits are located at 10 have not yet been well characterized. The cyo gene is one such gene as evidenced by spectroscopic (11) and immunological (12) findings that synthesis of cytochrome o is derepressed by high oxygen tension. Broman et al. (13) showed that a cya mutant contains a reduced level of cytochrome o and that the level is restored to normal when the mutant is grown in the presence of CAMP. These results suggest that transcription of the cyo gene is derepressed aerobically and is subject to catabolite repression. Therefore, it is interesting to study the transcriptional regulation of the cyo gene by the operon fusion technique (14) and to compare the nucleotide sequence of its regulatory region with those of other genes controlled by oxygen.

Effect
of Oxygen on cyo Gene Expression-The transcriptional activity of the cyo gene during aerobic growth was monitored by assay of the specific activity of P-galactosidase in strain ST2539, which has cyo-la&Y operon fusion (Fig. 2). The results demonstrated that cyo gene expression under aerobic conditions reached a maximum in the early exponential phase (ODBBO I O.l), and that the gene was repressed in later phase. This fact suggests that the repression of cyo gene expression resulted from the increased number of cells, or changes in the dissolved oxygen concentration in the culture. To substantiate the effect of oxygen concentration on cyo gene expression, three sets of oxygen-limited cultures were performed and their fl-galactosidase activity was monitored (Fig. 3). To evaluate only the oxygen effect, cells were harvested in the early exponential phase corresponding to the peak in Fig. 2  Effect on cyo Gene Expression-We examined the effects of various carbon sources on cyo gene expression. Strain ST2539 was grown aerobically with various compounds as the sole carbon source, and harvested in the early exponential phase at ODFso 5 0.1 (see Fig. 2), and then its P-galactosidase activity was measured. As shown in Table II repression by these compounds. Next, cyo gene expression in the Acya mutant harvested at the early exponential phase was monitored as /3-galactosidase activity. The activity in the mutant was one seventh of the wild-type level and was restored to the normal level by addition of 2.5 mM CAMP (Table  II).
The possibility that the reduced cyo gene expression under anaerobic conditions was due to a decreased level of intracellular CAMP was investigated.
When strain ST2539 was grown anaerobically with 2.5 mM exogenous CAMP, its level of cyo gene expression was not restored to the control level (data not shown).
Nucleotide Sequence-From the above results, cyo gene expression seemed to be subject to catabolite repression and to be controlled by the oxygen concentration in the medium. Therefore, we determined the nucleotide sequence of the regulatory region of the cyo gene. The regulatory region has been mapped within the 1.7-kb SalI-SalI fragment of pHNl1 Transcriptional Regulation of E. coli Cytochrome o Complex which contains the cyoA gene and its 5'-flanking region (8).
As shown in Fig. 4, There is a single open reading frame of sufficient length to encode subunit II between nucleotide position 44 and 988 (data not shown). The deduced M, of the polypeptide is 34,911 (315 residues), in good agreement with the M, = 35,000 of subunit II of the purified enzyme (4, 5). The first ATG triplet is located seven nucleotides downstream of a typical Shine-Dalgarno sequence, GAGGT (positions 33-37 in Fig. 5), that has five consecutive bases complementary to the 3'-terminal sequence of 16 S ribosomal RNA (30).
In the upstream region of the transcription start point, we identified two potential CAP-binding sites, positions -233 to -212 for CAP1 and -81 to -60 for CAP2 (Fig. 5), that resemble the 22-bp consensus sequence, 5'-AANTGT-GANNNNNNTCACANTT-3' (31). This is consistent with the fact that the cyo operon is subject to catabolite repression. At present, it is not clear which CAP site is functional.
The cyo regulatory region contains a single region of ATrich hyphenated dyad symmetry (positions -217 to -195) adjacent to the CAP1 site (Fig. 5). Thus this region could be a binding site of a regulatory protein for the control of expression by oxygen concentration. A computer search for other genes sharing a homologous nucleotide sequence to this  In the case of g&A-sdhC, where divergent transcription occurs, the sequence is numbered with respect to the sdhC transcription start point. As the transcription start points of fum.A, fumC, and cyd have not been determined, these sequences are not numbered. dyad symmetry in their 5'-flanking region was then carried out. Six genes having a similar region were found: the cyd gene for the cytochrome b558-b595-d complex (32), the sdh gene for succinate dehydrogenase (33), the gltA gene for citrate synthase (33), the lpd gene for E3 component of 2-oxoglutarate dehydrogenase (34), and the fumA and fumC genes for fumarase (35,36). The homologous regions of the sequences are shown in Fig. 6. The cytochrome bss&g5-d complex is an alternative terminal oxidase of the aerobic respiratory chain and is reported to be expressed in the mid-to late exponential phase (37). The other five enzymes are components of the citric acid cycle and have been shown to be repressed under anaerobic conditions (38)(39)(40). Transcript Analysis-The start site of the cyo mRNA was identified by determining the 5'-end of the cyo transcript by primer extension analysis. A 20-mer complementary to nucleotides 206-225 was used as a primer. As shown in Fig. 7, RNAs from the wild-type strain (ST2539) grown under aerobic conditions (lane I) and the Acya strain (ST2571) grown with 2.5 mM CAMP under aerobic conditions (lane 4) produced transcripts of identical length. Therefore, the major start site of mRNA synthesis was identified as A at position +l (Fig.  5). Aerobically grown wild-type cells (lane I) produced more of the transcript than did anaerobically grown cells (lane 2). Furthermore, the absence of bands in lane 3 indicates that the synthesis of the cyo transcript required CAMP. Upstream of the start point, there are potential promoter sequences, TAAAATG (positions -12 to -6) resembling the consensus Pribnow box, TATAATG (41) and TTTACA (positions -35 to -3O), homologous to the sequence TTGACA in the -35 region of many E. coli promoters. The latter sequence is most favorably positioned relative to the Pribnow box; that is, a 17-bp spacer region is located between the last base in the -35 region and the first base in the Pribnow box (42). DISCUSSION In this study, we characterized the transcriptional regulation of the cyo operon. We constructed a cyo-la&Y chromosomal operon fusion and examined la& expression under various growth conditions, such as aerobic and anaerobic conditions and with various carbon sources.
