171 , No. 12 Mutations Affecting Regulation of Cobinamide Biosynthesis in Salmonella typhimurium DAN I. ANDERSSONt AND JOHN R. ROTH * Department of Biology , University of Utah , Salt Lake City , Utah 84112 Received 28 April 1989/Accepted 5 September 1989 Transcription of the genes for cobalamin biosynthesis is reduced during aerobic-growth .
We isolated and characterized mutants that showed a 2-to 90-fold increase in aerobic expression of the cobinamide biosynthesis ( CobI ) genes , depending on the particular mutation and growth-conditions .
Four different classes of mutations were characterized .
All mutations ( CobRI through CobRIV ) were cis-acting , dominant mutations that mapped near the promoter end of the CobI operon .
Two of these classes of mutations ( Ill and IV ) caused an increase in anaerobic as well as aerobic transcription of the CobIl and CobIlI operons ; this led to increased biosynthesis of cobalamin under anaerobic-growth-conditions .
A recessive mutation ( cobF ) mapping far from the CobI operon increased anaerobic CobI operon expression by about fourfold .
Vitamin B12 ( cobalamin ) is synthesized by Salmonella typhimurium only under anaerobic-growth-conditions ( 15 ) ; expression of the biosynthetic genes is reduced in the presence of oxygen .
Neither the physiological significance nor the mechanism of this pattern of regulation is understood .
Cobalamin is used as a cofactor in four known reactions in S. typhimurium .
First , it is required by one of the two methyltransferases ( metE and metH gene products ) that independently can catalyze methylation of homocysteine to form methionine ( 19 , 24 , 25 , 27 ) .
Second , it is required for the cleavage of ethanolamine to acetaldehyde and ammonia , providing both a carbon and a nitrogen source ( 6 , 21 ) .
Third , Frey et al. ( 11 ) showed that B12 is involved in formation of the nonessential hypermodified Q base found at position 34 in the anticodon of tRNAASP , Asn , His , Tyr ; B12 is needed for conversion of epoxyqueuosine to queuosine .
Finally , it was recently shown ( R. M. Jeter , submitted for publication ) that S. typhimurium can use 1,2-propanediol as a carbon source under aerobic-growth-conditions only if vitamin-B12 is provided .
The pathway for breakdown of propanediol by S. typhimurium has not been characterized but probably includes the enzyme propanediol dehydratase , which is known to require B12 in other organisms ( 28 ) .
None of the known cobalamin-dependent reactions is essential for cell growth under most conditions , and none of them has so far been shown to be of special significance during anaerobic-growth .
Previous genetic analysis ( 15 , 16 ) has established that most of the cobalamin biosynthetic ( Cob ) genes map near the his operon at 41 min on the chromosome .
The genes appear to comprise three different operons ; the CobI operon is involved in cobinamide biosynthesis , the CobIl operon is involved in dimethylbenzimidazole ( DMB ) biosynthesis , and the Coblll operon encodes functions that join cobinamide and DMB to form cobalamin .
All three operons are transcribed counterclockwise with repect to the Salmonella genetic map .
Previous experiments established three exogenous factors that affect transcription of the Cob operons ( 10 ) .
The end product of the pathway , cobalamin , reduces transcription , t Present address : Department of Microbiology , University of Uppsala , Biomedicum , S-75123 Uppsala , Sweden .
and cyclic AMP ( cAMP ) serves to activate these genes .
Finally , transcription of the cobalamin biosynthetic genes is strongly reduced by the presence of oxygen , implying the existence of a specific regulatory mechanism that responds directly or indirectly to oxygen .
In the accompanying paper ( 1 ) we present experiments directed at how cells sense oxygen levels and how that information might be conveyed to the transcriptional apparatus of the CobI operon .
In this paper we report cisdominant mutations , mapping adjacent to the CobI operon , that increase CobI gene expression in the presence of oxygen .
We also describe a recessive mutation that maps far from the CobI operon and causes increased expression of CobI under both aerobic and anaerobic conditions .
MATERIALS AND METHODS Bacterial and bacteriophage strains .
The genotypes of bacterial strains are given in Table 1 .
All bacterial strains used are derivatives of S. typhimurium LT2 .
Two transpo-sition-defective derivatives of the specialized transducing phage Mu dl ( Ampr lac cts ) of Casadaban and Cohen ( 3 ) were used .
The derivatives , Mu d11734 Kanr ( 4 ) and Mu dl8Ampr ( 13 ) are referred to as Mu dJ and Mu dA , respectively .
Two transposition-defective derivatives of TnJO were used ; TnlOdell6dell 7Tetr ( 30 ) and TnlOdCam ( 9 ) .
TnJOdelJ6 dell 7Tetr is referred to as TnOdITet .
Finally , a Kanr deriv-ative of TnJO ( 18 ) was used and is referred to as TnJOKan .
We thank Charles Miller , Erica Barrett , Chris Higgins , respectively , for providing oxrA oxrB pepT , hyd fhl , and oxrC strains .
Throughout the paper , three lac operon fusions to the Cob operons were used ; all were formed by insertion of a Mu dJ element .
Fusion cob-24 : : Mu dJ is an insertion within the CobI operon ; fusions cob-62 : : Mu dJ and cob-66 : : Mu dJ are insertions in the CoblI and CobIlI operons , respectively .
We will refer to these insertions as CobI : : lac , CobII : : lac , and CobIII : : lac ; it should be kept in mind that CobI,-II , and-III refer to operons and phenotypic classes and not to particular genes .
