Redox Regulation of the Genes for 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 cobinamide biosynthetic genes ( the CobI operon ) was induced under three different physiological conditions : anaerobiosis ( anaerobic respiration or fermentation ) , aerobic respiration at low oxygen levels , and aerobic respiration with a partial block of the electron transport chain .
After a shift to inducing conditions , there was a time lag of approximately 50 min before the onset of CobI induction .
Under conditions of anaerobic respiration , the level of CobI transcription was dependent on the nature of both the electron-donor ( carbon and energy source ) and the acceptor .
Cells grown with electron-acceptors with a lower midpoint potential showed higher CobI expression levels .
The highest level of CobI transcription observed was obtained with glycerol as the carbon source and fumarate as the electron-acceptor .
The high induction seen with glycerol was reduced by mutational blocks in the glycerol catabolic pathway , suggesting that glycerol does not serve as a gratuitous inducer but must be metabolized to stimulate CobI transcription .
In the presence of oxygen , CobI operon expression was induced 6-to 20-fold by the following : inhibition of cytochrome o oxidase with cyanide , mutational blockage of ubiquinone biosynthesis , and starvation of mutant cells for heme .
We suggest that the CobI operon is induced in response to a reducing environment within the cell and not by the absence-of-oxygen per se .
The B12 biosynthetic genes are organized into three functionally related gene clusters located at 41 min on the chromosome ( 19 , 20 ) ; all genes are transcribed counterclockwise and appear to comprise three operons .
The CobI operon encodes cobinamide biosynthetic functions , the CoblI operon encodes dimethyl benzimidazole biosynthesis , and the CobIII operon encodes functions required to join cobinamide and dimethyl benzimidazole , forming cobalamin .
Most B12 biosynthetic genes appear to be included in these three operons .
It was established by Jeter et al. ( 19 ) that de novo biosynthesis of vitamin-B12 in Salmonella typhimurium occurs only under anaerobic-growth-conditions .
Under aerobic conditions , S. typhimurium can make B12 if provided with cobinamide , suggesting that only synthesis and/or decoration of the corrinoid ring ( CobI operon functions ) are absent aerobically .
Cells can synthesize di-methyl benzimidazole and assemble B12 from cobinamide and dimethyl benzimidazole ( CoblI and CoblII operon functions , respectively ) under both aerobic and anaerobic conditions .
Only four functions in S. typhimurium are known to require cobalamin ( 8 , 14 , 33 , 34 , 40 ; R. M. Jeter , submitted for publication ) .
and used fu-Escalante-Semerena Roth ( 13 ) Lac operon sions examine transcriptional regulation of the cobalamin to biosynthetic genes .
Their studies revealed that transcription of the cobinamide biosynthesis genes ( CobI operon ) is stimulated by cyclic AMP ( cAMP ) and that the end product ( cobalamin or a derivative ) signals repression .
Furthermore , CobI transcription is strongly repressed in the presence of oxygen , suggesting the existence of an oxygen-sensing transcriptional regulatory mechanism .
Previous studies of several other oxygen-regulated genes in S. typhimurium have revealed evidence for at least three t Present address : Department of Microbiology , University of Uppsala , Biomedicum , S-75123 Uppsala , Sweden .
The oxrA gene ( 39 ) , equivalent to thefnr gene of Escherichia coli ( 9 , 25 , 28 ) , encodes a protein that positively controls a number of genes in S. typhimu-rium , including some terminal reductases ( 4 , 9 , 25 , 38 ) .
The oxrB gene product is involved in anaerobic expression of the oxrA-dependent genes ( 39 ) .
Jamieson and Higgins ( 18 ) showed that mutations in the gene for phosphoglucose isomerase ( oxrC ) prevent anaerobic expression of hydrogenases 1 and 3 .
It was established earlier ( 13 ) that repression of CobI transcription in the presence of oxygen is independent of the oxrA gene .
In the accompanying paper ( 1 ) we confirm this and show that CobI transcription is also independent of the oxrB and oxrC genes .
To elucidate control of the CobI operon in response to oxygen , we examined the regulation of expression and the kinetics of induction of a CobI : : lac operon fusion under various growth-conditions .
