170 , No. 2 Novel Regulatory Loci Controlling Oxygen - and pH-Regulated Expression in Salmonella typhimurium ZARRINTAJ ALIABADI ,1 YONG KEUN PARK ,2 t JOAN L. SLONCZEWSKI ,3 AND JOHN W. FOSTER ' * Department of Microbiology , College of Medicine , University of South Alabama , Mobile , Alabama 36688,1 Department of Microbiology , Marshall University School of Medicine , Huntington , West Virginia 25704,2 and Department of Biology , Kenyon College , Kenyon , Ohio 430223 Three new loci were discovered , each of which participates in the regulation of anaerobic gene expression .
The regulatory gene earA negatively regulates the expression of the anaerobiosis-inducible gene aniG as well as that of at least three other genes , as determined by two-dimensional polyacrylamide gel electrophoresis .
The earA locus maps at 86 min .
The expression of aniG was also shown to be controlled by changes in external pH under aerobic and anaerobic conditions .
Maximal expression was observed under anaerobic conditions at an external pH of 6.0 .
Significant transcriptional activity was also observed under aerobic conditions at pH 6.0 .
This was in contrast to hyd , whose expression was dependent upon anaerobiosis and varied with external pH. The pH dependence disappeared under fully aerobic conditions .
Mutations in earA had no effect upon hyd expression .
The two other regulators identified were oxrF , which controls aniH , and oxrG , which , in concert with oxrA and oxrB , controls aniC and anil .
The oxrG locus was mapped to 88 min and appears to code for a positive regulator .
Various oxr mutants were subjected to two-dimensional polyacrylamide electrophoretic analysis of anaerobiosis-inducible proteins .
Several pathways of anaerobic control were observed by means of these techniques .
There have been several reports identifying genes in Salmonella typhimurium and Escherichia coli whose expressions are controlled by the presence or absence of molecular oxygen ( 1 , 10 , 35 , 37 ) .
Anaerobiosis-inducible genes with unidentified products have been referred to as ani ( 1 ) and oxd ( 35 ) .
These genes were discovered by generating random lacZ operon fusions in the chromosome with Mu d ( Ap lac ) phage .
Identified anaerobiosis-inducible genes include those involved with nitrate ( nar ; 4 , 14 , 34 ) and fumarate ( frd ; 21 , 22 ) reductases , hydrogen sulfide production ( phs ; 1 , 11 ) , cobalamine biosynthesis ( cob ; 13 ) , hydrogenase ( hyd ; 1 , 20 , 39 ) , some peptidases and peptide permeases ( tppB , pepT ; 15 , 18 , 19 , 35 ) , anaerobic L-a-glycerophosphate dehydrogenase ( glpAB ; 23 ) , NAD biosynthesis ( nadAB ; 17 ) , degradative threonine dehydratase ( 28 ) , formate hydrogen lyase ( fhl ; 31 ) , a tertiary amine oxidase ( torA ; 24 , 30 ) , menaquinone synthesis ( men ; 24 ) , and dimethyl sulfoxide reductase ( 23 ) .
Several genes have been implicated in S. typhimurium as participating in anaerobic transcriptional control .
These include oxrA , a positive regulator equivalent to fnr in E. coli ( 7 , 25 , 35 ) ; oxrB , another positive regulator ( 35 ) ; oxrC ( pgi ) , not a transcriptional regulator but involved with the generation of an inducing signal ( 19 ) ; and tppR , a positive regulator of tppB ( 18 ) .
The most characterized anaerobic regulatory locus is fnr of E. coli .
Mutations in fnr prevent the anaerobic induction of several respiratory enzymes , such as the nitrate and fumarate reductases , hydrogenase 2 , dimethyl sulfoxide reductase , and glycerol-3-phosphate dehydrogenase ( 3 , 7 , 23 , 25 , 34 ) .
The nucleotide sequence of fnr has revealed a considerable amount of homology with the cyclic-AMP-receptor-protein ( CRP ) .
This suggests that a nucleotide such t Present address : Department of Biology , Han Nam University , Daejon , South Korea .
as cyclic AMP may be involved in the regulation of anaerobic gene expression by binding to Fnr .
Not all anaerobically inducible genes are fnr + dependent , however .
Various studies have shown that tppB , torA , pepN , certain components of fhil , the gene encoding hydrogenase 3 , and phs do not require Fnr for their expressions ( 1 , 11 , 18 , 19 , 30 , 35 ) .
The molecular mechanism ( s ) controlling the expressions of these genes remains obscure .
In a previous study we characterized 16 ani-lacZ operon fusions into three classes on the basis of their control by oxrA and oxrB .
Class I loci were regulated by both oxrA and oxrB , class II loci were controlled by oxrA only , and class III loci were independent of both oxrA and oxrB .
We have also noted other regulatory subclasses including ( i ) genes whose anaerobic in minimal-medium expression requires vitaminfree Casamino Acids , ( ii ) some Casamino Acids-inducible strains whose expression is enhanced by the addition of 20 mM nitrate , ( iii ) genes whose anaerobic expression is repressed by glucose ( catabolite-repression ) , ( iv ) genes whose anaerobic expression increases in complex medium with the addition of glucose , and ( v ) ani loci , whose expression is repressed by the addition of 20 mM nitrate .
We have utilized three ani loci that exhibit different regulatory characteristics to identify three new regulatory loci .
The first is earA , which acts as a negative regulator of aniG , a class III gene whose anaerobic expression requires complex medium and is glucose repressed .
The second is oxrF , which controls aniH , a class III locus whose expression is oxrAB independent .
The third regulator is oxrG , which controls aniC and aniI , both group I loci whose anaerobic expressions require Casamino Acids and are enhanced by nitrate .
It was also determined that the hyd and aniG loci are regulated by alterations in external pH as well as by anaerobiosis .
