217825a0Nature2175131196803028258270028-0836196810.1038/217825a0ukNatureNatureNATUREnatureNature is a weekly international journal publishing the finest peer-reviewed research in all fields of science and technology on the basis of its originality, importance, interdisciplinary interest, timeliness, accessibility, elegance and surprising conclusions. Nature also provides rapid, authoritative, insightful and arresting news and interpretation of topical and coming trends affecting science, scientists and the wider public./nature/journal/v217/n5131issueJournal homeArchiveCurrent issueAdvance online publicationPrivacy policySubscribeNature Publishing GroupCurrent issue217825a0New Controlling Element in the Lac Operon of E. coli AU - IPPEN, KARIN AU - MILLER, JEFFREY H. AU - SCAIFE, JOHN AU - BECKWITH, JONATHANDepartment of Bacteriology and Immunology, Harvard Medical School, Boston, Massachusetts[ast]Harold C. Ernst Research Fellow.[dagger]On leave from MRC Microbial Genetics Research Unit, London.The lac promoter maps between the represser i gene and the operator o gene, so that the operator gene is probably transcribed.THE genetic elements (lac) determining lactose metabolism in E. coli map very close together in a small region of the chromosome (Fig. 1)1. Three structural genes, z (for b-galactosidase), y (for galactoside permease) and a (for thiogalactoside transacetylase), are regulated by the re-pressor product of the i gene. In the absence of b-galactoside inducers of the system, the repressor interacts with the lac operator (o) to prevent synthesis of the three proteins. Jacob and Monod have suggested that repression involves an inhibition of transcription of lac operon DNA into a messenger-RNA (mRNA) copy1. This model for regulation has recently received strong support from the experiments of Gilbert and Muller-Hill, who have isolated the repressor protein and have found that it binds specifically to lac operator double-stranded DNA2. Thus it seems likely that this represseroperator complex prevents transcription of the lac operon. There is genetic evidence indicating that it is not the operator itself which serves as a transcription initiation site. Whereas one would expect mutations in an initiation site to alter the potential for operon expression, operator-constitutive (oc) mutations, including some well characterized deletions (unpublished results of Davies and Jacob), do not affect the rate at which the operon can be expressed3. Such an initiation site must therefore lie outside the operator, either between i and o or between o and z. Jacob, Ullman and Monod3 have proposed that such a site, the promoter (p), maps between o and z. We have reported the isolation and characterization of mutants which have all the properties of mutants in an initiation site of the lac operon4. These mutants, which reduce the maximum rate of lac operon expression, do not alter either the operator or the i gene. On the basis of genetic mapping, we had concluded that these mutants altered a site between o and z, possibly identical to the promoter. This conclusion was based on the use of certain i oc mutants thought to be deletions extending from the operator into or beyond the i gene3. Davies and Jacob have now shown, however, that these mutants are not overlapping deletions (unpublished results). In fact, it now seems likely that most or all of them behave as either negatively complementing i mutations (personal communication from J. Davies and W. Gilbert) or as double mutants (unpublished results). The former class would be mutations in the i gene, which produce a defective sub-unit of the repressor, impairing the normal function of wild-type repressor sub-units. The findings of Davies and Jacob have led us to re-examine the mapping of our mutants. We present here conclusive evidence that the genetic site altered in our mutant maps between i and o. This evidence thus identifies a new controlling element determining the expression of the lac operon. We propose that it is in this region that the initiation of transcription takes place. Because, in the original formulation of the promoter concept, it was considered most likely that the site defined as promoter was an initiation point for mRNA transcription3, we prefer to retain the term promoter to describe this new region (Jacob and Monod, personal communication, agree with this use of the term promoter). It now seems probable that, in fact, the region between o and z corresponds only to an initiation site for translation. We have characterized four ultra violet-induced mutants which co-ordinately reduce the rate of expression of the z, y and a genes. Some of their properties are described in Table 1. Three of these, although independently isolated, are almost identical in their characteristics, and at least two, L8 and L37, do not recombine with each other (unpublished results of J. Beckwith). L8 and L37 do, however, seem to be point mutants9. The lesion in the fourth mutant, L1, confers an additional property on the strain, and, although the operator in L1is normal, the activity of its i gene is reduced. Table 1 Induced levels of lac enzymes as percentage of wild-type [Regulation m-Dominant LI 2 Partly constitutive Yes [pound]8, [pound]37 6 Normally inducible Yes [pound]29 4 Normally inducible