Results suggest that oxygen per se does not signal repression .
MATERIALS AND METHODS Bacterial strains and genetic methods .
The genotypes of all bacterial strains used are given in Table 1 .
All bacterial strains are derivatives of S. typhimurium LT2 .
A transposi-tion-defective derivative of the specialized transducing phage Mu dl ( Ampr lac cts ) of Casadaban and Cohen ( 5 ) was used .
This derivative , Mu dll734Kanr ( 6 ) , is referred to herein as Mu dJ .
One transposition-defective derivative of TnlO , TnJOdell6dell7Tetr ( 42 ) was used and is referred to as TnlOdTetr .
All transductions were performed with the highfrequency , generalized transducing phage mutant P22 HT105/1 int-201 as previously described ( 7 ) .
Transductants were purified and made phage free by streaking for single colonies on nonselective green indicator medium ( 7 ) .
Culture media , growth-conditions and assays .
Difco nutrient broth ( 0.8 % ) containing 0.5 % NaCl was used as a complex medium .
The no-carbon E medium of Vogel and Bonner ( 41 ) was used as a minimal-medium .
Solid media contained 15 g of Bacto-Agar ( Difco Laboratories ) per liter .
Carbon sources were present at 0.4 % ( final concentration ) and electron-acceptors were present at 20 mM .
Auxotrophic supplements were added at concentrations described elsewhere ( 12 ) .
The final concentrations of tetracycline were 20 , g/ml in complex media and 10 , ug/ml in minimal-medium .
To monitor B12 synthesis genetically , strains used carried a metE mutation , which eliminates the B12-independent homocysteine methyltransferase ; in such strains , thionene-inde-pendent growth depends on B12 , a cofactor of the alternative metH methyltransferase ( 23 , 36 , 40 ) .
Presence of Ado-B12 was monitored by testing growth on ethanolamine ; the enzyme ethanolamine lyase requires Ado-B12 ( 8 , 33 , 34 ) .
In aminolevulinic acid ( ALA ) starvation experiments , hemA mutant cells were grown overnight in high ( 100 ng/ml ) - ALA medium ( nonstarvation ) and transferred to low ( 5 ng/ml ) - ALA medium ( starvation ) and then grown for 6 to 10 h to achieve maximum CobI : : lac induction .
To block electron transport with cyanide , KCN was added to exponentially growing cells to a final concentration of 25 to 100 , uM , and then cells were grown for 6 to 8 h before being harvested for enzyme assays .
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 as described earlier ( 13 ) .
The addition of specified amounts of air sealed to anaerobic tubes was achieved by withdrawing with a syringe a given volume of the original gas mix ( 90 % N2-5 % H2-5 % C02 ) from the headspace above the culture and injecting the corresponding volume of air .
An-aerobic cultures were inoculated by injecting 0.05 to 0.1 ml of aerobic overnight culture into a prewarmed anaerobic culture tube ( 5 ml ) .
Cells were normally grown overnight to the stationary-phase .
Extensive control experiments showed that similar f-galactosidase activity was obtained from logarithimically growing cells and cells that had entered the stationary-phase after anaerobic-growth .
Cell density was monitored with a Lambda 4B UV/VIS spectrophotometer ( Perkin-Elmer Corp. ) .
f3-Galactosidase activity was assayed as described by Miller ( 27 ) by using CHCl3 and sodium dodecyl sulfate to permeabilize the cells .
All,-galactosidase activity values presented are Miller units .
673 a 0 a ( 0 0 time ( minutes ) 1,5-0 0 C 0c U. 0 ( D0 1,0 0co c0 1 ¬ 0,5-0 C ) 0 i 0,0 0 600 200 400 Time ( minutes ) FIG. 1 .
( a ) Kinetics of induction of a CobI : : lac operon fusion at different initial oxygen concentrations .
The values on the ordinate are Miller units of,3-galactosidase .
The initial oxygen concentrations were 0.09 % ( C1 ) , 0.48 % ( * ) , 4.8 % ( - ) , 9.5 % , ( A ) , and atmospheric oxygen ( 0 ) .
