169 , No. 6 Selection of Mutations That Alter the Osmotic Control of Transcription of the Salmonella typhimurium proU Operon JANICE DRUGER-LIOTTA , t VIRGINIA JOAN PRANGE , f DAVID G. OVERDIER , AND LASZLO N. CSON Department of Biological Sciences , Purdue University , West Lafayette , Indiana 47907 Received 1 December 1986/Accepted 12 March 1987 KA * We isolated 60 independent mutations , designated osnX , in Salmonella typhimurium that result in constitutive expression of the normafly osmoregulated proU operon .
1ach of the osmX mutations is closely linked to the proU locus and eis-dominant over the osniX + allele in diploid strains .
These results suggest that the mutations are probably in the 5 ' transcriptional control region of the proU operon .
Our failure to obtain either recessive or unlinked mutations that altered the osmotic control of transcription of the proU operon suggests that transcriptional regulation of the gene is not under the negative control of a repressor protein that is dispensable for celi viability .
We discuss possible models for the mechanism of osmotic regulation of transcription of the proU operon .
Organisms ranging from bacteria to algae , vascular plants , and invertebrate animals respond to osmotic sttess by increasing the internal concentrations of a number of low-molecular-weight compounds ( 15 , 39 ) .
These substances function as compatible solutes that balance the internal osmolarity of the organisms with that of the exterior .
Prom-inent among the compounds accumulated by a wide variety of organisms are the amino-acid proline and the quaternary amine glycine betaine ( N,N,N-trimethylglycine ) .
Bacteria can accumulate high levels of compatible solutes by mutations that result in their overproduction ( 9 ) or by transport from the exterior ( 10 , 25 ) .
Exogenous proline and glycine betaine stimulate the growth-rate or respiration rate of several species of bacteria in media of elevated osmolarity ( 7 , 25 , 32 , 35 ) .
We found that in the enteric bacterium Salmonella typhimurium , enhancement of growth-rate by proline during osmotic-stress is dependeht on the uptake of this metabolite by two independent transport systems , the ProU and the ProP systems , whose activities are stimulated by the exposure of the cells to osmotic-stress ( 10 , 12 , 13 ) .
Cairney et al. ( 4 , 5 ) reported that glycine betaine is also a substrate for the ProU and ProP transport systems during conditions of osmotic-stress , with the ProU system having a much greater affinity for glycine betaine than for proline .
Gowrishankar ( 16 ) , Cairney et al. ( 5 ) , and we ( 12 , 13 ) found that osmotic stimulation of the ProU is mainly system effected by enhanced transcription of the proU operon , although Cairney et al. ( 5 ) found that the osmotic strength of activity of the medium also directly regulates the transport the ProU system .
The structural gene ( s ) of the ProP system invariant level in media of various is expressed at a nearly osmolarities , and osmotic-regulation of this system is brought about by stimulation of the activity of the transport ( 4 , Escherichia coli , whose growth-rate is also proteins 13 ) .
and glycine betaine in media of inhibstimulated by proline itory osmolarity , has transport systems that are analogous to of S. typhimurium ( 16-19 , 27 ) .
the ProU and ProP systems t Present address : Department of Medicine , Northwestern University Medical School , Chicago , IL 60611 .
t Present address : Department of Plant Pathology , University of Florida , Gainesville , FL 32608 Studies conducted with S. typhimurium and E. coli strains carrying fusions of the lacZ gene to the proU promoter indicated that exposure of the cells to media of elevated osmolarity causes a rapid increase in the steady-state transcription rate of the proU operon , provided that the exogenous osmolytes are not freely diffusible across the cell membrane ( 5 , 13 , 16 ) .
Substances , such as glycerol or ethanol , that can diffuse across the membrane can not trigger the induction of proU , suggesting that the signal for induction of the proU operon is not increased intracellular osmolarity .
However , the above studies did not provide much insight into the mechanisms that mediate the osmotic regulation of the proU operon .
We used a genetic approach to elucidate steps involved in osmoregulation of transcription of the proU operon and isolated mutants of S. typhimurium exhibiting abnormal regulation of proU expression .
In this manuscript , we describe the characterization of mutants that express the proU operon constitutively .
The isolation of one of these mutants has been described previously in preliminary form ( 12 ) .
MATERIALS AND METHODS Media and growth-conditions .