The earA ( external acidification regulator ) mutations relinquished aniG from pH control but not oxygen regulation .
To our knowledge this is the first report of a gene involved in regulating the transcriptional response of enterobacteria to changes in pH. 84 MATERIALS AND METHODS Bacterial strains and cultural conditions .
The bacterial strains used in this study are listed in Table 1 .
Cultural conditions , including minimal E medium and complete LB-medium , were as described previously ( 12 , 27 , 36 ) .
Nocarbon E medium was supplemented with 10 mM MgSO4 and lactose at 0.2 % ( NCE-lactose medium ) ( 12 ) .
For labeling with H235SO4 , the minimal MOPS ( morpholinepropanesulfonic acid ) medium of Neidhardt et al. ( 29 ) containing 28 , uM S04 was used .
Antibiotic concentrations were 30 , ug/ml for ampicillin , 100 , ug/ml for kanamycin , and 10 , ug/ml ( minimal-medium ) or 20 , ug/ml ( complex medium ) for tetracycline .
The pH of liquid media was maintained by including buffers with the appropriate pKa values as described by Slonc-zewski et al. ( 32 ) .
The buffers ( at 100 mM ) included citrate ( pH 5.0 to 5.3 ) , MES [ 2 - ( N-morpholino ) ethanesulfonic acid ] ( pH 5.7 to 6.3 ) , PIPES [ piperazine-N , N ' - bis ( 2-ethanesul-fonic acid ) ] ( pH 6.7 to 7.3 ) , or MOPS ( pH 7.7 to 8.3 ) .
Anaerobiosis was achieved by overlaying a tube of broth with 1 ml of sterile paraffin oil .
Plates were incubated in an anaerobic GasPak system ( BBL Microbiology Systems ) .
Aerobic conditions were achieved with vigorous shaking ( 220 rpm ) of 5 ml of culture medium in a 50-ml flask .
MacConkey nitrate medium was prepared as recommended ( 2 ) .
Transductions employing bacteriophage P22 HT 105/1-int were performed as indicated earlier ( 16 ) .
After transductions , nonlysogenic segregants were identified by cross-streaking against phage H5 on the green medium of Chan et al. ( 6 ) .
H5 is a virulent , clear-plaque-forming derivative of P22 which will not infect strains lysogenic for P22 .
Many of our Mu dl and Mu dA lysogens were replaced by the smaller ( 11.3-kilobase ) Mu dJ derivative , which carries kanamycin resistance .
Construction and use of Mu dJ ( originally referred to as Mu dI1734 ) was described by Castilho et al. ( 5 ) .
Isolation of regulatory mutants .
Two basic methodologies were employed to isolate regulatory mutations .
The first method involved plating 107 cells of an appropriate lacZ fusion strain per plate onto NCE-lactose medium with incubation under repressive conditions , i.e. , aerobically for ani-lacZ strains .
The method screens for spontaneous regulatory mutations which allow for constitutively high or deregulated expression of the lacZ fusion , resulting ingalactosidase levels suitable to support growth on lactose as a sole carbon source .
The second method used to isolate regulatory mutants involved screening a random pool of TnlO insertions for insertions into regulatory loci ( 12 ) .
P22 HT 105/1-int phage propagated on the TnlO insertion pool was crossed with the ani-lacZ recipient on MacConkey tetracycline medium , and duplicate plates were incubated aerobically ( looking for constitutive expression ) and anaerobically ( looking for a lack of anaerobic induction ) .
Verification of strains containing potential oxr : : TnlO insertions was achieved by showing that subsequent transduction of the TnlO to the original ani-lacZ strain resulted in a 100 % linkage between Tcr and the deregulated Oxr-phenotype .
The P-galactosidase assay performed on these strains was described by Miller ( 27 ) .
Enzyme activity is expressed in Miller units , i.e. , nanomoles of o-nitrophenyl-3-D-galactosidase per minute per unit of optical density at 600 nm ( A6 .
For P-galactosidase assays , cells were grown under aerobic or anaerobic conditions to an optical density at 600 nm of 0.4 to 0.5 before the assay .
Values represent an average of duplicate measurements .
TnWO insertions near regulatory loci were identified by transducing ani-lacZ oxr mutants with P22 lysates propagated on a pool of TnWO insertion mutants with selection under aerobic conditions on MacConkey tetracycline medium .
Light-colored transductants that retained the appropriate Mu d antibiotic resistance marker were purified and backcrossed with the original mutant strain to determine linkage to the oxr locus .
Deletions extending from a TnWO insertion through an oxr locus were constructed by using the method of Maloy and Nunn ( 26 ) .
TnJO-directed Hfr donors were constructed as originally outlined by Chumley et al. ( 9 ) and used in mapping as described previously ( 1 ) .
The procedure for the two-dimensional analysis of cellular proteins was essentially as described previously ( 33 ) .
Cells were grown in minimal MOPS with vigorous aeration to 2 x 108 cells per ml .
The culture was then split ; 5 ml was left under aeration , and 5 ml was transferred to a culture tube and overlaid with sterile paraffin oil .
Cells were labeled at the point of shift with 50 , Ci of H235S04 and allowed to grow for one doubling .
Approximately 1.5 ml of labeled cells was suspended in 13 , ul of a sodium dodecyl sulfate ( SDS ) lysing solution , boiled , and run in a pH 5 to 7 isoelectric focusing system followed by 11.5 % SDS-polyacrylamide electrophoresis as described earlier ( 33 ) .
Comparisons were made between aerobic and anaerobic samples with equivalent protein ( 5 to 15 , g ) and equivalent disintegrations per minute ( 1 x 106 to 3 x 106 ) .
The coordinates given are those provided in a previously published standard two-dimensional profile of S. typhimu-rium polypeptides ( 33 ) .