The experimental details are described in Materials and Methods .
( b ) Cell density ( A6w ) as a function of time at different oxygen concentrations .
The symbols and oxygen concentrations used are as in panel a. RESULTS Expression of the CobI : : lac fusion is induced during anaerobic-growth .
As shown previously ( 13 ) , transcription of the CobI genes increases in response to anaerobisis .
During fermentative growth-on-glucose , CobI : : lac expression increased sixfold compared with growth under aerobic conditions .
Anaerobic-growth-on-glycerol-fumarate resulted in a 170-fold increase in CobI : : lac expression , compared with that in cells grown with aerobic respiration of glycerol .
All assays are done in the absence of B12 to prevent its repressive effect .
Anaerobic respiration at low oxygen levels induces CobI transcription .
Several conditions were checked to determine whether a high level of CobI : : lac transcription could be achieved under conditions other than strictly anaerobic respiration ( Fig. 1 ) .
For example , would aerobic respiration to low oxygen levels , in the absence of other electron-acceptors , result in high expression levels ?
A CobI : : lac fusion strain ( TT10852 ) was grown in sealed tubes containing glycerol as the carbon source and various initial concentrations of oxygen as the sole electron-acceptor .
Induction of the CobI operon occurred after a time lag that depended on the amount of oxygen provided initially ( Fig .
We conclude that the lag reflects the time necessary for the cells t 0c U. 0 ( D0 1,0 0co c0 1 ¬ 0,5-0 C ) 0 i 0,0 0 600 200 400 Time ( minutes ) FIG. 1 .
( a ) Kinetics of induction of a CobI : : lac operon fusion at different initial oxygen concentrations .
The values on the ordinate are Miller units of,3-galactosidase .
The initial oxygen concentrations were 0.09 % ( C1 ) , 0.48 % ( * ) , 4.8 % ( - ) , 9.5 % , ( A ) , and atmospheric oxygen ( 0 ) .
The experimental details are described in Materials and Methods .
( b ) Cell density ( A6w ) as a function of time at different oxygen concentrations .
The symbols and oxygen concentrations used are as in panel a. 1 00-200-300 time ( minutes ) FIG. 2 .
Kinetics of CobI : : Iac induction after a shift from aerobic to anaerobic conditions .
,3-Galactosidase levels ( left ordinate ; El ) are plotted the root of the Miller unit value at time t after the as square shift minus the value at time immediately before the shift .
The zero right ordinate shows cell density ( * ) measured as A ,6 .
Strain TT10852 was grown to the log phase in aerobic glycerol-nitrate medium and purged for 2 min with N2 to remove any oxygen present , and then a 0.2-ml inoculum was transferred to 5 ml of anaerobic prewarmed glycerol-nitrate medium .
The tubes were incubated at 37 °C with shaking , samples ( 0.5 ml ) were removed at regular time intervals , and,-galactosidase activity and cell density were determined .
consume oxygen and reduce its concentration to a level that causes induction .
After induction of the operon , enzyme levels increased to a final level of approximately 400 U of P-galactosidase .
The rate of increase of P-galactosidase levels and the final level attained were independent of the initial oxygen concentration .
The question of whether induction occurs at low oxygen concentrations or only after all oxygen is consumed is difficult to answer directly .
However , observation of the growth-rate of cells suggests strongly that there is still residual oxygen present at the time of induction .
Recall that the sole carbon source , glycerol , can only be utilized by S. typhimurium in the presence of an electron-acceptor .
Since oxygen was the only electron-acceptor provided and since cell growth continued through the induction period , we conclude that some oxygen remained throughout the induction period .
These results suggest that high expression of the CobI : : lac operon does not require an alternative electron-acceptor but can occur when the oxygen level is low .
From this type of experiment , we can not assess the absolute oxygen concentration at which induction occurs , but clearly it must be lower than the lowest initial oxygen concentration given ( 0.09 % ) .
Time course of CobI : : lac transcription after oxygen deprivation .