The composition of the rich the medium , LB , and minimal-medium 63 and growth-conditions used have been described ( 13 ) .
The carbon source mM unless otherwise in minimal-medium was 10 glucose K indicated .
Medium of low-osmolarity was medium de-lactose scribed by Kennedy ( 23 ) .
Bromothymol blue ( BTB ) indicator medium was based on that described by Miller ( 31 ) , that the basal medium was K medium in which the except was yeast extract replaced with 5 g of Casamino Acids ( Difco Laboratories ) per liter .
The osmolarity of medium 63 , K medium , and BTB indicator was increased by the addition of NaCl at the indicated concentrations , with the pH of the medium readjusted to that of the respective medium without NaCl .
Solid media were obtained by the addition of 20 g of agar ( Difco ) per liter .
Cultures were grown aerobically at 370C .
MacConkey lactose indicator medium was purchased from Difco .
Antibiotics were used at the following concentrations ( milligrams per liter ) : sodium ampicillin , 25 ; tetracycline , 15 ; kanamycin sulfate , 75 ; chloramphenicol , 12.5 ; and nalidixic-acid , sodium salt , 4 .
Resistance and sensitivity to these compounds are denoted by superscript r and s , respectively .
When used , the proline analogs L-azetidine-2-carboxylate and 3,4-dehydro-DL-proline were present at 1 mM .
The bacterial strains used were derivatives of S. typhimurium LT2 .
Their constructions are summarized in Table 1 .
The generalized transducing phages P22 HT1/105 intlO3 and KB1 were used in strain constructions .
The transductions were performed by the methods of Davis et al. ( 11 ) and Boro and Brenchley ( 3 ) .
Phage Mu d11734 ( Kanr lac ) ( 6 ) was used to replace the original Mu dl B : : Tn9 phage in the proU1844 insertion in strain TL456 , generating proU-lacZ operon fusions that were stable to transposition .
This was accomplished by transducing strain TL456 to Kanr with P22 lysate of strain TT10286 ( hisD9953 : : Mu d11734 ) .
Strains in which the Mu dl B : : Tn9 phage were replaced by Mu dI1734 arose as a result of two crossover events across the homologous regions common to the phage near their two termini ( 6 ) .
Approximately 1 % of the Kanr transductants obtained on LB kanamycin plates proved to be derivatives in which Mu dl B : : Tn9 has been replaced by Mu d11734 .
Such derivatives were recognized by their His ' Amps Kanr phenotype .
One transductAnt obtained was designated TL501 ; P-galactosidase assays indicated that the lacZ in this strain responded to osmotic regulation as in strain TL456 .
Selection of proU constitutive mutants .
We used two different selection proceduies to generate mutants that express the proU operon constitutively .
The first one was based on fact that proline-auxotrophic strains that also lack the PutP and ProP transport systems are unable to grow in medium 63 with 0.1 mM proline because of the defects in proline transport ( 30 ) .
However , these mutants can grow normally with 0.1 mM proline in the presence of an additional 0.3 M NaCl or 0.45 M sucrose , because the osmotic-stress induces the ProU system , which can mediate the rapid uptake of 0.1 mM proline ( 10 ) .
By selecting derivatives of a AproBA putP proP mutant that were 4ble to grow with 0.1 mM proline in the absence of osmotic stres5 , we were able to obtain one strain expressing the ProU system constitutively .
For the selection , 0.1-ml samples of an overnight LB culture of strain TL185 were spread on medium 63 plates containing 0.1 mM proline and 0.2 mM arginine .
( The latter amino-acid was added to repress the arginine biosynthetic enzymes and thus minimize the occurrence of prolineprototrophic derivatives in which mutations in the arginine-biosynthetic enzymes could bypass the AproBA block [ 22 ] .
) Mutants able to grow on the above medium appeared at the approximate frequency of 1 per 106 cells plated .
Seventy-five mutants that were obtained by this selection were restreaked on the selective medium .
Of these , 8 could grow with 0.1 mM proline as rapidly as a control proP + putP + proU + AproBA strain , whereas the remaining 67 strains showed only very slight growth .
Only the eight strains exhibiting the most rapid growth were characterized further .