S. typhimurium strains used Genotype Source Strain ( 1 ) ( 1 ) ( 1 ) aniC052 : : Mu dA supD10 aniAlOOS : : Mu dA supD10 aniG1072 : : Mu dA supD10 aniD1047 : : Mu dA oxrB8 zxx-895 : : TnS aniHJ069 : : Mu dA aniIO070 : : Mu dA aniG1072 : : Mu dJ supD10 aniGl072 : : Mu dJ aniG + earA209 metC1945 : : TnlO aniG1072 : : Mu dJ zig-1935 : : TnlO ( TnlO 86 % linked to earA + ) oxrAl leu485 pepT : : Mu dl zda-888 : : TnlO oxrC : : TnS JF900 JF927 JF1105 JF1114 JF1140 JF1141 JF1294 JF1295 JF1411 JF1493 RESULTS Control of hyd and aniG loci by external pH. Growth under anaerobic conditions in unbuffered medium such as LB will result in a drop in external pH levels from 7.4 to approximately 6 .
To determine whether any of our anaerobiosis-inducible fusions , especially those only induced in LB , were regulated by pH , our ani-lacZ strains were grown in LB-medium buffered as described in Materials and Methods .
Two of our fusions , aniA ( hyd ) and aniG , clearly exhibited a pH effect ( Fig. 1 ) .
The hyd-lacZ strain , JF927 , expressed P-galactosidase only under anaerobic conditions , but wa strongly influenced by the pH of the growth medium ( Fig. 1A ) .
Maximum induction occurred between pH 5.7 and 6.3 .
Two other independently isolated hyd : : Mu dJ fusions also exhibited this phenomenon .
Regulation of aniG was more complex ( Fig. 1B ) .
While induction was best under anaerobic conditions , significant induction was also observed under aerobic conditions when the external pH of the growth medium was adjusted to pH 6.0 .
Thus , hyd and aniG are controlled by at least two environmental conditions , oxygen and pH. However , each manifests this control in different ways and via distinct mechanisms .
For example , the earA mutations described below eliminate pH control for aniG but have no effect upon hyd .
Isolation of earA regulatory mutants .
The aniG locus is complex in its regulation .
It requires a complex medium for induction , is anaerobiosis inducible , and , as described above , is regulated by the external acidification of growth medium .
As described in Materials and Methods , regulatory mutations affecting aniG were isolated by plating the lacZ operon fusion strains under aerobic conditions on NCE medium containing lactose .
Mutations resulting in the increased expression of the ani-lacZ-fusions would allow the mutant cells to grow under the normally repressive condi tions if these mutations eliminated the complex medium requirement as well as either pH or oxygen control .
Regulatory mutations affecting aniG were readily obtained and are listed in Table 2 along with the,-galactosidase levels obtained under aerobic and anaerobic conditions .
The assay conditions actually combined all components of aniG expression , since under anaerobiosis the pH of LB during-growth drops to 6.3 to 6.4 .
To determine whether the mutations were linked to aniG , possibly representing promoter or operator lesions , the aniG : : Mu dJ from each regulatory mutant was transduced into S. typhimurium LT2 , and Kmr transductants were subsequently scored for oxygen regulation .
Three strains containing mutations linked to aniG ( presumably promoter or operator mutations ) were saved ( JF1357 , JF1358 , and JF1519 ) .
The remaining mutations were not linked to aniG and possibly represent lesions in one or more trans-acting regulatory loci .
These mutations , while allowing a large increase in aniG expression , did not eliminate oxygen control .
For reasons discussed below , this class of mutations affects the external acidification regulation of aniG and is referred to as ear mutations .
The Mu dJ insertion into aniG was removed from an ear mutant via cotransduction with a TnJO insertion near aniG + ( JF1494 ) selecting for Km ' Tcr colonies .
Reintroduction of aniG : : Mu dJ as well as other ani-lacZ-fusions into one such aniG + earA strain ( JF1411 ) was used to determine the effect of earA on other ani loci in our collection .
Only aniG proved to be dependent upon earA for its regulation .
The earA product : negative regulator of aniG affecting pH control .
Strains containing TnlO insertions located near earA were isolated as described in Materials and Methods .
One aniG : : Mu dJ strain ( JF1493 ) containing a TnlO linked 86 % to earA + was used to generate deletions through this regulator .
After selection for fusaric acid resistance ( FA ' ) and Tcs , deletion mutants were screened on MacConkey medium under aerobic and anaerobic conditions .
Many of the FAr Tc ' mutants constitutively expressed aniG-lacZ at derepressed levels , indicating earA is a negative regulator .
The TnJO insertion mutant was also used to map the earA gene on the S. typhimurium chromosome by constructing Hfr donors with origins of transfer at the point of insertion .
These studies place earA at 86 min .
The earA mutations allowed almost constitutive expression of aniG in minimal-medium , where aniG is not normally expressed due to the lack of a complex medium requirement , glucose repression , and alkaline-pH repression .
We found that aniG-lac expression in earA mutants , including earA deletions , consistently remained oxygen repressible by two-to threefold in minimal a aniH : : Mu dA strains were grown in minimal-medium , and aniC : : Mu dA and anil : : Mu dA strains were grown in LB .
Units and growth are as described in Table 2 , footnote a. medium or in complex medium buffered to either pH 6.3 or ( 1 ) ( 1 ) This paper This paper This paper This paper SF165 SF242 C. Miller ( 35 ) D. Jamieson ( 19 ) J. BACTERIOL .
500 a2 .0 I5 8 I. 5 5.3 5.7 6 6.3 67 7 7.3 7.7 8 6.3 PH 1.8-1.7 ¬ 1.6-1.5 ¬ 1.4-1.3 ¬ 1.2 1.1-I1-0.9-0.6 ¬ 0.7-0.6 ¬ 0.5-0.4 ¬ 0.3-0.2 ¬ 0.1-0 ' i 5 B X t 5.3 5.7 6 6 .