Culture media and growth-conditions .
Difco nutrient broth ( 0.8 % ) containing NaCl at a final concentration of 0.5 % was used as a complex medium .
The E medium of Vogel and Bonner ( 29 ) was used as a minimal-medium ; with carbon sources other than glucose , no-citrate E medium was used .
672 Solid media contained 15 g of Bacto-Agar ( Difco ) per liter .
Carbon sources and electron-acceptors were present at the following concentrations : glucose ( 10 mM ) , glycerol ( 10 mM ) , and fumarate ( 20 mM ) .
Auxotrophic supplements were added at the concentrations recommended elsewhere ( 8 ) .
The final concentrations of antibiotics in complex media were 20 , ug of tetracycline , 50 , ug of kanamycin , 20 , ug of chloramphenicol , and 30 pLg of ampicillin per ml .
The final concentrations of antibiotics in minimal-medium were 10 jig of tetracycline , 125 , ug of kanamycin , 5 , ug of chloramphen-icol , 15 , ug of ampicillin , and 500 , ug of streptomycin per ml .
When added to plates , lactose analogs were used at final concentrations of 40 mg of 5-bromo-4-chloro-3-indolyl-P-D-galactopyranoside ( X-gal ) and 0.12 g of phenylethyl - , - D-thiogalactoside ( PETG ) per liter .
Cultures were grown aerobically in 0.5-ml volumes in 10-ml tubes at a speed setting of 8 in a Gyrotory shaker ( New Brunswick Scientific Co. ) .
Anaerobic cultures were grown under completely anoxic conditions as described previously ( 10 ) .
Cell density was monitored with a Klett-Summerson photoelectric colorimeter ( Klett Manufacturing Co. , Inc. ) .
Transductional and conjugational methods .
All transduction crosses were performed with the high-frequency , generalized transducing phage mutant P22 HT105/1 int-201 ( 23 ) as previously described ( 5 ) .
Transductants were purified and made phage free by picking light-colored single colonies from nonselective green plates ( 5 ) .
All conjugational crosses were performed by mixing liquid cultures ( 0.1 ml each ) of the donor and recipient strains for 1 h .
This mixture was plated on selective medium .
3-Galactosidase was assayed as described by Miller ( 17 ) by using CHC13 and sodium dodecyl sulfate to permeabilize the cells .
All values presented are Miller units ( 17 ) .
Cobalamin was assayed by using a commercial radio-isotope dilution assay ( Quanta phase ; Bio-Rad Laboratories ) as recommended by the manufacturer .
Diethyl sulfate ( DES ) mutagenesis was performed as described previously ( 8 ) , except that the time of mutagenesis was increased to 45 min .
Hydroxylamine ( NH20H ) mutagenesis was by the method of Hong and Ames ( 12 ) .
Phage lysates were mutagenized to a 0.1 % survival rate .
Generation of a CobI-carrying F ' plasmid .
A composite transposon consisting of the CobI operon , a part of the CobIII operon , and two flanking TnJO elements was used to introduce the CobI genes onto an F ' plasmid by transposasemediated transposition .
The general method of construction has been described previously ( 22 ) .
To construct F'CobI , the CobI region was first flanked by derivatives of TnJO .
The TnJOKan element ( 18 ) was placed outside the promoter end , and the cob4 : : TnJO insertion ( in CobIll ) was placed at the distal end , of the CobI region .
This constructed a composite CobI transposon capable of moving into a plasmid .
An F ' 128 pro ' lac + plasmid and a transposase-producing plasmid ( pNK972 ) were introduced the strain bearing the composite TnJO CobI + TnlO .
From this final strain ( TT14685 ) the F ' plasmid was transferred by conjugation into recipient strain TT14686 ( metE cobI217 recAl rpsLl142 ) , selecting for Lac ' exconjugants , which were replica printed to identify strains that had received the ( unselected ) Cob ' phenotype with the F ' plasmid that conferred the ( selected ) Lac ' phenotype .
One such Cob ' exconjugant strain that had not acquired resistance to tetracycline or kanamycin ( strain TT14687 ) was saved ; the F ' plasmid of this strain carries the CobI operon , which is presumably flanked only by copies of ISIO derived from the original TnWO elements .
This F ' plasmid was shown to contain the CobI operon and part of the CobIII operon by its ability to complement and repair ( by recombination ) point and insertion mutations .
A lacZ950 : : TnlO insertion from strain TT12242 was crossed into the Lac operon of the final F ' plasmid to abolish 3-galactosidase production .
This final plasmid ( in strain TT14688 ) was used as a donor of F ' CobI + for the dominance tests of the four cobR mutations .
The CobI-containing F ' plasmid from strain TT14688 was introduced into strains TT14676 through TT14679 ( cob-24 : : Mu dJ cobR ) by conjugation .
Cells were plated on minimal glucose-tetracycline-kanamycin plates , anaerobically selecting for Cob ' .
These exconjugants were purified nonselectively and assayed for P-galactosidase .
Strain TT14680 , containing a cob-24 : : Mu dJ fusion and a chromosomal duplication extending from pyrC to pncX , was used as a recipient in transduction crosses .
Donor phage grown on strain TT14682 ( zcd-3677 : : TnJOdCam cobFI ) introduced the cobFI mutation .
Transductants were selected ( Cam ' ) on minimal plates containing glucose , X-gal , PETG , and chloramphenicol .