In addition to mutations causing increased expression of the ProU system , the ability of the above strains to grow with 0.1 mM proline might have been due to a number of other mutations : reversion of the proP1654 mutation of strain TL185 , alterations in the substrate specificity of permeases that normally transport other amino-acids , enhanced membrane leakiness , increase in the affinity of prolyl tRNA synthetase for proline , etS .
To test whether the growth of the eight potential proU constitutive strains on 0,1 mM proline depended on the functioning of the ProU permease , they were transduced to Tetr by phage P22 grown on strain TL187 ( proU : : TnlO ) .
Of the eight strains , one ( strain TL395 ) gave rise to proU : : TnJO transductants that were unable to grow with 0.1 mM proline .
Extensive characterization ( see Results ) indicated that strain TL395 expressed the ProU system constitutively ; the mutation resulting in this phenotype has been designated osmXl .
The proU : : TnlO transductants of the other seven strains were still able to grow well with 0.1 mM proline .
Thus , their respective parental strains probably carried one of the other types of suppressor mutations for the proline transport defect and were not studied further .
The second selection procedure for the generation ofproU constitutive mutants was based on the isolation of derivatives of a 4F ( proU-lacZ ) strain that exhibited increased expression of the lac genes on BTB lactose indicator me-dium .
Strains carrying the proU184441acZ fusion form blue colonies on this indicator medium , but they form dark yellow colonies if the osmolarity of the medium is increased with .0.3 M NaCI .
In an attempt to obtain insertions of TnlO into potential regulatory genes for proU , we mutagenized strain TL501 ( proU-lacZ ) with TnlO by the procedure of Davis et al. ( 11 ) .
Samples of the culture , infected with phage P22 : : TnJO lysate of strain NK337 , were spread at a dilution that gave -500 colonies per plate on 100 BTB-lactose-tetracycline plates .
Colonies that exhibited more intense yellow color than the parental strain were visible after 48 h .
Such colonies were purified on alizarin yellow-crystal violet indicator medium to obtain phage P22-free cells ( 11 ) and tested again on BTB lactose plates .
Strains TL1022 through TL1032 obtained in this manner were saved as potential mutants expressing 4F ( proU-lacZ ) constitutively , with the mutant alleles designated osmX2 through osmXl2 .
We carried out two sets of transductions to test whether the proU constitutivity in these strains w4s due to the TnlO insertion .
( Strains TL1024 through TL1032 proved to be resistant to phage P22 , probably because they became lysogens during the TnlO mutagenesis ; we used phage KB1 in the initial transductions with these strains .
) We transduced the wild-type strain TL1 to Kapr by phage P22 grown on the osmX Mu dl1734 lysogens and screened transductants for the inheritance of proU constitutivity or Tetr .
The Kanr transductants obtained with each donor strain proved to have inherited the ProU constitutive phenotype of the respective parents ( as judged by colony color on BTB lactose plates ) without inheriting Tetr .
These results indicate that the mutation causing ProU constitutivity is closely linked to the proU operon and that the constitutivity is not caused by the TnlO insertion , but arose independently from acquisition of TnlO .
The Kanr transductants of strain TL1 obtained with each of the ProU constitutive mutants proved to have normal sensitivity to phage P22 , indicating that the phage P22 resistance of the parental strains is not due to the ProU constitutivity .
Complementation tests of the mutations causing ProU constitutivity .
We carried out complementation tests of the mutations giving rise to proU constitutivity by transducing them into strain TL819 , which carries two copies of the proU + operon within a tandem duplication of the 42-to 58-min region of the chromosome .
The merodiploid proU + lproU + strairp TL819 was constructed by a procedure developed by Hillyard and Roth ( personal communication ) , using TnlO as the join point for the duplication .
The steps of the construction are shown in Fig. 1 .
The osmX mutations were transduced along with the closely linked proUl844 : : Mu d11734 insertion into strain TL819 , resulting in osmX + pro U + losmX pro U : : Mu d11734 diploids .
To ensure that th transductants were carrying the chromosomal duplication , the recipient strain was grown and the transductants were selected in medium containing tetracycline .
We found that the transductants obtained in the above crosses inherited proU constitutivity at a very high frequency : -70 % for osmXl and > 90 % for osmX2 thrqugh osmX12 , as established by their Lac + phenotype on BTB-lactose-tetracycline plates .
These linkages of proU constitutivity to the proU1844 insertion were similar to those seen with recipients carrying only one copy of the osmX + proU + alleles ( see Results ) .