Transcriptional control of aniA ( hyd ) and aniG by external pH. Cells were grown to equivalent densities under aerobic ( open symbols ) or anaerobic ( closed symbols ) conditions in LB buffered to the pH indicated .
Samples were washed and suspended in assay buffer to avoid pH effects on the enzymatic activity .
( A ) JF927 [ aniA ( hyd ) - lacZJ ; ( B ) JF1295 ( aniG-lacZ ) .
P-Galactosidase activities are in Miller units as described ( 8 ) .
Regulatory mutations affecting the expression of aniG 1-Galactosidase activitya % Linkage tob : zig-1935 Strain Pertinent genotype aniG Aerobic Anaerobic JF1294 aniG : : Mu dJ 29 760 JF1299 aniG : : Mu dJ earA201 2,910 5,200 62 JF1300 aniG : : Mu dJ earA202 3,970 4,960 55 JF1301 aniG : : Mu dJ earA203 3,520 6,940 69 JF1302 aniG : : Mu dJ earA204 2,530 4,100 90 JF1303 aniG : : Mu dJ earA205 2,760 5,913 90 JF1304 aniG : : Mu dJ earA206 2,410 6,390 0 72 JF1295 aniG : : Mu dJ 75 1,170 JF1355 aniG : : Mu dJ earA207 2,010 4,540 0 73 JF1356 aniG : : Mu dJ earA209 2,370 5,960 0 86 JF1357 aniG : : Mu dJ aniGP ° 210c 1,650 3,140 97 0 JF1358 aniG : : Mu dJ aniGPK212C 2,080 4,310 97 0 JF1518 aniG : : Mu dJ earA208 1,950 3,740 0 92 JF1519 aniG : : Mu dJ aniGP ° 211c 1,680 3,270 98 0 a , BGalactosidase activity is measured in Miller units ( 27 ) ( see the text ) .
Cells were grown to equivalent optical densities under aerobic and anaerobic conditions as described in the text .
b Cotransduction frequencies were determined using a minimum of 100 scored Tcr transductants per cross .
c Superscript PO indicates the mutation probably resides in the promoter-operator region of aniG .
Effects of an earA deletion on the pH , oxygen , and complex medium dependence of aniG-lacZ expression Strain Medium9 Genotype lacto n f eaerobic fold increase 17 1 2.5 35 2.8 6.8 sAn '' a d i s a G ¬ Anaerobic Aerobic JF1295 aniG-lacZ 926 13-52-603-109 691 53 10 1 .
LB , pH 6.3 21 17 39 102 JF1506 aniG-lacZ AearA 1,702 1,784 3,350 4,666 2,615 5,203 3,616 7,100 8,630 6,814 3.0 7 .
LB , pH 6.3 a Glycerol was added to 40 m b Miller units ( see the text ) M and potassium nitrate was added to 20 mM where indicated .
However , pH control was eliminated ( Table ' 3 ) .
The 13-galactosidase values provided in Table 3 illustrate several points .
First , the anaerobiosis-inducible level for an earA + strain ( JF1295 ) in LB reflects both oxygen and pH controls ( line 1 ) .
The values agree nicely with those found in the LB , pH 7.9 aerobic culture ( line 5 ) and the LB , pH 6.3 ana'erobic culture ' ( line 6 ) .
Second , the AearA strain ( JF1506 ) lost external acidification control relative to earA + cells ( Table 3 , lines 5 and 6 , earA + ; lines 10 and 11 , AearA ) , ` but still maintained a two-to threefold anaerobiosis inducibility in all media including minimal glucose medium .
Third , the ` data illustrate the complex medium dependence of aniG expression ( Table 3 , lines 3 and 4 ) and the loss of this dependence in earA mutants ( lines 8 and 9 ) .
Isolation of oxrF and oxrG regulatory mutants .
A procedure similar to that 'em ployed for aniG was used to isolate regulatory mutations affecting aniH , a locus anaerobically induced ` in minimal-medium but unaffected by oxrA or oxrB .
One mutation , oxrF405 , proved to be unlinked to aniH ( Table 4 ) .
This mutation caused a complete deregulation of aniH but did not affect any of the other ani-lacZ operon fusions .
This was accomplished by methods similar to those noted above for earA mutations .
We have been unable to map oxrF as yet , but it clearly appears to be a unique regulator since it does not exhibit the characteristics of other known regulators .
The oxrG locus was identified after the transduction of strain JF900 ( aniC : : Mu dA ) with a P22 lysate propagated on a pool of TnJO insertion mutants .
One Tcr transductant ( JF1417 ) contained an insertion which proved to prevent the induction of aniC ( Table 4 ) .
The OxrG phenotype was 100 % o cotransducible with Tcr .
The oxrG : : TnJO insertion is 57 % linked to oxiE at 88 min .
Since the oxyR locus resides in that region , we wondered whether our oxrG locus was actually oxyR .
The oxyR gene is a positive regulator that controls the hydrogen-peroxide adaptive respoInse ( 8 ) .
An insertion into oxyR would lead to extreme sensitivity to H202 .
This did not occur with the oxrG : : TnJO strain .
The oxrG + and oxrG : : TnJO strains gave equivalent zones of inhibition surrounding filter paper disks soaked in 3 % H202 .
Transfer of the oxrG : : TnJO insertion to other ani strains revealed that aniI is also controlled by OxrG ' ( Table 4 ) .
It is significant that aniC and aniI , which map at 93 and 30 to 35 min , respectively , are both class I anaerobiosis-inducible genes .