All transductant colonies were light blue , indicating that the cobF mutation , which cause dark blue colonies in haploid strains , is recessive to cobF + .
To confirm the structure of each merodiploid , six transductants were purified , and a segregation test was performed .
Cells were grown overnight in a nonselective complex medium and plated on rich plates .
These colonies were replica printed to rich plates with-tetracycline , ampicillin , or chloramphenicol or to minimal E plates with X-gal and PETG to score for antibiotic resistance and color ( i.e. , the expression level of the CobI : : lac fusion ) .
This demonstrates the structure of the duplication and the presence of both the cobFl and cobF + alleles .
RESULTS Isolation of mutants with increased aerobic CobI operon transcription .
Spontaneous and chemically induced mutants with increased aerobic expression of a lac fusion to the CobI operon were isolated .
The level of 3-galactosidase produced by the parental CobI : : lac fusion ( strain TT10852 ) in the presence of oxygen was sufficiently high that these strains formed dark blue colonies on X-gal plates and were able to grow on minimal lactose plates .
To screen and select for mutants with increased aerobic CobI : : lac transcription , the P-galactosidase inhibitor PETG was added to the plates to inhibit P-galactosidase activity .
The concentration of PETG used ( 0.12 g/liter ) caused the parent strain to form light blue colonies under aerobic conditions and prevented growth of the parent strain on lactose as the carbon source .
Hydroxylamine-induced mutants were isolated after local mutagenesis of the CobI region .
A phage lysate was grown on strain TT10927 , which carries a TnJO insertion outside the promoter end of the CobI operon ; this phage lysate was used to transduce the CobI : : lac fusion strain TT10852 ( cob-24 : : Mu dJ ) , selecting for kanamycin ( Mu dJ ) and tetracycline ( TnJO ) resistance and screening for dark blue colonies on plates containing X-gal and PETG .
Such mutants were found at a frequency of approximately lo-3 .
DES-induced mutants were isolated after mutagenesis ( 8 ) of cells of the CobI : : lac fusion strain ( TT10852 ) .
After DES treatment , cells were plated on minimal glucose plates containing X-gal and PETG .
Dark blue colonies were found at a frequency of 10 ' to 10-4 .
To assure that all mutants are independent , only one colony was picked from each mutagenized culture .
Spontaneous mutants were isolated after a positive selection for growth on lactose .
The CobI : : lac fusion strain ( TT10852 ) was plated on lactose minimal plates containing PETG , and large Lac ' colonies were picked .
This type of mutant was found at a frequency of 1o-7 to 10-6 .
Classification of regulatory mutants .
These mutants were classified according to the magnitude of their effect on CobI : : lac expression .
Four different classes of cobI-linked mutants were found and given phenotype designations CobRI through-IV .
We isolated and initially characterized nine mutants of class I , three mutants of class II , seven mutants of class III , and four mutants of class IV .
From each class , one typical mutant was characterized in detail ( strains TT14676 through 9 , respectively ) .
The chosen mutations , representing classes I through IV , were designated cobRI through 4 , respectively .
After either hydroxylamine or DES mutagenisis , strains with mutations that were unlinked to the CobI operon were isolated , and the mutations were designated cobF , since all proved to affect a single locus .
These mutants were found at a frequency of 10-5 to 1O-4 ( without mutagenesis ) and lo - ( after DES treatment of cells ) .
We isolated nine mutants with mutations at this locus , and the DES-induced mutant cobFl ( strain TT14691 ) was studied in detail .
In all cases , the regulatory mutants were able to synthesize cobalamin de novo under anaerobic-growth-conditions ( data not shown ) .
Therefore none of these mutations disrupt any vitamin-B12 biosynthetic functions .
Despite higher aerobic expression of the CobI operon , none of the mutations permitted B12 biosynthesis in the presence of oxygen .
To determine whether the mutations affect any general metabolic functions , we tested ability of these mutants to grow on glucose , glycerol-fumarate , and glycerol-nitrate both aerobically and anaerobically ( data not shown ) .
The mutants belonging to classes CobRI through IV ( strains TT14696 through 9 , respectively ) were indistinguishable from the parent strain ( TT10927 ) under all conditions tested .
In contrast , cobF mutants grew about twofold slower than the wild type on glucose under anaerobic conditions , which suggests that they are impaired for fermentation ; under all other conditions tested the cobF mutants showed growth behavior similar to that of wild-type cells ( data not shown ) .
Effect of cobR and cobF mutations on CobI transcription .
Table 2 shows the effect of the cob mutations on the expression of P-galactosidase from the CobI : : lac fusion under a variety of conditions .
The cobRI mutant , which was isolated after hydroxylamine mutagenesis , showed increased aerobic expression of the CobI : : lac fusion but had very little effect on expression under anaerobic conditions .
The DES-induced mutant , cobR2 , showed its major effect when cells were grown on glycerol aerobically .
This mutant also showed a slight increase in CobI transcription anaerobically on glucose .
The cobR3 mutant , found after DES mutagenesis , showed an increase in CobI : : lac transcription when cells were grown aerobically on either glucose of glycerol .
Under anaerobic conditions , CobI : : lac expression was increased during fermentation of glucose but not during anaerobic respiration to glycerol-fumarate .
The spontaneous mutant ( cobR4 ) was the most extreme type , showing a large increase in CobI expression during aerobic-growth on either glucose or glycerol .
This mutant showed a smaller increase during fermentation of glucose .