Yes Three of these mutants, L1, L8 and L37, have been described before4. The fourth, L29, was isolated as one of two lac mutants found after ultraviolet mutagenesis of a lac+ strain. The o and i characters of L1 were determined in diploid studies. These mutations are cis-dominant in that they only reduce the rate of expression of the operon on the same chromosome and these effects are not relieved by the introduction of a second lac region into the cell4. An explanation for the properties of these mutants is that they have altered a site essential for the initiation of either transcription or translation of the operon. The result is a reduced efficiency of reading. In order to determine conclusively the position of the mutants relative to the operator and the i and z genes, we have employed a deletion analysis of this region. To avoid confusion we shall call these promoter mutants. We have described a genetic system in which it is possible to isolate deletions removing the distal end (a end) of the operon and extending varying distances toward ;and past the operator (Fig. 1)5. These deletions are isolated in a strain in which a 80dlac transducing phage is integrated in the chromosome near the locus (T1) determining sensitivity to the bacteriophages T1 and 80. In the strain described in Fig. 1, the lac region is inverted from its normal orientation on the chromosome so that the i gene is furthest from the T1 locus. By selecting T1R mutants of such a strain, it is possible to isolate deletions extending from the T1 locus or beyond into the defective prophage5,6. These deletions are presumed to include all the genetic material between the T1 locus and the determined deletion end, for they remove all point mutant sites in the y gene and in the z gene as indicated. Fig 1 The lac operon transposed (in80 dlac) to the 80 site on the bacterial chromosome. The T1R mutations were selected as before7. All i gene and oc mutations were isolated by Jacob, Monod and co-workers. LacMS272 is an extremely polar mutant isolated by Malamy8. Lac2 and lacYA559 have been described before9. The orders of i and o mutations which are not established by deletion mapping have been determined by Davies and Jacob (results in preparation). The deletions isolated in this way, which we have used in our mapping studies, are described in Fig. 1. Table 2 shows the frequency of recombination of these deletions with mutants affecting various components of the lac operon and the i gene. These include an is mutation, two i- mutants, four promoter mutants, three oc mutants and three z- mutants. One of the mutants, oc307, was shown by Davies and Jacob (unpublished results) to map the furthest to the left of any known oc mutant. In most cases, the deletion strains were infected with an F factor (F-lac-pro) carrying the mutation to be mapped. Cultures of these partial diploids were then screened for recombinants. It should be noted therefore that the frequencies reported in Table 2 do not correspond to recombination frequencies measured in the standard way. The crucial deletion for determining the position of the promoter mutants relative to the operator is X8554. Two promoter mutants, L8 and L37, give frequencies of 0.0650.086 per cent recombinants with this deletion. In contrast, no recombinants have been detected between this deletion and any of the oc mutants as indicated in Table 2. Either this deletion covers the entire operator region, or else its terminus lies very close to the end of the operator nearest the i gene. In general, the other frequencies presented in Table 2 are entirely consistent with the order of mutations as determined by their ability to recombine with a series of other point mutations and deletions. The map order is also in complete agreement with respect to the order of i and o mutations determined by Davies and Jacob (unpublished results). Furthermore, the results unambiguously show that the position of all four promoter mutations is between i and o. We have presented genetic evidence showing that mutants affecting the maximum level of lac operon expression map in the region between i and o. The properties of these mutants can be accounted for by two different general explanations. (1) These mutants define a site necessary for expression of the lac operon, which can be altered by mutation to reduce operon activity. (2) The mutants map outside all controlling sites for the operon, but in some undefined way affect operon expression negatively. Although it is very difficult to conceive of any simple mechanisms for the second explanation, it is, of course, imperative to show directly that these mutants define an essential site. There is some evidence indicating that the region between i and o is essential to operon expression. It is possible to isolate lac-constitutive mutations which are caused by deletions (Table 2) cutting into either the i gene (class I) or the o region (class II) from a point outside the i end of the lac region5. In such a system we have isolated a large number of i lac+ deletions. In addition, we have isolated one deletion of class II. This deletion removes the i gene and the promoter region, but leaves part of the operator intact. As expected, the expression of the operon is impaired by the deletion of the promoter