Their expressions in minimal media require Casamino Acids , are enhanced by nitrate , and ` are also controlled by the oxrA and products .
oxrB The fact that the insertion will not permit anaerobic induction suggests that OxrG is a positive regulator of aniC and I. Two-dimensional electrophoretic analysis of anaerobiosis-inducible proteins .
We have published ' a compilation of environmental stress-inducible proteins including those that are anaerobiosis inducible ( ANI proteins ; 33 ) ' .
The resulting protein index was used in this study to compare the protein patterns obtained with various oxr mutants .
The aim was to begin mapping the control circuitry involved with anaerobic gene expression .
Figure 2 is representative of normally observed ANI proteins as presented earlier ( 33 ) .
Next to each ANI protein is the protein number used in reference 33 .
The gels obtained from oxr mutants ( Fig. 3 and 4 ) and the tabular presentation of these results ( Table 5 ) may be easily compared with the reference photograph in Fig. 2 .
When available , two-dimensional gel electrophoresis was performed with several different mutants fro 'm each class , all with identical results .
Analysis of the oxrA gels ( Fig. 3A and B ) reveals that 15 of the 33 ANI proteins are under OxrA control .
Most of the affected proteins , 14 , were uninducible , reflecting the lack of positive control by OxrA .
However , one protein ( no. 13 , ANI-3 ) was expressed constitutively , suggesting that OxrA may negatively control some ani genes .
The oxrB gels ( Fig. 3C and D ) suggest that OxrB positively controls the production of at least six ANI proteins , all of which were also controlled by OxrA .
In addition , three previously unidentified proteins were observed , of which two were constitutively expressed ( 25 x 62 ; 55 x 62 ) ` and the other was now inducible ( 94 x 37 ) .
This new inducible protein was also observed with oxrA .
A similar phenomenon was observed for an oxrC mutant ( Fig .
' 3E and F ) , although it should be noted that oxrC ( pgi ) is not a transcriptional regulator but is involved in generating an undefined signal from glucose necessary for anaerobic induction of some genes .
With oxrC , the levels of 14 ANI proteins were altered and 5 new proteins were observed .
Five of the ANI proteins were constitutively produced ( three of these were also negatively regulated by oxrA ) .
Nine were uninduced or induced to a lesser extent than normal ( five of these were also affected by oxrA ) .
Of the five new proteins , three were constitutively expressed ( 43 x 63 ; 64 x 44 ; 93 x 27 ) and two were now ANI ( 57 x 57 ; 83 x 32 ) .
While aniG represents a gene not normally induced in minimal-medium , the earA mutation allowed this gene to be expressed under this condition .
Thus , two-dimensional polyacrylamide gel electrophoresis of an earA mutant would be expected to reveal new proteins not normally observed .
Representative gels presented in Fig. 4 show this to be true .
Figures 4A and B show an earA + pattern under aerobic and anaerobic conditions , while Fig. 4C and D show polypeptides from an earA mutant .
Two new proteins were observed , both of which were constitutively expressed .
Surprisingly , one of the normal ANI proteins was also constitutively produced .
The gel electrophoresis performed with the oxrF mutant strain did not reveal any newly expressed proteins or any normal ANI proteins with abnormal control .
The reason for this is not apparent .
One possible explanation is that the protein ( s ) controlled by oxrF is expressed poorly or degraded rapidly , preventing its observation on two-dimen-sional gels .
Two-dimensional gels of the oxrG mutant also revealed nothing out of the ordinary .
It should be remembered , however , that the two known genes which are targets for OxrG require vitamin-free Casamino Acids for their induction .
ANI proteins such as these would not normally be detected by our method of labeling in unsupplemented minimal-medium .
Thus , the lack of detectable difference in the polypeptide profile was not unexpected .
Effects of oxrG and oxrH mutations on ani-lacZ expression 13-Galactosidase activitya Aerobic Anaerobic 46-327-927-836 16 250 13 5 7 48 6 12 Strain Genotype JF1140 JF1410 aniH : : Mu dA aniH : : Mu dA oxrF405 aniC : : Mu dA aniC : : Mu dA oxrG : : TnIO aniI : : Mu dA aniI : : Mu dA oxrG : : TnJO JF900 JF1417 JF1141 JF1509 10 293.4 El ~ 1 23 40 03 .
Standard two-dimensional profile of anaerobiosis-induc-ible proteins .
Growth conditions are described in the text .
The first dimension was an isoelectric focusing gel ( negative pole to left , positive pole toward the right ) containing 1.6 % ampholytes ( pH 5 to 7 ) and 0.4 % ampholytes ( pH 3 to 10 ) .
The second dimension was an 11.5 % polyacrylamide gel .
The boxed proteins are those whose productions increase in response to anaerobiosis ( ANI proteins ) as indicated in reference 33 .
The ANI protein indicated with a circle was also temperature inducible ; arrows indicate ANI proteins which increase in response to one or more starvation stresses .
Each ANI protein is numbered in a manner corresponding to the published catalog ( 33 ) - r ~ ~ a A B ¬ .
¬ IB t d 1 = w A * A0 ' .
A a w * n. - 4 ¬ .
Us ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ w .0 * ¬ 0 a0 4 < .4 0 I 0 0 p : ,4 P4 * S. .
* a * a 4 , , 4 * 0 I A pi : .4 ~ ~ j.-w ¬ A ~ ~ : : ... ~ ~ ~ ~ ~ ~ ~ ~ ~ 11 .
... I - ; : , t , : 4 * ' -LCB- X. , w. .
ey S.V 4 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ i ~ ~ i a ¬ m : : .4 .
A ' - W. 10F E 77mw -.4 - .
I-W * B Us s. 4 U b -- .
4 , wo n - : 0 w 1w S , 0 FIG. 3 .
Two-dimensional profiles of anaerobiosis-inducible proteins from oxrA , oxrB , and oxrC mutants .