Like the cobRI , -2 , and -3 mutants , it showed no increase in CobI : : lac transcription compared with the wild type during anaerobic-growth-on-glycerol-fumarate .
All four CobR mutants still showed transcription control in response to external cobalamin ( Table 2 ) , similar to the wild type ( 10 ) .
Furthermore , externally added cAMP stimulated CobI : : lac transcription of the mutants as well as that of the wild type ( Table 2 ) .
This suggests that the mutations affect only oxygen control and do not interfere with the other mechanisms that regulate the operon ( 10 ) .
The cobFl mutation , unlinked to the CobI operon , showed its major regulatory effect during anaerobic fermentation of glucose .
The mutation had minimal effects on expression under the other conditions tested .
The behavior of cobF mutants appears to contradict the manner in which these mutants were isolated .
The cobF mutants were identified under aerobic conditions as colonies that showed high expression of a CobI : : lac fusion , judged by X-gal hydrolysis .
Nevertheless , when cobF mutants were grown aerobically in liquid medium , 3-galactosidase assays showed very little increase compared with results with the wild type ( Table 2 ) .
The major effect of cobF on CobI expression was seen anaerobically .
The mutants were probably detected by vir-tue of the increased,-galactisidase produced by cells withi the colonies , which had limited access to oxygen and were in fact growing anaerobically .
In cobF mutants these anaerobic cells make a larger contribution to the P-galactosidase levels in the colony than they would in a cobF + colony .
To exclude the possibility of a nonspecific effect of the cobFl mutation on Lac expression , we tested the effect of the cobFl mutation on Lac fusions to the cysG , thr , trp , metE , hisD , bio , and gal genes .
The cobFl mutation had no effect on the expression of any of these operon fusions ( data not shown ) .
To look for interactions , we constructed double mutants carrying both the cobFl mutation and one of the cobR mutations ( cobRI through 4 ) .
When grown aerobically on glucose , the cobF-cobR double mutants showed 1.3-to 2-fold higher CobI expression than did the parental cobR mutant ( data not shown ) .
Thus the effects of these two mutation types are at least partially additive .
We examined the effect of the cobFl mutation on the expression of genes that are induced anaerobically and known to be controlled by the oxygen-regulatory genes oxrA,-B , and-C ( 14 , 26 ) .
( The oxrA gene of S. typhimurium is equivalent to the fnr gene of Escherichia coli .
) A pep T7 : : lac ( 14 , 26 ) fusion ( oxrA,-B , and-C dependent ) , a hyd4 : : lac ( 14 ) fusion ( oxrC dependent ) , and an flhl-S : : lac ( 14 ) fusion ( oxrC dependent ) were all unaffected by the cobFl mutation ( data not shown ) .
Finally , we found that oxrA,-B , and-C mutations had no effect on expression of a CobI : : lac fusion under either aerobic or anaerobic conditions ( data not shown ) .
Mapping of regulatory mutations .
The CobI-linked regulatory mutations were mapped with respect to a TnJO element upstream of the CobI operon ( zeb-1845 : : TnlO ) and a Mu dJ insertion within the CobI operon ( cob-24 : : Mu dJ ) .
To determine linkage between cobR mutations and the Mu dJ insertion , strains TT14676 through TT14679 were used as donors to transduce the cobRI through 4 mutations , respectively , with the Mu dJ insertion into strain TR6583 ; kanamycinresistant transductant colonies were selected ( Mu dJ ) and screened to identify recombinants that form dark blue colo-nies aerobically on plates containing X-gal and PETG .
The cotransduction frequency of the Mu dJ and cobRI through 4 mutations was widely variable ; the frequencies varied from 22 to 60 % cotransduction , suggesting that the cobR mutations map far apart or that particular alleles have a strong effect on transduction frequencies .
To determine whether the regulatory mutations were promoter proximal or distal to Mu dJ , we determined the linkage of the regulatory mutations to the TnJO located outside the promoter end of the operon .
Phage grown on strains TT14696 through TT14699 ( zeb-1845 : : TnlO cobRI through 4 ) was used to transduce strain TT10852 ( cob-24 : : Mu dJ ) , selecting for tetracycline and kanamycin resistance ( to maintain the recipient Mu dJ ) and screening for light blue ( cobR + ) and dark blue ( cobRI through 4 ) colo-nies .
The cotransduction frequency between the TnlO element and cobRI through 4 mutations in this cross was between 83 and 91 % .
The linkage between the zeb-1845 : : TnJO and cob-24 : : Mu dJ insertions ( determined in other crosses ) was 27 % , suggesting that the cobRi through 4 mutations were promotor proximal to the Mu dJ element and very close to the zeb-1845 : : TnJO insertion .
To determine the exact order of zeb-1845 : : TnJO , cobRI through 4 , and cob-24 : : Mu dJ , we performed crosses with a strain ( TT14709 ) that contains the deletion DEL1011 ( his phs ) extending from the zeb-1845 : : TnJO insertion clockwise ( away from the CobI operon ) into the his operon .
This deletion leaves the CobI operon intact .
Because of its large size , this deletion can not be repaired by a single transduced fragment .
Donor P22 phage was grown on strains with zeb-1845 : : TnlO and cob-24 : : Mu dJ and one of the cobR mutations ( strains TT14692 through TT14695 ) ; this phage was used to transduce the deletion strain to kanamycin resistance , screening for dark blue colonies .
If the regulatory mutations were on the left side of the TnJO element ( nearer the his locus ; Fig. 1 ) , they could not be coinherited with the Mu dJ insertion because there is no homology available for recombination .