We interpret these results to mean that the osmX mutations are dominant to the osmX + allele , bepause if they were recessive , the osmX mutations would have appeared to be unlinked to the proUJ844 : : Mu dl1734 insertions when the diploid strain TL819 was the recipient .
To confirm that the presumed osmX c ( proU-lacZ ) / osmX + proU + heterozygotes were indeed diploid , we obtained haploid Tet5 segregants from the diploid strains .
In each case , we were able to recover segregants that were osmX F ( proU-lacZ ) ( i.e. , Kanr , yellow on BTB-lactose , unable to grow on medium 63 with 0.1 mM proline and 0.3 M NaCl ) , or osmX + proU + ( i.e. , Kans , blue on BTB-lactose , unable to grow on medium 63 with 0.1 mM proline , able to grow on medium 63 with 0.3 M NaCl and 0.1 mM proline ) , indicating that the parental strain carried the osmX ` I ( proU-lacZ ) and osmX + proU + alleles .
The assays for P-galactosidase were carried out with mid-logarithmic-phase cells ( A6w , 0.6 ) by the method of Miller ( 31 ) , except that the results are expressed as nanomoles of product formed per minute per milligram cell protein , with the protein concentration determined by the method of Lowry et al. ( 26 ) .
TL839 ; Kanr Tetr yellow on BTB-lactose P22 ( TL817 ) - + TL1 ; Tetr P22 ( TL1081-TL1092 ) - * TL819 ; Kanr Tetr yellow on BTB-lactose TL1113 1654/523 A47 P22 ( TL835 ) Y * TL819 ; Kanr Tetr TR5281 + A523 A47 J. R. Roth ( 30 ) TT513 + + + zec-2 : : TnlO J. R. Roth ( 34 ) TT1672 + + + + + melA361 : : TnlO R. Menzel and J. R. Roth ( 30 ) T710286 + + + + + hisD9953 : : Mu dI1734 K. Hughes and J. R. Roth aThe genotype osmX denotes cis-acting mutations resulting in increased basal expression of proU + or 4D ( proU-IacZ ) alleles .
The other genetic symbols have been defined ( 6 , 20 , 35 ) .
b The notation `` P22 or KB1 ( M ) - * N ; A B C. .
denotes that a phage P22 or KB1 lysate of strain M was used to transduce strain N , selecting phenotype A , and scoring phenotypes B , C , etc. .
zfi-9 : : TnlO TL839 + + 1654 4523 + + A47 + TnlO in unknown location TL1022-TL1032 2-12 1844 : : Mu dI1734 TL1041-TL1052 1-12 1844 : : Mu dI1734 + + + A47 TL1061-TL1072 1-12 1844 : : Mu dI1734 1654 A523 1654 A523 A47 TL1081-TL1092 1-12 1844 : : Mu d11734 zfi-9 : : TnlO + TL1095 + / + + / + + + DUP [ zec-2 * TnlO * zfi-9 ] DUP [ zec-2 * TnlO * zfi-9 ] A47 TL1101-TL1112 +11 -12 + / 1844 : : Mu dI1734 1654 A523 + / 1844 : : Mu d11734 + + + / + + + DUP [ zec-2 * TnIO * zfi-9 ] RESULTS Phenotype of the osmXl mutant .
We obtained one constitutive mutant , strain TL395 ( osmXl proU + AproBA putP proP ) , by direct selection for derivatives of strain TL185 ( osmX + proU + AproBA putP proP ) with increased proline transport activity in media of low-osmolarity ( see Materials and Methods ) .
The growth-rate of strain TL395 in liquid medium 63 containing 0.1 mM proline and no NaCl was sixfold greater than that of the parental strain TL185 ( Table 2 ) .
In the absence of osmotic-stress , proline-prototrophic putP proP mutants are normally resistant to the proline analogs L-azetidine-2-carboxylate and 3,4-dehydro-DL-proline .
Strain TL675 ( osmXl putP proP ) was inhibited by these two compounds in medium 63 , whereas the osmX + control strain TL673 ( putP proP ) was able to grow unimpaired .
This increased sensitivity of strain TL675 to the proline analogs is due to increased expression of the ProU system , which is known to transport these two compounds ( 5 , 10 ) .
osmXl results in increased bsal transcription of the proU operon .