( A , C , and E ) Profiles from aerobically grown cells ; ( B , D , and F ) anaerobically grown cells .
( A and B ) Strain SF165 ( oxrAl ) ; ( C and D ) JF1114 ( oxrB8 ) ; ( E and F ) SF242 ( oxrC : : Tn5 ) .
SDS-boiled lysates were prepared as described ( 33 ) and ` run in a two-dimensional system including-a pH 5 to 7 ( right to left ) isoelectric focusing gel and 11.5 % SDS-polyacrylamide gel electrophoresis .
Acidic proteins are situated to the right of each gel .
Boxes indicate normally regulated ANI proteins ; arrowheads indicate abnormally regulated ANI proteins ; stars designate new proteins .
Multiple labelings were done for each strain , and a minimum of two gels were run for each labe ` ling .
Each run included a control parent sample for comparative purposes .
Only consistently induced proteins are indicated .
Tabular results from this figure are presented in Table 5 oxrA oxrB earA oxrF oxrG N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N 6 .
SIN-27 C N ¬ N N N NN N N N N N N ¬ ¬ N N N N N N N + / - N N NN N N N 25 x 62 ( C ) 55 x 67 ( C ) 94 x 37 ( I ) N N N N N N N N N N N N N N N N N N N N N N N N N N N N N C N 49 x 83 ( C ) 51 x 83 ( C ) 11 .
ANI-28 New proteinsc N + / - C + / ¬ ¬ ¬ N N + / - + / - C N N N N C N C ¬ ¬ N N N N + / - N N N N N N 43 x 64 ( C ) 64 x 44 ( C ) 57 x 59 ( I ) 93 x 27 ( C ) 88 x 32 ( I ) C N N N N N N ¬ ¬ + / - N N N N N N N N N N 94 x 37 ( I ) a From reference 33 .
bN , Normal induction ; - , no induction ; C , constitutive ; + / - , induction less than normal .
c Coordinates are those in Fig. 1 of reference 33 .
DISCUSSION Investigations into the transcriptional regulation of anaer - obic gene expression have revealed a complex system of controls .
Three independent pathways for anaerobic induction were previously described .
Studies have shown oxrA and oxrB to be pleiotropic positive control loci which are associated primarily with regulating respiratory functions bN , Normal induction ; - , no induction ; C , constitutive ; + / - , induction less than normal .
c Coordinates are those in Fig. 1 of reference 33 .
The oxrC ( pgi ) gene is involved in the generation of an inducing signal associated primarily with fermentative activities , and tppR is a specific regulator of tppB ( 18 , 19 , 35 ) .
The present study has uncovered several new anaerobic regulatory loci and has attempted to outline overlapping controls through two-dimensional polyacrylamide gel electrophoretic analysis of anaerobiosis-inducible proteins .
Two of the anaerobiosis-inducible loci identified in an earlier report , hyd and aniG , were also found to respond to changes in external pH , with maximal expression occurring under acidic conditions .
None of the other ani-lacZ-fusions responded to these acidification conditions .
It is interesting that hyd only responded to a pH stimulus if it was placed in an anaerobic situation , whereas aniG responded under both aerobic and anaerobic environments .
Maximal expression of aniG , however , required both anaerobiosis and an acid-pH .
It-was observed that growth in complex unbuffered LB under anaerobic conditions resulted in a progressive drop in external pH to approximately 6 .
This would explain in part why the locus was induced so well under our assay conditions .
However , it seems clear that aniG is not normally expressed in minimal-medium because of some requirement present in complex medium .
After growth , the minimal-medium used consistently measured pH 7.2 , a hydrogen-ion concentration which should be satisfactory for showing some anaerobic induction .
Yet aniG expression was barely detectable in minimal-medium .
The earA mutations , therefore , not only relinquish aniG from pH control but also eliminate the complex medium requirement , since these mutants showed largely deregulated aniG expression in minimal-medium .
These results indicate , first , a novel control mechanism involving transcriptional responses t changes in external pH ; second , overlapping or coordinate control by anaerobiosis ; and third , an ill-defined complex medium requirement .
The earA locus we have discovered is novel in that it is directly involved with the transcriptional control of pH-regulated loci .
The gene codes for a negative regulator of aniG and at least two additional genes as indicated by two-dimensional analysis of earA mutant polypeptides .
Since mutations in earA eliminate both pH and complex medium control of aniG , the EarA protein may sense both conditions or may itself be regulated by them .
Its normal role would be to repress the transcription of certain genes during alkaline growth-conditions or in minimal-medium .
We propose that an additional regulatory protein exists that interacts with some component of LB ( e.g. , peptide ) .
) complex is required to release aniG from EarA repression under acidic conditions .
Mutations resulting in the loss of this second regulator would prevent the induction of aniG .
Furthermore , earA mutations resulting in constitutive aniG expression would be dominant to mutations in this second regulator .
We have recently identified mutations consistent with our predictions and are characterizing the responsible locus .
It will be important to determine what functions as a corepressor or inactivator of the EarA protein .
For example , what role does proton motive force , of which ApH is a component , play in activating or inactivating the EarA repressor ?
The earA locus has been cloned and was found to code for a 33-kilodalton protein which is membrane associated .
This raises interesting questions concerning the mechanism by which EarA senses external acidification and subsequently releases aniG from negative control ( J. Foster , manuscript in preparation ) .
The other pH-regulated locus , aniA ( hyd ) , which exhibited strict requirements for both acidic and anaerobic conditions , was unaffected by earA mutations .
Thus , it would seem there are also multiple pathways for pH regulation .
Control of hyd by pH could very likely be oxrA independent since hyd is still significantly inducible ( 26-fold ) in an oxrA background as compared to an oxrA + ( 80-fold ; 1 ) background .