If the mutations were located on the right side of the TnJO element , between TnJO and the Mu dJ insertion , they would show linkage to the Mu dJ , since recombination could occur between the TnlO insertion site and the regulatory mutations .
All four mutations showed 3 to 5 % cotransduction with Mu dJ ; the inferred gene order is therefore zeb-1845 : : TnJO , cobRI through 4 , cob-24 : : Mu dJ ( Fig. 1 ) , which suggests that the cobRI through 4 mutations are located at the promotor-proximal end of the CobI op-eron .
The cobFI mutation was initially mapped by Hfr crosses .
We first isolated a TnlOdTet element ( 9 ) that was linked to the cobFl mutation and performed Hfr mapping on this strain ( TT14700 ) by the method of Chumley et al. ( 7 ) .
These crosses suggested that the cobF mutation was located between 22 and 34 min ( data not shown ) .
To refine the mapping of the cobFl mutation , we transduced the linked xcd-3678 : : TnJOdTet into two strains ( TT10285 and TT10399 ) that carry different duplications , on extending from pyrC to pncX ( 23 to 27 min ) and the other extending from purB to pmi ( 25 to 31.5 min ) .
If the zcd-3678 : : TnlOdTet element were located in the duplicated region , transductants inheriting the TnJO element would be unstable due to the segregation of the duplication , and tetracycline-sensitive derivatives would be found at a low frequency .
Unstable Tetr transductants were found with the recipient carrying a duplication extending from pyrC to pncX , suggesting that cobFl was located between 23 and 27 min on the chromosome .
The other duplication strain ( 25 to 31.5 min ) gave stable transductants .
These results placed the cobF mutation between 23 and 25 min .
We further mapped the cobF mutation by determining cotransduction frequencies between the zcd-3678 : : TnJOdTet element and Mu dA insertions in this area .
No cotransduction was detected between this TnlO element and insertions in the put or purB genes .
Weak linkage ( 3 % cotransduction ) was seen to the pyrC2688 : : Mu dA insertion ( TT9531 ) .
The linkage either cobFl mutation showed not to pyrC or purB mutations .
From this data , we infered a gene order of pyrC , zcd-3678 : : TnlOdTet , cobF , purB ( Fig. 2 ) .
Dominance To determine whether the cobRI through tests .
4 mutations were dominant or recessive , we constructed an F ' plasmid including the CobI operon ( see Materials and Methods ) .
The F ' plasmid was then transferred ( from strain TT14688 ) into strains TT14676 through TT14679 ( cobRI through 4 cob-24 : : Mu dJ ) , and the P-galactosidase expression of these merodiploids was determined .
Expression of CobI : : lac in these strains was similar to that in the haploid mutants ( data not shown ) , indicating that the cobRI through 4 mutations are dominant to the wild type .
To determine whether the mutations are cis or trans acting , we introduced the CobI : : lac fusion into the F ' plasmid , which was mated into the cobRI through 4 strains ( TT14696 through TT14699 , respectively ) .
There was no effect of the chromosomally located cobRI through 4 mutations on the CobI : : lac fusion located in the F ' plasmid ; thus all four regulatory mutations are cis dominant .
A dominance test performed with duplications of this region gave similar results ( data not shown ) .
Effect of the cobRi through 4 and cobF regulatory mutations on CobI : : lac transcription P-Galactosidase activitya under the indicated conditionsb Aerobic Anaerobic Glycerolfumarate 820 1,100 1,066 838-720-812 Strain Genotype Glycerol ( NA ) 5 40 20 73 340 5 Glucose Glucose NA + Bl2c + cAMPd 16 26 TT10852 TT14676 TT14677 TT14678 TT14679 TT14691 CobI : : lac CobI : : lac cobRI 3 14 2 2 14 104-22-177-399 ND 22-130-310 60 CobI : : lac cobR2 CobI : : lac cobR3 CobI : : lac cobR4 CobI : : lac cobFl 4 40 270 5 2 3 5 NDe a The numbers are Miller units ( 17 ) and represent the means of three separate measurements .
b The cells were grown as described in Materials and Methods .
c Cobalamin was added to a final concentration of 1.5 x 1O-7 M. d cAMP was added to a final concentration of 5 x 10-3 M. eND , Not determined .
his phs-H -- I cobRl-4 I I ZX cob-24 : : MudJ ( Cobl ) 1 zeb-1845 : : Tnl 0 I I 83-91 % 22-60 % 27 % FIG. 1 .
Genetic mapping of the cobRI through -4 regulatory mutations .
The values represent cotransduction frequencies between the different markers .
The crosses were performed as described in Results .
Recessivity of the cobF mutations was demonstrated by using chromosomal duplications .
The cobFI mutation was introduced into duplication strain TT10285 by using the 40 % linked zcd-3677 : : TnJOdCam insertion ( TT14682 ) ( Fig. 3 ) .
Since the cobFl mutation is only 40 % cotransducible with the zcd-3677 : : TnJOdCam insertion , we examined six transductants .
None of the potential cobFJIcobF + merodiploid strains showed increased CobI : : lac expression under conditions that retained the duplication , suggesting that , if the cobFl mutation were inherited , it must be recessive to the wild-type allele .
The presence of the cobFl mutation was shown in two different ways .
When duplications were allowed to segregate ( Fig. 3 ) , three of the six strains yielded ampicillin-and tetracycline-sensitive segregants that formed dark blue colonies on X-gal medium .