To test whether the osmXl mutation resulted i increased transcription of the proU operon , we transduced the proU-lacZ fusion from strain TL671 ( proUJ844 : : Mu d11734 ) into strain TL769 ( osmXl proU + zfi-9 : : TnJO ) .
The donor strain TL671 ( osmX + ) forms white ( Lac - ) colonies on lactose MacConkey agar plates , because the proU-lacZ fusion is expressed only at basal levels on this medium of low-osmolarity .
The transductants fell into two groups with respect to their Lac phenotype : 78 % formed Lac-colonies on lactose MacConkey agar , but the remaining 22 % formed crimson ( Lac ' ) colonies , indicating that they were expressing the proU operon constitutively .
This result suggests not only that the osmXl mutation results in increased transcription of the proU operon in media of low-osmolarity , but also that it is closely linked to the proU operon .
Strain TL769 + / 1844 : : Mu d11734 + + + / + + + DUP [ zec-2 * TnIO * zfi-9 ] P srl h / * D : - MUudI + I TL813 V zfi-9 : ` TnIO ( P22.TTS13 ) m-2 : : Tn/2 hIse ( HIs KanR TetR ) -- 9 - : : - T , n % V7 pwU + h/SD : : MrdI 5,1 + Hh + , KonR TL815 p , oU + zfl-9 * TnIO * zoc-2 t ( j ( His + KanR TetR h/sD : : MudI fr zfi-9 : : Tnl0 P22.TL815-0-TT1O286 TL817 o , / + ( TetR Hise KanR ) A h/sD : : MudI Aoro A Prou zf / -9 * Tn 1O * m-2 P22 .
TL817 - TL714 h/s + A proU + so , + pu > + TL714 hs / \ ( P2 i 2 .
znf-9 * TnvO * zmc-2 TL817 ) / \ pO1fJ 5,1 + TotR , KonS TL819 o ( 1 + / Jis + srl + ( TetR , Kans , segregates Tets ) hIls zfl-9 * TnlO * zoc-2 P22 .
( TL1O81-1092 ) -- TL819 OSrX + pf U + At hIs .
osmXPrPU : ; MudI h1s + zfl-9 * TnlO * JV-2 o s , r / + TL12-TL1112 ( KanR , TetR ) r / ) is + w + / A osm.X proU : : MudXI Sr / + osmXO Zfl-9 * Th?O * jec-2 FIG. 1 .
Construction of osmX 4 ( proU-lacZ ) / osmX + proU + diploid strains .
We transduced strain TL813 ( zfi-9 : : TnlO hisD : : Mu d11734 ) to His + by phage P22 grown on strain TT513 ( zec-2 : : TnlO his + ) .
The TnlO 's in the zfi-9 : : TnlO and zec-2 : : TnlO insertions are oriented in the same direction on the chromosome ( D. Overdier , unpublished results ) .
Therefore , we were able to obtain His + transductants carrying a duplication between the zec-2 and zfi-9 insertions as a result of a recombinational event between zfi-9 : : TnlO on one of the daughter strands of a replicated chromosome and zec-2 : : TnlO on a transducing particle , and a second recombinational event within the his region on the transducing particle and the his region on the other daughter strand .
Transductants carrying duplications ( e.g. , strain TL815 ) were recognized by their His ' Kanr phenotype and by the fact that they gave rise to His-Kan ' and His ' Kans segregants under nonselective conditions .
Strain TL815 carries two copies of TnlO .
To obtain a strain with a single TnlO acting as the join point of the chromosomal duplication , we transduced strain TT10286 ( hisD : : Mu dI1734 ) to Tetr by a phage P22 lysate of strain TL815 .
Transductants ( e.g. , strain TL817 ) carrying chromosomal duplications inherited the His + phenotype together with the Tetr and remained Kanr .