To determine the time required for induction of CobI : : lac transcription , strain TT10852 ( cob-24 : : Mu dJ ) was grown aerobically in glycerol-nitrate medium and then shifted to completely anaerobic conditions ( Fig. 2 ) .
Induction of,-galactosidase started approximately 50 min after the shift in growth-conditions .
In this experiment the square root of the enzyme activity was plotted versus time after the shift ; this value showed an initial linear dependence on time ( as it does for other operons ) and allowed us to extrapolate to estimate the time delay before the onset of induction .
The induction time measured represents the sum of the time required for cells to consume tiny amounts of residual oxygen , signal conditions to the CobI operon , and then transcribe and translate,-galactosidase from the CobI : : lac hybrid operon .
We estimate that the time for the culture to become anaerobic was negligible in this experiment , since the only oxygen present was the traces introduced with the innoculum .
Transcription and translation of the hybrid op-eron ( 5 to 7 kilobases ) , assuming average rates , would require only a few minutes .
Therefore , most of the 50-min delay observed probably represents the time needed for synthesis or turnover of cellular components that are involved in signalling anaerobic conditions to the CobI promoter .
Note that induction did not require cell growth ; when P-galactosidase first appeared , 50 min after the shift , cell density had not increased compared with that at the time of the shift ( 0 min ) .
When induction was practically complete ( 300 min ) , the cell density had increased less than 50 % .
Partial blocks of electron transport result in aerobic induction of CobI transcription .
The experiment shown in Fig. 1 demonstrated that induction of CobI transcription can occur during aerobic respiration with low oxygen levels .
One result of a decreased-oxygen level would be a decrease in the rate of electron transport .
This suggested the possibility that the CobI operon might be induced in the presence of high oxygen levels if the electron transport chain were inhibited .
To test this idea , partial blocks were introduced into the electron transport chain in three different ways .
First , the aerobic terminal oxidase cytochrome o ( 2 , 17 , 29 ) was blocked with low concentrations of the respiratory poison KCN ( 25 to 100 , uM ) .
This concentration of cyanide does not inhibit the anaerobic terminal oxidase cytochrome d ( 2 , 17 , 29 ) .
Second , a block in ubiquinone biosynthesis was introduced by transducing a ubiF mutation ( from strain KR42 [ 43 ] ) into strain TT10852 .
Two quinones are present in cells ; ubiquinone is synthesized and used aerobically , and mena-quinone is synthesized and used preferentially under anaerobic-growth-conditions ( 4 , 17 , 32 ) .
Third , cells with a hemA mutation were starved for heme by limiting the supply of ALA , which is required for heme biosynthesis in this mutant .
All of these conditions caused increased transcription of a CobI : : lac fusion despite heavy aeration of the cultures ( Table 2 ) .
Although the data presented are for samples taken after 6 to 8 h of growth limitation , these levels reflect the highest level obtained during the course of the culture ; no transient changes in expression level were seen .
Experiments of this general type were performed earlier by Kuritz-kes et al. ( 24 ) to test regulation of the E. coli glpAB genes ( anaerobic dehydrogenase ) ; in this situation , results similar to those reported here were interpreted as suggesting a role of heme-proteins in regulation .
Finally , we tested whether uncoupling electron transport and oxidative phosphorylation would permit CobI induction in the presence of oxygen .
Cells were grown in glycerol minimal-medium aerobically , and the uncoupler 2,4-dinitro-phenol was added to the culture .
At the lowest 2,4-dinitro-phenol concentration used ( 0.1 mM ) , the growth-rate was not reduced significantly , whereas at the highest 2,4-dinitro-phenol concentration ( 2 mM ) cell growth stopped .
The cells were incubated for 6 to 8 h after the addition of 2,4-dinitrophenol and then assayed for P-galactosidase activity .
This respiratory uncoupler had no effect on aerobic CobI transcription at any concentration tested ( 0.1 to 2 mM ) .
Levels of CobI expression depend on the particular electron-acceptor present during-growth .