Broman et al. (13) showed spectroscopically that the synthesis of cytochrome o was subject to catabolite repression. Here, we showed by operon fusion that the expression of the cyo gene is subject to catabolite repression at the transcriptional level. Furthermore, to substantiate these results, we determined the nucleotide sequence of the regulatory region and found two potential CAP-binding sites. We also demonstrated qualitatively that CAMP was required for the production of the cyo transcript. All these results indicate that the cyo operon is subject to catabolite repression at the transcrip-Regulation Deletion studies on this region will answer the question.
The synthesis of cytochrome o was shown spectroscopically (11) and immunologically (12) to be derepressed by high oxygen tension. Here, we demonstrated by operon fusion that the level of cyo gene expression varied with the progression of growth phase. The cyo gene was highly expressed in the early exponential phase, and a high oxygen concentration (more than 78 FM) was shown to be required for expression of the cyo gene by the experiments using the oxygen-limited cultures. We think the repression of cyo gene expression in the late exponential and stationary phases was caused by lowered oxygen concentration.
Furthermore, the amounts of transcripts were more under aerobic conditions than under anaerobic conditions, confirming the effect of oxygen concentration qualitatively.
The fact that the addition of CAMP to the medium of anaerobic cultures did not restore the flgalactosidase activity indicated that the repression under anaerobic conditions was not due to depletion of intracellular CAMP. Thus, the regulatory effects of catabolites and oxygen tension on the expression of cyo gene are distinct.
At present, the promoter regions of five genes coding for enzymes that are expressed under aerobic conditions have been sequenced; these are the 5'-flanking region of sdhC (a subunit of succinate dehydrogenase), gltA (citrate synthase), Ipd (E3 component of 2-oxoglutarate dehydrogenase), and fumA and fumC (fumarase). These genes encode components of the citric acid cycle and all have a region of AT-rich dyad symmetry in their 5'-flanking region like that of the cyo operon. Although the relative positions of these regions from the transcription start point differ and the hyphenated dyad symmetry is not perfect, 17 bp (sdhC/gltA, fumA) or 15 bp (Ipd, fumC) of the total 23 bp are homologous. An attractive possibility is that this region is a recognition motif for an oxygen effector( Furthermore, there is a similar sequence in the upstream region of the cyd gene (19 of 23 bp are homologous), which codes for the alternative terminal oxidase, cytochrome bssx-b,,,-d complex. The cyo and cyd genes are not expressed in a coordinated manner: the former is extremely oxygen-dependent, whereas the latter is moderately oxygen-dependent.
However, both genes appeared to be regulated by the oxygen concentration of the medium. Additional experiments are underway to determine if the regulatory region of the cyo operon shares a common trans-effector with the genes mentioned above and if the region of dyad symmetry actually functions.
The respiratory chain of E. coli can use several compounds effectively as electron acceptors. Of the commonly available electron acceptors, such as oxygen, nitrate, and fumarate, oxygen is used preferentially under aerobic conditions. This study showed that the preferential utilization of oxygen is partially, if not entirely, ascribed to the transcriptional control. The expression of the cyd gene encoding the cytochrome b,5x-b5r,s-d complex was found to be elevated from the midexponential to the stationary phase (37). The nar operon encoding nitrate reductase and the frd operon encoding fumarate reductase are expressed only under strictly anaerobic conditions (43,44). The regulatory pattern of expression of the cyo gene is different from those of the other genes encoding terminal enzymes of the respiratory chain, but rather similar to those of the enzymes participating in the citric acid cycle (38,45). When glucose is present in the medium, only the enzymes participating in the glycolytic pathway are induced and energy is obtained mainly by this pathway. But when glucose is not available and the cells are growing aerobically, nonfermentable compounds are utilized as carbon sources and enzymes of the citric acid cycle and the cytochrome bsn2-o complex are induced by a mechanism of oxygen induction that overcomes the regulation by CAMP. We believe that the preferential utilization systems of glucose and oxygen are independent, but contribute mutually to regulation of energy economy of the cells by restricting the formation of enzymes for the glycolytic pathway and citric acid cycle/terminal oxidase pathway.