We obtained strain TL819 by transducing strain TL714 to Tetr by a phage P22 lysate of strain TL817 .
osmXI confers ability to grow in 0.1 mM proline on a proBA putP proP mutant ' Specific growth-rate ( generations/h ) in medium 63 containing Proline Proline Proline Proline ( 2 mM ) ( 0.1 mM ) ( 2 mM ) NaCl ( 0.3 M ) + NaCI ( 0.3 M ) TL185 ( osmX + proU + AproBA AputP proP ) 0.14 1.0 0.59 0.83 TL395 ( osmXI proU + AproBA AputP proP ) 0.87 1.0 0.67 0.71 TL406 ( osmXI proU1844 : : Mu dl B : : Tn9 AproBA AputP proP ) < 0.09 1.0 0.10 0.71 aStrains were inoculated into the indicated media at a dilution of 1:20 , and the optical density of the cultures was measured as a function of time Strain ( relevant genotype ) ( osmXl proU + AproBA putP proP ) was able to grow with 0.1 mM proline on medium 63-glucose .
None of its proUJ844 : : Mu d11734 transductants , whether they were osmX + or osmXl , were able to do so , presumably because the Mu d11734 fusion inactivated the ProU system .
P-galactosidase levels of osmX 4 ( proU-lacZ ) mutants .
In addition to the osmXI mutant that was obtained by direct selection for increased ProU activity , we isolated 11 other osmX mutants by selection for derivatives of a 1D ( pro U-lacZ ) strain that exhibited increased lac expression on BTB-lactose indicator plates ( see Materials and Methods ) .
We determined the 13-galactosidase activities of these strains grown under various conditions of external osmolarity ( Table 3 ) .
For this experiment , cells were grown in K medium containing NaCl at the indicated concentrations rather than in medium 63 , because K medium elicits approximately 20-fold lower basal expression of pro U-lacZ then medium 63 ( V. J. Prange and L. N. Csonka , unpublished results ) .
In K medium without NaCl , all the osmX proU-lacZ mutants except strain TL1049 had a higher basal level of 1-galactosidase activity than the control strain TL671 ( osmX + proU-lacZ ) ( Table 3 ) .
The osmX mutants fell into three patterns of , B-galactosidase synthesis .
Strains TL1041 , TL1042 , TL1051 , and TL1052 showed an 8-to 13-fold increase in the basal level of expression of proU in K medium , but they were able to raise the expression of the operon nearly normally by exposure to 0.2 or 0.5 M NaCl .
Strains TL1044 , TL1045 , TL1046 , TL1047 , and TL1048 had an 8-to 200-fold-higher basal level of expression of proU , but exposure to osmotic-stress , did not elicit normal induction of the operon .
The third group , which included strains TL1043 , TL1049 , and TL1050 , exhibited very slightly increased or unchanged basal levels of 0-galactosidase .
Possibly the mutations in these three strains damaged the promoter of the proU operon or the lacZ gene of Mu dl734 .
Table 3 also lists the phenotypes of these strains on BTB-lactose and MacConkey-lactose indicator media .
Although strains TL1043 , TL1049 , and TL1050 exhibited a very slight , if any , increase in P-galactosidase activity , they appeared to express the lac genes at an increased level , as judged by their colony color on the indicator media of low-osmolarity , and therefore probably carry alterations in the osmotic control of proU expression .
The reason for the discrepancy between the assay results and the phenotypes of these strains on the indicator media is unclear .
Mapping of the osmX mutations .
To gain insight into the nature of the osmX mutations , we carried out extensive fine-structure mapping to localize them on the chromosome .
In these mapping experiments , we used the zfi-9 : : TnlO insertion as the reference point , and therefore it was necessary to establish the position of this TnlO with respect to outside markers .
This was done in two reciprocal three-point mapping experiments , in which strain TL501 ( proU1844 : : Mu dl1734 nalB + ) was transduced to Tetr by phage P22 grown on strain TL657 ( proU + nalB zfi-9 : : TnlO ) and strain TL657 transduced to Kanr by phage P22 lysate of strain TL501 ( Table 4 ) .
The sole gene order that was consistent with both transductions was proU + nalB zfi-9 : : TnlO .
Previous mapping placed the proU operon at 57 units of the chromosome between pheA and nalB ( 34 ) , so the overall gene order is pheA proU nalB zfi-9 : : TnlO recA .
As the-first step in mapping the osmX mutations , we transduced strain TL839 ( osmX + proUJ ) to Kanr by phage P22 lysates grown on strains TL1041 through TL1052 ( osmX 2454 TABLE 3 .