If a decreased rate of electron transport can stimulate CobI operon expression , on 2 0 ¬ S = 0 0 CD a 0 0 a 10 0 S So a 0 a0 la 0 I ¬ I I I I 0 400 I 0 400-a Treatment Induction ratio 0 0 a ~ ~ ~ ~ ~ ~ ) 0 m Cyanide added ' 5 98 20 Ubiquinone removedc 5 31 6 Heme removedd 5 28 6 a The values are P-galactosidase activity before and after induction is expressed as Miller units ( 27 ) under the indicated conditions .
b Strain TT10852 ( cob-24 : : Mu dJ ) was grown with vigorous aeration in glycerol minimal-medium , and KCN was added .
The cells were grown for 6 to 8 h after KCN addition and then assayed for P-galactosidase activity .
The concentrations of cyanide used decreased the growth-rate approximately 10 to 50 % .
c Strains TT14742 ( cob-24 : : Mu dJ ubiF + ) and TT14743 ( cob-24 : : Mu dJ ubiF ) were grown in glycerol minimal-medium with vigorous aeration to the midlog growth phase and then assayed for p-galactosidase activity .
The presence of the ubiF mutation decreased the growth-rate approximately 50 % o compared with that of the wild type .
d Heme starvation was achieved by inoculating strain TT14741 ( cob-24 : : Mu dJ hemA341 ) , grown overnight in glycerol minimal-medium with 100 ng of ALA per ml , into the same medium containing only 5 nm of ALA per ml .
The cells were grown for 6 to 8 h aerobically and then assayed for,-galactosidase activity .
Since the cells were starved for ALA , the cell growth-rate decreased continuously and cell density increased only two to threefold during this time .
Growth had stopped at the time cells were assayed for , B-galactosidase activity .
might expect gene expression to vary as a function of the particular electron-acceptor used .
Enteric bacteria are known to use several electron-acceptors as alternatives to oxygen ( 3 , 9 , 25 ) .
A poor electron-acceptor ( i.e. , one with a low midpoint potential ) , which supports a decreased rate of electron transport , might be expected to lead to an increased CobI transcription .
Escalante-Semerena and Roth ( 13 ) showed previously that , when glycerol is the carbon source , the nature of the electron-acceptor does affect CobI transcription .
Their data show a surprisingly linear relationship between CobI transcription and the midpoint potential of the electron-acceptor used ( Fig. 3 ; the results presented are for cells growing on glycerol [ 13 ] ) .
The same general correlation was also seen for other carbon sources ( compare Fig. 4 and 5 ) .
( L-L 300 ¬ 0 X c - > n 0 < 0 o 0DJm ( D c C ) -0 j ' 0 o 0 m U ) 0a 200-0 0 Z - < 0 100 ¬ Levels of Cobl expression depend on the electron-donor ( carbon source ) present during anaerobic respiration .
Figure 4 shows CobI expression levels when cells were grown anaerobically on various carbon sources with nitrate as the electron-acceptor .
With glycerol as the carbon source , 480 U of,-galactosidase was produced .
Growth on all other carbon sources resulted in considerably lower levels of CobI expression .
Our initial suspicion was that these results were due to a cAMP effect , since it had previously been shown ( 13 ) that CobI transcription was stimulated by cAMP .
To check this possibility , a strain was constructed that contains an adenyl cyclase mutation ( cya : : TnlO ) and a crp * mutation .
The crp * mutation alters the Crp protein such that it activates gene expression even without cAMP ( 15 , 30 , 31 ) ; the cya mutation prevents cAMP synthesis .
Strain TT11297 ( cob-24 : : Mu dJ cya : : TnJO crp * ) showed the same behavior ( data not shown ) as strain TT10852 ( cob-24 : : Mu dJ ) ; thus the carbon source dependence of anaerobic CobI transcription is not mediated by cAMP .
The preceding experiments suggested that CobI induction might be signaled in a sense by accumulation of electrons ( or reduced cellular components ) .
Maximum induction was obtained with a good donor and a poor acceptor ; minimum induction was obtained with a poor donor and a good acceptor .
This notion leads to the prediction that higher expression would be seen on any carbon source if a poorer electron-acceptor was used .
Figure 5 shows CobI operon expression levels obtained with three different electron-acceptors and four different carbon sources .