Pecher et al. ( 31 ) noted a similar pH phenomenon with an hyd : : lac operon fusion in E. coli using medium with low buffering capacity .
Their study could not differentiate between higher rates of formate accumulation at acid-pH or a pH effect on the system controlling hyd transcription .
Our data were obtained by using a heavily buffered medium and suggest that pH rather than formate accumulation was responsible for the effect .
The data presented for aniA ( hyd ) and aniG lend support to the theory that the transmembrane potential is sensed by one or more effector molecules ( EarA ?
A recent report by Slonczewski et al. described pH-regulated gene expression in E. coli ( 32 ) .
Both of the exa ( external acidification ) loci described also exhibited an an-aerobic control component similar to what we describe for aniG .
The selection method used in our study , MacConkey medium , was much different from that used by Slonczewski et al. ( 32 ) for isolating external acid-inducible fusions .
Their method involved altering the medium pH through buffering .
MacConkey medium relies upon acid production by the organisms to incorporate dye into a Lac ' colony .
Thus , our method screens both for genes which respond to anaerobi-osis as well as for genes which are regulated by fluctuations in external pH. Further analysis of these and similar pHcontrolled loci could lead to a greater understanding of the mechanism ( s ) of pH homeostasis .
It is not insignificant that in the course of producing gastrointestinal disease S. typhi-murium must experience exposure to acidic conditions while passing through the stomach , as well as an anaerobic environment in the intestine .
The second novel regulatory gene we discovered , oxrF , controls aniH , an oxrAB-independent locus .
While we have not mapped or characterized this gene in detail , it clearly represents a new pathway for anaerobic gene expression .
The oxrG locus was shown to be a positive regulator of aniC and anil .
Both of these target genes are oxrA and oxrB dependent and require Casamino Acids for induction , and their expressions are enhanced by nitrate .
There is some similarity to the narC ( nitrate reductase ) locus in E. coli , whose expression is repressed by oxygen and enhanced under anaerobiosis by the addition of nitrate .
The narL locus has been identified as controlling the nitrate-dependent component of narC induction ( 34 ) .
However , this locus has not been identified in S. typhimurium .
In addition , it appears that all three loci , oxrA , oxrB , and oxrG , are required for any expression of the aniCI regulon , whereas narC requires only oxrA and narL .
It is possible that the oxrG product is involved in the Casamino Acids or nitrate signals rather than oxygen .
Since OxrA is analogous to Fnr in E. coli , it appears to be a transcriptional activator of anaerobic gene expression .
It is not clear , as yet , whether OxrB and OxrG are actually transcriptional activators or , as is the case with OxrC , are involved with the generation of an induction signal-molecule .
In either instance , the determination of their functions will be important in piecing together the complexities of anaerobic gene expression .
In summary , there are a minimum of nine pathways for anaerobic gene expression as characterized by genetic and biochemical techniques : ( i ) oxrA dependent , ( ii ) oxrAB dependent , ( iii ) signal oxrABC dependent , ( iv ) oxrABG dependent , ( v ) oxrAC dependent , ( vi ) oxrC dependent , ( vii ) earA dependent , ( viii ) oxrF dependent , and ( ix ) tppR dependent .
Of the 33 ANI proteins normally detected , 22 are controlled by one of these pathways .
However , 11 were not affected by any of the regulatory mutations tested , suggesting the presence of additional pathways .
Our studies with earA indicate overlapping controls involving pH and anaerobiosis as well .
It is also significant that the two-dimensional polypeptide analysis indicates that oxrA , oxrB , and oxrC have the capacity for positive and negative regulation depending upon the target gene .
This is not unlike what has been observed with cyclic AMP-CRP .
Yamamoto and Droffner ( 38 ) have presented data suggesting that DNA supercoiling plays a significant role in anaerobic gene expression .
Their theory centers around the isolation of strict aerobic mutants containing mutations in gyrA and gyrB which lead to little or no DNA gyrase activity .
Our data do not necessarily conflict with theirs , since the alteration of supercoiling under anaerobic conditions could be the signal that triggers the transcription of various oxr regulatory loci .
However , evidence from E. coli indicates fnr ( oxrA ) is transcribed under both aerobic and anaerobic conditions ( 22 ) .
Thus , it would appear that at least one anaerobic regulator does not respond itself to supercoiling .
Altered chromosome structure could , however , facilitate Fnr access to target promoter sequences .
Undeniably the phenomena of anaerobic gene expression as well as pH regulation are intricate processes whose mechanistic complexities will require a major effort to understand .
VOV L. 1701 ~ ~ ~ N , 1988 O79L08. ,8 OXAYDGpEHNREGULATION IN S. TYPHIMURIUM 849 ear + +02 la ear + 0 ... WI o 10 0 0 b. : ... , 0-1 ~ ~ - t ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ) P ... Uss ~ ~ IS ear 206 + 02 ear 206 -02 AW ~ ~ ~ ~ ~ ~ 1 FIG. 4 .
Two-dimensional profiles of ANI proteins from earA ' and earA mutants .
Top panels are two-dimensional gels of SDS lysates of strain JF1294 ( earA ' ) .
The two lower panels are from JF1304 ( earA206 ) .
Boxes indicate normal ANI proteins .
Arrowheads indicate constitutively produced proteins found in the earA mutant .
Each panel is only a section of a complete gel .
ACKNOWLEDGMENT This work was supported by Public Health Service grant GM 34147-03 from the National Institutes of Health LITERATURE CITED 1 .
Aliabadi , Z. , F. Warren , S. Mya , and J. W. Foster .
Oxygen-regulated stimulons of Salmonella typhimurium identified by Mu d ( Ap lac ) operon fusions .
Barrett , E. L. , C. E. Jackson , H. T. Fukumoto , and G. W. Chang .
Formate dehydrogenase mutants of Salmonella typhimurium : a new medium for their isolation and new mutant classes .