Expression of P-galactosidase in osmX 4 ( proU-lacZ ) strains P-Galactosidase sp act ( nmol/min per mg of protein ) in K medium containing : Colony color Strain plates lMaacConkey ¬ No NaCl 0.6 6 6 8 0.2 M NaCI 6 31 11 23 23 0.5 M NaCl 75 68 80 69 71 51 50 34 34 24 TL671 ( osmX + ) ( control ) TL1041 ( osmXl ) TL1042 ( osmX2 ) TL1051 ( osmXll ) TL1052 ( osmX12 ) Blue Yellow Pink Crimson Light yellow Light yellow Light yellow Yellow Yellow Yellow Yellow Yellow Crimson Crimson Crimson Crimson Crimson Crimson Crimson Light crimson 8 40 31 72 23 25 7 TL1044 ( osmX4 ) TL1045 ( osmXS ) TL1046 ( osmX6 ) TL1047 ( osmX7 ) TL1048 ( osmnX8 ) 113 22 22 5 7 3 TL1043 ( osmX3 ) TL1049 ( osmX9 ) TL1050 ( osmX1O ) 1 Yellow Yellow Yellow Light crimson Pink Light crimson 0.5 2 0.6 0.8 0.6 2 TABLE 4 .
Three-point mapping of proU , nalB , and zfi-9 : : TnJO loci % of transductantsb with phenotype : Transductiona Selection Kans Nal ' NalS Tetr Kans Nals Nals Tets Kanr Nalr Nalr Tetr Kanr Nals Nalr Tets Tetr 47 5 10 38 TL657 ( nalB zfl-9 : : TnJO ) TL501 ( proU1844 : : Mu d11734 ) TL501 ( proU ) 844 : : Mu d1734 ) - TL657 ( nalB zfi-9 : : TnlO ) 0 A P22 lysate of the first strain was used to transduce the second strain .
b One hundred transductants were scored in both crosses 24 44 Kanr 31 1 % of transductantsb with phenotype : Donor strain OsmX-Tetr OsmX-Tet ' OsmX + Tetr OsmX + Tets TL1041 ( osmXl ) 19 54 8 19 TL1042 ( osmX2 ) 16 84 0 0 TL1043 ( osmX3 ) 29 71 0 0 TL1044 ( osmX4 ) 44 56 0 0 TL1045 ( osmX5 ) 55 45 0 0 TL1046 ( osmX6 ) 39 59 2 0 TL1047 ( osmX7 ) 21 75 2 2 TL1048 ( osmnX8 ) 19 81 0 0 TL1049 ( osmX9 ) 24 76 0 0 TL1050 ( osmX10 ) 37 63 0 0 TL1051 ( osmXII ) 33 65 0 2 TL1052 ( osmX12 ) 29 68 0 3 a Phage P22 lysates of strains TL1041 through TL1052 ( osmX proUI844 : : Mu dI1734 ) were used to transduce strain TL839 ( zfi-9 : : TnlO ) to Kanr .
b One hundred transductants were scored from each cross .
The inheritance or lack of inheritance of the osmX mutation , denoted OsmX-and OsmX + , respectively , was scored by colony color on BTB plates .
The osmX mutations , as scored by colony color on BTB-lactose indicator plates , were found to be linked to the proU : : Mu dn1734 insertion ( Table 5 ) .
In seven strains , the osmX mutation was 100o linked to the Mu dI insertion , but with five donors ( strains TL1041 , TL1046 , TL1047 , TL1051 , and TL1052 ) we were able to recover osmX + proU : : Mu d11734 transductants .
Thus , the osmXl , osmX6 , osmX7 , osmXll , and osmX12 alleles can be separated from the Mu dI1734 phage , and therefore they can not be genetic alterations within the phage .
Of the strains that did not show complete linkage of the osmX mutation to the Mu dl1734 insertion , strain TL1041 ( osmXI ) was unique in that it yielded a higher proportion of osmX + transductants ( 27 % ) than the other four strains , which gave rise to osmX + transductants at a frequency of 2 to 4 % .
We carried out a second set of crosses to determine on which side of the proU1844 : : Mu d11734 insertion the osmX mutations were located .
In these crosses , strain TL456 ( proUJ844 : : Mu d B : : Tn9 ) was transduced to Tetr by strains of the genotype osmX proU1844 : : Mu d11734 zfi-9 : : TnlO ( strains TL1081 through TL1092 ) ( Table 6 ) .
With each donor strain , the majority of the Kanr transductants ( i.e. , those-that inherited proU1844 : : Mu d11734 ) were osmX + .