As predicted , for all four of the carbon sources , expression of a CobI : : lac fusion increased as poor electron-acceptors were used ( Fig. 5 ) .
Possibility of a special role for glycerol .
Despite the relationships described above , the CobI operon was consistently in the of 4 induced to higher levels presence glycerol ( Fig .
and 5 ) than with the other carbon sources tested .
Does mean that glycerol is a good electron-donor with effective transport and oxidation , or might glycerol play some special role in induction ?
One argument against a special role of the pathway is the much poorer induction with glycerol phosphat carbon source FIG. 5 .
Dependence of CobI : : lac transcription on the electron-acceptor and donor present during-growth .
The electron-acceptors were oxygen ( A ) , dimethyl sulfoxide ( 0 ) , and fumarate ( El ) .
The cells were grown aerobically with the respective electron-donor and acceptor .
Samples ( 0.2 ml ) were inoculated into 5-ml anaerobic tubes containing the same medium and incubated for 6 to 8 h before 3-galactosidase activity determined .
was 0 Z - < 0 100 ¬ 0 A B F G C D E carbon source FIG. 4 .
Dependence of CobI : : lac transcription on the electron-donor present during-growth .
The cells were grown overnight aerobically with nitrate as the electron-acceptor and the indicated electron-donor .
Samples ( 0.1 ml ) were inoculated into 5-ml anaerobic tubes , and the cultures were grown overnight in the same medium and then assayed for 1-galactosidase activity .
0 0 t0 coD 0 ) a 9 0000 a 0 ) CuIM 1000 midpoint potential ( mV ) FIG. 3 .
CobI : : lac transcription as a function of the midpoint potential of the electron-acceptor present during-growth .
Strain TT10852 was grown overnight in glycerol minimal-medium containing different electron-acceptors .
Samples ( 0.1 ml ) were inoculated into anaerobic tubes containing 5 ml of the same medium .
These cultures were grown overnight at 37 °C .
Cultures grown with electron-acceptors other than oxygen were made anoxic as described previously ( 13 ) .
DMSO , Dimethyl sulfoxide ; TMAO , trimethyl-amine oxide .
These two carbon metabolized the sources are via same pathway , and both the glycolysis pathway via enter dihydroxy acetone phosphate ( 26 ) .
We suspect that oxidation of glycerol phosphate might be slower simply because of a reduced rate of transport .
It seemed possible , however , that glycerol might act as a gratuitous inducer of CobI transcription .
Therefore we tested whether good induction requires metabolism of glycerol by introducing mutational blocks various in the at steps glycerol catabolic pathway ( 10 , 11 , 16 , 21 , 22 , 26 , 44 ) .
Strains defective in glycerol kinase ( glpK strain TT14736 ) ; aerobic glycerol-phosphate dehydrogenase ( glpD strain TT14738 ) , and anaerobic glycerol-phosphate dehydrogenase ( gipA strain TT14737 ) were assayed .
Since the glpK and glpD mutants are unable to grow on glycerol aerobically , these experiments performed shift experiments .
Cells were as aerobically minimal glucose medium and were grown on shifted to anaerobic glycerol minimal-medium ( Table 3 ) .
Even though no growth could occur , because of the absence of electron-acceptor , wild-type cells ( strain TT10852 ) an showed strong induction of CobI transcription ; this induction reduced by single in either was not blocks of the redundant dehydrogenases ( aerobic glpD strain TT14738 or anaerobic glpA strain TT14737 ) .
When both dehydrogenases were removed ( strain TT14739 ) , CobI : : lac induction was inefficient .
These results suggested that the high levels very TABLE 3 .
Effect of glp mutations on anaerobic CobI : : Iac induction in the presence of glycerol Strain Descripti.o3 acstiividtyaase induct n - Galacto Foldion TT14735 Wild type 241 48 TT14736 glpK 65 13 TT14737 glpA 273 55 TT14738 glpD 213 43 TT14739 glpAD 18 4 a The vilues shown are Miller units ( 27 ) .
All strains carry a Mu dJ insertion in the CobI operon , which forms a CobI : : lac operon fusion .