region. The properties of this strain will be described elsewhere (unpublished results of J. H. Miller, J. R. Beckwith and E. R. Signer). In addition to confirming the mapping results described here, the properties of this deletion also point to the importance of the promoter region in the expression of the lac operon. Thus it seems most likely that the mutants, L1, L8, L37 and L29, do lie in an essential region (the promoter) determining the initiation of operon expression. Table 2. PERCENTAGE OF RECOMBINANTS AMONG TOTAL COLONIES Deletion JT8604 JT8507 ^8554 X8508 Mutant ZF^559 0 0 0 0 2 + 0 0 0 ZMS27'2 + + 0 0 <4 0-26 0-037 < 0-001 n.t. 4)7 n.t. n.t. < 0-001 n.t. <& n.t. n.t. < 0-002 n.t. LI, [pound]29 + + + 0 LS + + 0-065 0 [pound]37 0-79 0-34 0-086 0 is n.t. n.t. 0-19 0-009 1522 n.t. n.t. 0-16 0-094 [ast]FJ694 + + 0-24 + The deletions are isolated in a strain carrying the 80 lac as described in the text and also carrying a lac-pro A,B deletion To estimate the percentage of recombinants we constructed strains m which the mutant to be mapped against the deletion was carried on an F-lac-pro A,B+ episome. with the dominant i mutation (i522) and the oc mutants, aliquots of a culture of a strain partially diploid, in this way, for the lac region, were spread on to minimal medium (M63) containing glucose as carbon source, 5-bromo-4-Sro-indoxyl-b-D)-galactoside (Xg; 0.004 per cent) and 0.004 M sodium citrate, necessary for the growth of T1 R strains on this medium. Because Xg is not an inducer8, the constitutive diploid strains were blue and any i+o+ recombinants were white on these plates. White colonies appearing on such medium were purified and assayed for b-galactosidase to ensure that they carried a wild-type lac region. The same method was used to map the i deletion X8508 against i3. The frequency of the i gene mutations with X8554 is actually the frequency of homogenotization of the i+ or i allele To map the promoter mutants, L8 and L37, cultures of partially diploid strains were spread on lac-EMB plates and pure lac+ (wild-type) colonies were picked and verified by assay. The percentage of recombinants is based on samples from many independent diploid cultures; the frequencies are not seriously biased by "jackpots" caused by recombination events occurring early in the growth of the culture. Some mutations (in F donors) were mapped by crossing against the F- lac Sr strains and selecting lac+ recombinants (spot tests). Positive results are indicated as + and negative as 0 These spot tests detect recombinants at a much higher efficiency than the scoring techniques described here. Some combinations were not tested (n.t.). The evidence pointing to a new site essential to lac operon expression which maps between i and o provides strong indication that the transcription process is initiated before the operator region. This conclusion does not exclude the possibility that the initiation of operon translation also occurs in this region. If protein synthesis does begin before o, however, then the operator should be translated; there is no evidence to support this prediction. We therefore favour the view that there are two sites essential to the expression of the lac operon. One would be the promoter, lying between i and o, which is the site of initiation of mRNA synthesis. Initiation of translation may then take place at a second site in the oz boundary region, denned by Jacob, Ullman and Monod3. With this picture of the organization of the controlling sites of the lac operon, it is possible to visualize a fairly straightforward mechanism for the expression and regulation of operon activity. According to this scheme, the promoter region serves as the initiation point for transcription, possibly by acting as a binding site for RNA polymerase. Promoter mutants would reduce the site's affinity for the polymerase. While it is possible that the represser hinders binding of RNA polymerase, the results presented here suggest a very simple alternative. In binding to the operator, the represser could directly block the progress of the RNA polymerase into the structural genes of the lac operon. This work was supported by a grant from the US Public Health Service and a grant from the US National Science Foundation. J. Miller was supported by a Public Health Service training grant to the Department of Biochemistry and Molecular Biology, Harvard University.Jacob, , F., and Monod, , J., J. Mol. Biol., 3, 318 (1961).PubMedISIChemPortGilbert, , W., and Muller-Hill, , B., Proc. US Nat. Acad. Sci., 56, 1891 (1966); Gilbert, , W., and Muller-Hill, , B., Proc. US Nat. Acad. Sci. (in the press).ChemPortJacob, , F., Ullman, , A., and Monod, , J., J. Mol. Biol., 13, 704 (1965).ISIChemPortScaife, , J., and Beckwith, , J. R., Cold Spring Harbor Symp. Quant. Biol., 31, 403 (1967).ISIBeckwith, , J. R., Signer, , E. R., and Epstein, , W., Cold Spring Harbor Symp. Quant. Biol., 31, 393 (1967).ISIFranklin, , N. C., Dove, , W. F., and Yanofsky, , C., Biochem. Biophys. Res. Commun., 18, 910 (1965).ArticleISIBeckwith, , J. R., and Signer, , E. R., J. Mol. Biol., 19, 254 (1966).PubMedISIChemPortMalamy, , M., Cold Spring Harbor Symp. Quant. Biol., 31, 89 (1967).Newton, , W. A., Beckwith, , J. R., Zipser, , D., and Brenner, , S., J. Mol. Biol., 14, 290 (1965).PubMedISIChemPort