Bilous , P. T. , and J. H. Weiner .
Dimethyl sulfoxide reductase activity by anaerobically grown Escherichia coli HB101 .
Pascal , J. Ratouchniak , and M. Chippaux .
Alteration by mutation of the control by oxygen of the nar operon in Escherichia coli .
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. W. Watanabe , and Y. Ogata .
Specialized transduction of tetracycline resistance by phage P22 in Salmonella typhimurium .
Properties of a high frequency transducing lysate .
Chippaux , M. , V. Bonnefoy-Orth , J. Ratouchniak , and M. C. Pascal .
Operon fusions in the nitrate reductase operon and study of the control gene nirR in Escherichia coli .
Christman , M. F. , R. W. Morgan , F. S. Jacobson , and B. N. Ames .
Positive control of a regulon for defenses against oxidative-stress and some heat-shock proteins in Salmonella typhimurium .
Chumley , F. G. , R. Menzel , and J. R. Roth .
Hfr-formation directed by TnlO .
The number of anaerobically regulated genes in Escherichia coli .
Clark , M. A. , and E. L. Barrett .
The phs gene and hydrogen sulfide production by Salmonella typhimurium .
Davis , R. W. , D. Botstein , and J. R. Roth ( ed .
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 .
Fimmel , A. L. , and B. A. Haddock .
Use of chl-lac fusions to determine regulation of gene chlC in Escherichia coli K-12 .
Gharbi , S. , A. Belaich , M. Murgier , and A. Lazdunski .
Multiple controls exerted on in-vivo expression of the pepN gene in Escherichia coli : studies with pepN-lacZ operon and protein fusion strains .
Holley , E. A. , and J. W. Foster .
Bacteriophage P22 as a vector for Mu mutagenesis in Salmonella typhimurium : isolation of nad-lac and pnc-lac gene fusions .
Holley , E. A. , M. P. Spector , and J. W. Foster .
Regulation of NAD biosynthesis in Salmonella typhimurium : expression of nad-lac gene fusions and identification of a nad regulatory locus .
Jamieson , D. J. , and C. F. Higgins .
Anaerobic and leucine-dependent expression of a peptide transport gene in Salmonella typhimurium .
Jamieson , D. J. , and C. F. Higgins .
Two genetically distinct pathways for transcriptional regulation of anaerobic gene expression in Salmonella typhimurium .
Jamieson , D. J. , R. G. Sawers , P. A. Rugman , D. H. Boxer , and C. F. Higgins .
Effects of anaerobic regulatory mutations and catabolite-repression on regulation of hydrogen metabolism and hydrogenase isoenzyme composition in Salmonella typhi-murium .
Jones , H. M. , and R. P. Gunsalus .
Transcription of the Escherichia coli fumarate reductase genes ( frdABCD ) and their coordinate regulation by oxygen , nitrate , and fumarate .
Jones , H. M. , and R. P. Gunsalus .
Regulation of Esche-richia coli fumarate reductase ( frdABCD ) operon expression by respiratory electron-acceptors and the fnr gene product .
Kuritzkes , D. R. , X.-Y .
Zhang , and E. C. C. Lin .
Use of 4 ' ( glp-lac ) in studies of respiratory regulation of the Escherichia coli anaerobic sn-glycerol-3-phosphate dehydrogenase genes ( glpAB ) .
Kwan , H. S. , and E. L. Barrett .
Roles for menaquinone and the two trimethylamine oxide ( TMAO ) reductases in TMAO respiration in Salmonella typhimurium : Mu d ( Apr lac ) insertion mutations in men and tor .
Lambden , P. R. , and J. R. Guest .
Mutants of Escherichia coli unable to use fumarate as an anaerobic electron-acceptor .
Maloy , S. R. , and W. D. Nunn .
Selection for loss of tetracycline resistance by Escherichia coli .
Miller , J. R. ( ed .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 28 .
Merberg , D. , and P. Datta .
Altered expression of biode-gradative threonine dehydratase in Escherichia coli mutants .
Neidhardt , F. C. , P. L. Bloch , and D. F. Smith .
Culture medium for enterobacteria .
Burini , and M. Chippaux .
Regulation of the trimethylamine N-oxide ( TMAO ) reductase in Esche-richia coli : analysis of tor : : Mudl operon fusions .
Pecher , A. , F. Zinoni , C. Jatisatienr , R. Wirth , H. Hennecke , and A. Bock .
On the redox control of synthesis of anaerobically induced enzymes in enterobacteriaceae .
Slonczewski , J. L. , T. N. Gonzalez , F. M. Bartholomew , and N. J. Holt .
Mu d-directed lacZ-fusions regulated by low pH in Escherichia coli .
Spector , M. P. , Z. Aliabadi , T. Gonzalez , and J. W. Foster .
Global control in Salmonella typhimurium : two-dimensional electrophoretic analysis of starvation - , anaerobiosis - , and heat-shock-inducible proteins .
Requirement of Fnr and NarL functions for nitrate reductase expression in Escherichia coli K-12 .
Strauch , K. L. , J. B. Lenk , B. L. Gamble , and C. G. Mifler .
Oxygen regulation in Salmonella typhimurium .
Vogel , H. J. , and D. M. Bonner .
Acetylornithase of Escherichia coli : partial purification and some properties .
Winkleman , J. W. , and D. P. Clark .
Anaerobically induced genes of Escherichia coli .
Yamamoto , N. , and M. L. Droffner .
Mechanisms determining aerobic or anaerobic-growth in the facultative anaerobe Salmonella typhimurium .
Zinoni , F. , A. Beier , A. Pecher , R. Wirth , and A. Bock .
Regulation of the synthesis of hydrogenase ( formate hydrogen-lyase linked ) of Escherichia coli .