The cells were grown aerobically in glucose minimal-medium and a 0.2-ml inoculum was transferred to 5 ml of anaerobic glycerol minimal-medium and incubated for 6 to 8 h before P-galactosidase activity was determined .
The , B-galactosidase activity in wild-type cells grown aerobically on glucose was 5 U. of anaerobic induction of CobI in the presence of glycerol require conversion of this carbon source at least to dihydroxy acetone phosphate and that glycerol per se does not act as a gratuitous inducer .
Consistent with this , strains carrying a glycerol kinase mutation ( glpK strain TT14736 ) also showed lower inducibility ( Table 3 ) .
These results led us to conclude that metabolism of glycerol is required for it to exert its strong effect on B12 and that the magnitude of the glycerol effect on CobI operon expression is due to the high rate of glycerol transport and oxidation .
l200 ¬ DISCUSSION Our results show that high CobI transcription levels can be achieved under three conditions : anaerobic respiration , aerobic respiration at low oxygen levels , and aerobic respiration with high oxygen and a partial blockage of the electron transport chain .
These results suggest that molecular oxygen per se does not signal repression of CobI transcription ; regulatory mechanisms that involve direct interaction of molecular oxygen with a regulatory protein are unlikely .
Three methods of electron served to impairing transport stimulate CobI expression : ( i ) lowering the concentration of a good electron-acceptor , oxygen ( Fig. 1 ) ; ( ii ) inhibiting the an aerobic electron chain and transport ( Table 2 ) ; ( iii ) using alternative electron-acceptor that was poorer than oxygen ( Fig. 3 and 5 ) .
These partial blocks of electron transport result in in the redox state of a of probably changes variety both membrane bound molecules , ( cytochromes , quinones ) Such and cytoplasmatic ( NAD , FAD , glutathione , etc. ) .
molecules could modulate CobI transcription in a manner dependent on their state of oxidation-reduction , or a regulatory protein could sense the redox state more directly .
Consistent with this idea is the observation that various different electron-donors ( carbon sources ) support widely levels of CobI transcription anaerobically , for example , under conditions of anaerobic This is not respiration ( Fig. 4 ) .
in on the the result of differences cAMP levels during-growth different carbon sources , since the same of pattern expression was seen in a strain unable to synthesize cAMP .
Mutational blocks in the glycerol-metabolizing pathway levels induction ( Table 3 ) partly prevent the high of CobI seen in the of presence glycerol .
This suggests that metabo-lism ( probably oxidation ) of glycerol is a prerequisite for stimulation of induction .
It seems unlikely that acts glycerol as a gratuitous inducer .
The fact that glycerol allows better induction than glycerol phosphate may simply reflect the fact that glycerol is transported more rapidly ( 26 ) and therefore is a quantitatively better electron-donor .
The carbon source dependence of induction thus may reflect differences in the cell 's ability to transport and metabolize these sources of reducing equivalents .
Conceivably different carbon sources can donate electrons at different rates to NAD or FAD and affect the redox state of these nucleotide pools as well as those of other redox-sensitive components of the cell .
A carbon source that is transported efficiently and donates electrons at a high rate can result in a more reduced cellular interior , which , directly or indirectly , leads to higher rates of CobI operon transcription .
The time lag for CobI induction after a shift to anaerobic conditions was unexpectedly long , approximately 50 min ( Fig. 2 ) .
Similar kinetic experiments performed with nitrate reductase ( 35 , 37 ) suggest a much shorter time lag of about 5 min .
The longer time delay for CobI induction after a shift suggests that CobI induction requires significant changes in the physiology of the cell , even though no increase in cel mass is necessary .
The 10-fold difference in time lag between CobI and nitrate reductase induction also implies that the regulatory mechanism is different in these two cases .
This is further corroborated by the fact that the oxrA gene , which is needed for expression of nitrate reductase ( 38 ) , is not required for CobI expression ( 13 ) .
Andersson , D. I. , and J. R. Roth .
Mutations affecting regulation of cobinamide biosynthesis in Salmonella typhimu-rium .