We conclude that these merodiploid transductants had received the cobFl mutation .
It was also possible to recover the cobF mutation by crosses from these three transduction merodiploid strains .
are not We conclude that the cobF + lcobF1 merodiploids and therefore the constitutive for CobI : : lac expression cobFI mutation is recessive to the wild-type allele .
Effect of the cobR3 and cobR4 mutations on CobIl and CobIHI transcription .
The presence of the cobR3 mutation , which increased CobI expression 13-fold , caused about a 2-fold increase in the expression of Lac fusions to the CobII and CobIlI operons ( Table 3 ) .
Similarly , the cobR4 mutation , which increased CobI : : lac expression 90-fold , stimulated CoblI and CobIII transcription approximately 10-fold .
This result could be explained by transcriptional readthrough from the CobI operon into the CobII and CoblII operons ; alternatively the CobIl and CobIII operons could have independent promoters stimulated by a protein encoded within CobI .
We designed two different tests to examine these possibilities .
We constructed a duplication of the Cob region that carried a cobR mutation in one copy and a CobIII : : lac fusion in the other .
The duplication was arranged such that transcription from the CobR mutant control region did not proceed toward the CobIlI fusion ; thus the effects of CobR mutations on CoblIl would have to be in trans .
Under these conditions , both the cobR3 and cobR4 mutations had only a minor effect ( about twofold ) on expression of the CobIII : : lac fusion ; this small effect was seen for both cobR mutations even though cobR3 stimulated CobI expression 10-fold and cobR4 stimulated CobI expression 90-fold .
We conclude that the major effect of cobR mutations on CobIlI acts in cis .
The minor trans effect was also seen when the fusion was placed on an F'CobI plasmid and the cobR mutations were placed in the chromosome .
Cobalamin production in the wild type and cobR3 and cobR4 regulatory mutants .
The cobR3 and cobR4 mutations increased aerobic transcription levels of the CobI operon so that it approached the level attained in wild-type cells durin FIG. 1 .
Genetic mapping of the cobRI through -4 regulatory mutations .
The values represent cotransduction frequencies between the different markers .
The crosses were performed as described in Results .
zcd-3678 : : Tnl 0 cobF ` K7 I pyrC 23 ' 3 % 30 % < 1 % < 1 % FIG. 2 .
Genetic mapping of the cobFl mutation .
The values shown are cotransduction frequencies between the different genetic markers , which were determined as described in Results .
s s S cobF + Amp/Tet/Cam CobFO phenotype Segregants S S S cobF Amp/Tet/Cam CobF-phenotype S S R Amp/Tet/Cam cobF I ) C CobF-phenotype FIG. 3 .
Dominance test of the cobFI mutation .
The genetic crosses were performed as described in Results .
The thick lines represent duplicated material .
Despite these high levels of Cob operon transcription , the regulatory mutants ( like wildtype cells ) were unable to produce B12 aerobically .
Table 4 shows the levels of cobalamin determined by using a radio-active dilution assay ( see Materials and Methods ) .
The same quantitative results were obtained by using an alternative 57Co labeling assay ( data not shown ) .
The first thing to notice is the complete absence of vitamin-B12 in aerobically grown wild-type cells .
Second , a low amount of cobalamin ( about 100 molecules per cell ) was synthesized in the wild-type cells during anaerobic-growth .
Even though the cobR3 and cobR4 mutants transcribed the CobI,-II , and-III operons at high rates aerobically , they did not synthesize B12 under these conditions .
However , anaerobically the cobR3 and cobR4 mutants synthesized two-to threefold more cobalamin than the wild type .
When the vitamin-B12 precursor DMB was added anaerobically , the wild type synthesized ca. 400 molecules per cell and the cobR4 mutant synthesized almost 2,000 molecules per cell .
Effect of regulatory mutations cobR3 and cobR4 on transcription of CobII : : lac and CobIII : : lac fusions , B-Galactosidase levelb of strains carrying the indicated operon fusion : CobI : : lac CobII : : lac CobIII : : lac cobR + 3 30 20 cobR3 40 63 51 cobR4 270-377-244 a The following strains were assayed : cobR + strains TT10852 ( CobI : : lac ) , TT10857 ( CobII : : lac ) , and TT10858 ( CobIII : : lac ) ; cobR3 strains TT14678 ( CobI : : lac ) , TT14705 ( CobII : : lac ) , and TT14706 ( CobIII : : lac ) ; and cobR4 strains TT14679 ( CobI : : Iac ) , TT14707 ( CobII : : lac ) , and TT14708 ( CobIII : : 1 ac ) .
b The numbers presented are Miller units ( 17 ) and represent the means of three separate measurements .
The cells were grown aerobically in glucose minimal-medium as described in Materials and Methods .
An in-vivo test also indicated an increased production of B12 in the cobR3 and cobR4 mutants .
We tested the anaerobic-growth-rate of a metE mutant in which methionine production ( and cell growth ) was entirely dependent on the B12-dependent MetH enzyme .
Under these conditions the presence of cobR3 and cobR4 mutations ( strains TT14698 and TT14699 , respectively ) led to a significantly faster growth-rate ( 10 to 20 % ) compared with that of the wild type ( strain TT10927 ) , indicating an increased production of co-balamin in the mutants .
Externally supplied cobalamin was transported and accumulated to high levels ( approximately 5,000 molecules per cell ) ( Table 4 ) .
This high cobalamin level had no detrimental effect on cell growth either aerobically or anaerobically ( data not shown ) .
DISCUSSION We isolated regulatory mutants that showed an increase in aerobic expression of the cobinamide biosynthetic genes TABLE 4 .
Cobalamin production in the wild type and the cobR3 and cobR4 mutants Cobalamin levels in cells grown under indicated conditionsa : Aerobic Anaerobic-DMB + DMB NA + B12 + DMB TR6583 Wild type < 5 < 5 124 5,122 420 TT14681 DEL299 ( Cob - ) < 5 < 5 < 5 ND < 5 TT14698 cobR3 < 5 < 5 335 ND ND TT14699 cobR4 < 5 < 5 237 ND 1,944 a The cobalamin levels are expressed as molecules per cell and represent the means of duplicate measurements .
DMB and cobalamin were added to a concentration of 50 and 0.2 , ug/ml , respectively .
The cells were grown in minimal glucose medium .
To estimate B12 molecules per cell , B12 levels were determined for a volume of culture of known cell titer ; all B12 was shown to be cell associated .
NA , No addition ; ND , not determined P.yrC R R R cobF cobF/I `` I Amp/Tet/Cam cobFt phenotype N Product L s s S cobF + Amp/Tet/Cam CobFO phenotype Segregants S S S cobF Amp/Tet/Cam CobF-phenotype S S R Amp/Tet/Cam cobF I ) C TABLE 3 .
Effect of regulatory mutations cobR3 and cobR4 on transcription of CobII : : lac and CobIII : : lac fusions Strain Description cobR allelea ( CobI ) .
These mutations increased aerobic CobI expression from 4-to 90-fold , depending on the mutant allele and the carbon source used for growth ( Table 2 ) .
The cobRI through 4 mutations were all located at the promotor-proximal end of the CobI operon ( Fig. 1 ) .
It is unlikely that these mutations create fusions to foreign promotors , since the regulatory profile of the mutants was otherwise similar to that of the wild type ; the mutants were still subject to transcriptional regulation of the CobI operon by externally added cobalamin and cAMP , and they continued to show partial anaerobic induction of CobI transcription ( Table 2 ) .
In dominance tests , all CobI-linked mutations were dominant to the wild type and only acted on the CobI operon when located in cis .
These characteristics would be expected for mutations affecting a promotor , operator , or attenuator ; cobR mutants could even affect a strongly cis-acting regulatory protein .
The transductional linkage data and preliminary physical mapping suggest that the mutations are distributed across a large region .
The exact location and nature of these mutations are currently being determined .
The cobRI through 4 mutants are Cob ' as determined phenotypically by their ability to synthesize methionine anaerobically via the B12-dependent MetH enzyme .
For the two mutants with the highest increase in CobI : : lac expression , we measured cobalamin production under both aerobic and anaerobic-growth-conditions .
Neither the cobR3 nor the cobR4 mutant could synthesize vitamin-B12 aerobically ( Table 4 ) , even though their transcription levels approached those of the wild type grown anaerobically ( Table 2 ) .
This suggests that aerobic cobalamin biosynthesis is limited by more than transcription of the Cob operons .
It is conceivable that oxygen-sensitive intermediates or enzymes are involved in this pathway ; oxygen-sensitive intermediates have been found in-vitro in other organisms ( 2 ) .
Alternatively , there may be essential steps in B12 biosynthesis that are carried out by proteins encoded by genes outside of the Cob operons whose expression is not increased by the cobR3 and cobR4 mutations .
Anaerobically , however , the cobR3 and cobR4 mutants were able to synthesize significantly more ( two-to threefold ) cobalamin than was the wild type ( Table 4 ) .
When one of the intermediates in the pathway ( DMB ) was added , both the wild type and the cobR4 mutant showed an increase ( three-and eightfold , respectively ) in synthesis of B12 .
This result indicates that the amount of B12 synthesized anaerobically is limited by the amount of DMB synthesized by the CobIl operon .
This situation has been reported for B12 synthesis by other organisms ( 20 ) .
We determined the effect of the cobR3 and cobR4 mutations on the expression of Lac fusions to the CobIl and CobIll operons ( Table 3 ) .
Both mutations led to an increase in expression of the CobII and CobIII operon when present in cis , which is mainly the result of transcriptional readthrough from the CobI operon into CobIH and CoblIl ( Tom Doak , personal communication ) .
The small trans effect ( about twofold ) noted here could be due to a transcriptional activator , encoded in the CobI operon , which acts on the CobII and CobIlI operons .
Alternatively , increased biosynthesis of cobinamide or its precursors could have an indirect stimulatory effect on CobIl and CoblIl transcription .
One class of regulatory mutations proved to be unlinked to the CobI operon .
These mutations , designated cobF , mapped near the pyrC gene at about 24 min on the chromosome ( Fig. 2 ) .
The cobFl mutation increased CobI : : lac transcription 1.5-to 4-fold , depending on whether oxygen was present ( Table 2 ) .
Double mutants ( cobFl and cobRI , -2 , -3 , or 4 ) showed a small increase in aerobic CobI transcription , compared with that of either of the single mutants alone , which suggests that the effects of cobFl and cobRI through 4 on CobI : : lac expression are independent and additive .
The cobFl mutation seems to be specific for the CobI operon and did not increase the expression of any other genes tested .
A dominance test showed that the mutation was recessive to the wild type ( Fig. 3 ) .
These results are compatible with the cobF gene encoding a CobI-specific trans-acting repressor .