Anraku , Y. , and R. B. Gennis .
The aerobic respiratory chains of Escherichia coli .
Bilous , P. T. , and J. H. Weiner .
Dimethyl sulfoxide reductase activity by anaerobically grown Escherichia coli HB101 .
Bishop , D. H. L. , K. P. Pandya , and H. K. King .
Ubiquinone and vitamin K in bacteria .
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 control sequences .
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 .
Chippaux , M. , D. Giudici , A. Abou-Jaoude , F. Casse , and M. C. Pascal .
A mutation leading to the total lack of nitrate reductase activity in Escherichia coli K12 .
Cozzarelli , N. R. , W. B. Feedberg , and E. C. C. Lin .
Genetic control of the L-a-glycerophosphate system in Esche-richia coli .
Cozzarelli , N. R. , J. P. Koch , S. Hayashi , and E. C. C. Lin .
Growth stasis by accumulated L-a-glycerophosphate in Esche-richia coli .
Davis , R. W. , D. Botstein , and J. R. Roth .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 13 .
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 .
Garges , S. , and S. Adhya .
Sites of allosteric shift in the structure of the cyclic-AMP-receptor-protein .
Hayashi , S.-I. , and E. C. C. Lin .
Production inhibition of glycerol kinase in Escherichia coli .
Ingledew , W. J. , and R. K. Poole .
The respiratory chains of Escherichia coli .
Jamieson , D. J. , and C. F. Higgins .
Two genetically distinct pathways for transcriptional regulation of anaerobic 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 .
Kistler , W. S. , and E. C. C. Lin .
Anaerobic L-a-glycero-phosphate dehydrogenase of Escherichia coli : its genetic locus and its physiological role .
Koch , J. P. , S.-I .
Hayashi , and E. C. C. Lin .
The control of dissimilation of glycerol and L-Qx-glycerophosphate in Esche-richia coli .
Kung , H.-F. , C. Spears , R. C. Greene , and H. Weissbach .
Regulation of the terminal reactions in methionine biosynthesis by vitamin-B12 and methionine .
Kuritzkes , D. R. , X.-Y .
Zhang , and E. C. C. Lin .
Use of F ( glp-lac ) in Studies of respiratory regulation of the Escherichia coli anaerobic sn-glycerol-3-phosphate dehydrogenase genes ( glpAB ) .
Lambden , P. R. , and J. R. Guest .
Mutants of Escherichia coli K12 unable to use fumarate as an anaerobic electron-acceptor .
Lin , E. C. C. 1976 .
Glycerol dissimilation and its regulation in bacteria .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 28 .
Newman , B. M. , and J. A. Cole .
The chromosomal location and pleiotropic effects of mutations of the nirA + gene of Escherichia coli K12 : the essential role of nirA + in nitrite reduction and in other redox reactions .
A structur-ally and functionally diverse group of electron-transfer proteins .
Postma , P. W. , and J. W. Lengeler .
Phosphoenolpyruvate : carbohydrate phosphotransferase system of bacteria .
Postma , P. W. , and B. J. Scholte .
Regulation of sugar transport in Salmonella typhimurium , p. 249-257 .
In E. Qua-glioriello , F. Palmieri , S. Papa , and M. Klinkenberg ( ed .
) , Function and molecular aspects of biomembrane transport .
Elsevier Biomedical Press , Amsterdam .
The organization of the quinone pool .
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 .
Ruiz-Herrera , J. , and I. Salas-Vargas .
Regulation of nitrate reductase at the transcriptional and translational levels in Escherichia coli .
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 .
Showe , M. K. , and J. A. DeMoss .
Localization and regulation of synthesis of nitrate reductase in Escherichia coli .
Requirement of Fnr and NarL functions for nitrate reductase expression 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 .
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 .
Zak , V. L. , and A. R. Kelln .
A Salmonella typhimurium mutant dependent upon carbamyl aspartate for resistance to 5-fluorouracil is specifically affected in ubiquinone biosynthesis .
Zwaig , N. , and E. C. C. Lin .
A method for isolating mutants resistant to catabolite-repression .