However , since cobF mutants showed impaired growth under fermentative conditions , we suspect that these mutations owe their effect to secondary consequences of abnormal metabolism .
We have isolated plasmids complementing the cobF mutation and are determining the nature of the products encoded by this plasmid .
Finally , we showed that mutations in the oxrA,-B , or-C genes ( which affect other oxygen-controlled genes in S. typhimurium ) have no effect on CobI transcription either aerobically or anaerobically .
( The oxrA gene of S. typhimu-rium is equivalent to the fnr gene of E. coli .
These results imply that expression of the cobalamin biosynthesis genes is controlled by a mechanism distinct from those identified for other genes regulated in response to oxygen .
Andersson , D. I. , and J. R. Roth .
Redox regulation of the genes for cobinamide biosynthesis in Salmonella typhimurium .
Battersby , A. R. , and L. A. Reiter .
Synthetic studies relevant to biosynthetic research on vitamin-B12 .
An approach to isobacteriochlorins via nitrones .
Casadaban , M. J. , and S. N. Cohen .
Lactose genes fused to exogenous promotors in one step using a Mu-lac bacteriophage : in-vivo probe for transcriptional Proc .
Castilho , B. A. , P. Olfson , and M. J. Casadaban .
Plasmid insertion mutagenesis and lac gene fusion with mini-Mu bacteriophage transposons .
Chan , R. K. , D. Botstein , T. Watanabe , and Y. Ogata .
Specialized transduction of tetracycline resistance by phage P22 in Salmonella typhimurium .
Properties of a high frequency transducing lysate .
Chang , G. W. , and J. T. Chang .
Evidence for the B12-dependent enzyme ethanolamine deaminase in Salmonella .
Chumley , F. G. , R. Menzel , and J. R. Roth .
Hfr formation directed by TnlO .
Davis , R. W. , D. Botstein , and J. R. Roth .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 9 .
Elliott , T. , and J. R. Roth .
Characterization of TnlOdCam : a transposition-defective TnlO specifying chloramphenicol resistance .
Escalante-Semerena , J. C. , and J. R. Roth .
Regulation of cobalamin biosynthetic operons in Salmonella typhimurium .
Frey , B. , J. McCloskey , W. Kersten , and H. Kersten .
New function of vitamin-B12 : cobamide-dependent reduction of ep-oxy queuosine in tRNAs of Escherichia coli and Salmonella typhimurium .
Hong , J. S. , and B. N. Ames .
Localized mutagenesis of any specific small region of the chromosome .
Hughes , K. T. , and J. R. Roth .
Conditionally transposi-tion-defective derivative of Mu dl ( Ap lac ) .
Jamieson , D. J. , and C. F. Higgins .
Two genetically distinct pathways for transcriptional regulation of anaerobi gene expression in Salmonella typhimurium .
Jeter , R. M. , B. M. Olivera , and J. R. Roth .
Salmonella typhimurium synthesizes cobalamin ( vitamin-B12 ) de novo under anaerobic-growth-conditions .
Jeter , R. M. , and J. R. Roth .
Cobalamin ( vitamin-B12 ) biosynthetic genes of Salmonella typhimurium .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 18 .
Morisato , D. , J. C. Way , H.-J .
Kim , and N. Kleckner .
TnlO transposase acts preferentially on nearby transposon ends in-vivo .
Mulligan , J. T. , W. Margolin , J. H. Krueger , and G. C. Walker .
Mutations affecting regulation of the methionine biosynthetic genes .
Isolation by use met-lac fusions .
Renz , P. , J. Horig , and R. Wurm .
On the biosynthesis of the 5,6-dimethylbenzimidazole moiety of vitamin-B12 , p. 317-322 .
In B. Zagalek and W. Friedrich ( ed .
Walter de Gruyter , Inc. , New York .
Roof , D. M. , and J. R. Roth .
Ethanolamine utilization in Salmonella typhimurium .
Roof , D. M. , and J. R. Roth .
Functions required for vitamin-B12-dependent ethanolamine utilization in Salmonella typhimurium .
A method for detection of phage mutants with altered transducing ability .
Shoeman , R. , T. Coleman , B. Redfield , R. C. Greene , A. A. Smith , I. Saint-Girons , N. Brot , and H. Weissbach .
Regulation of methionine synthesis in Escherichia coli : effect of metJ gene product and S-adenosyl-methionine on the in-vitro expression of the metB , metL , and metJ genes .
Shoeman , R. , B. Redfield , T. Coleman , N. Brot , H. Weissbach , R. C. Greene , A. A. Smith , I. Saint-Girons , M. M. Zakin , and G. N. Cohen .
Regulation of the methionine regulon in Escherichia coli .
Strauch , K. L. , J. B. Lenk , B. L. Gamble , and C. G. Miller .
Oxygen regulation in Salmonella typhimurium .
Taylor , R. T. , and H. Weissbach .
In D. Boyer ( ed .
Academic Press , Inc. , New York .
Toraya , T. , S. Honda , and S. Fukui .
Fermentation of 1,2-propanediol and 1,2-ethanediol by some genera of Entero-bacteriaceae , involving coenzyme B12-dependent diol dehydratase .
Vogel , H. J. , and D. M. Bonner .
Acetylornithase of Escherichia coli : partial purification , and some properties .
Way , J. C. , M. A. Davis , D. Morisato , D. E. Roberts , and N. Kleckner .
New TnJO derivatives for transposon mutagenesis and for construction of lacZ operon fusions by transposition .