If the osmX mutations were between Mu dl1734 and zfi-9 : : TnlO , then we would expect that the majority of the Kanr transductants would also inherit the osmX mutation .
Since this was not the case , the results imply that the gene order is osmXproU : : Mu d11734 zfi-9 : : TnlO .
Gowrishankar determined that the proU operon of E. coli is transcribed clockwise , as depicted in the conventional linkage map ( 16 ) .
Using a modification of the method involving the construction of phage Mu dl-8-directed chromosomal duplications developed by Hughes and Roth ( 20 ) , we found that the proU operon of S. typhimurium is also transcribed in the clockwise direction , i.e. , toward the zfi-9 : : TnlO insertion ( M. T. Ederer and L. N. Csonka , unpublished results ) .
This result , together with our finding that the gene order is osmX proU zfi-9 : : TnlO , indicates that the osmX mutations the side of are on promoter the proU operon .
We tested whether it might be possible to recombine the proU + allele next to each of the osmX mutations in a set of in which the donor strain osmX + proU + transductions was zfi-9 : : TnlO and the recipient was osmX proU1844 : : Mu d11734 .
On the basis of the established gene order , osmX proU + transductants could arise in the above transductions as a result of a single crossover between the osmX mutations and the proU : : Mu dI1734 insertion .
If the recipient osmX strain carries proBA , putP , and proP mutations , osmX proU + transductants might be recognized by their ability to grow with 0.1 mM proline on medium 63 .
Accordingly , we transduced strains TL1061 through TL1072 ( AproBA putP proP osmX osmU : : Mu d11734 ) to Tetr by phage P22 grown on strain KS180 ( osmX + proU + zfi-9 : : TnlO ) .
With strain TL1061 ( osmXl ) as the recipient , 16 % of the transductants were Kan ' and able to grow on medium 63 with 0.1 mM proline , indicating that they had inherited the proU + allele and retained the osmXI mutation .
However , the osmXl mutant was again unique , because we were unable to obtain any osmX proU + transductants with any of the other strains .
Our failure to obtain such transductants with these strains could be due to the close linkage of the osmX2 through osmX12 mutations to the proU1844 : : Mu dI insertion or to the fact that some of the osmX mutations might be point mutations , deletions , or other rearrangements within the 5 ' coding region of the proU operon .
In the above crosses , with strains TL1061 ( osmXl ) , TL1062 ( osmX2 ) , TL1066 ( osmX6 ) , TL1069 ( osmX9 ) , TL1071 ( osmXII ) , and TL1072 ( osmX12 ) as donors , 1 to 5 % of the transductants inherited the osmX + allele while retaining the proU1844 : : Mu d11734 insertion .
On the basis of our postulated gene order , these rare transductants arose as a result of a quadruple crossover event in which the recipient inherited the TnJO element and the osmX + allele without loss of the intervening Mu dl1734 insertion .
We have previously noted that it was possible to separate the osmXl , osmX6 , osmX7 , osmXJl , and osmX12 alleles from the proU1844 : : Mu d11734 insertion .
Here we found this to be true of the osmX2 and osmX9 alleles , indicating that they also lie entirely outside the Mu dn1734 phage .
To test whether the osmX mutations might be in a gene that encodes a regulatory protein mediating the transcriptional control of the proU operon , we carried out complementation tests with them .
For these with respect to the proU1844 : : Mu dI and zfi-9 : : TnlO insertionsa % of transductantsh with phenotype : OsmX ~ OsmX + OsmX-OsmX + Kan5 Kan ' Kanr Kanr TL1081 ( osmXl ) 0 87 1 12 TL1082 ( osmX2 ) 2 48 15 35 TL1083 ( osmX3 ) 3 33 23 41 TL1084 ( osmX4 ) 0 62 1 27 TL1085 ( osmX5 ) 0 74 1 25 TL1086 ( osmX6 ) 0 57 15 28 TL1087 ( osmX7 ) 1 63 6 30 TL1088 ( osmX8 ) 1 45 3 51 TL1089 ( osmX9 ) 0 49 2 49 38 TL1090 ( osnmX1O ) 0 62 0 TL1091 ( osmXll ) 0 61 8 31 TL1092 ( osmX12 ) 2 62 7 29 a Phage P22 lysates of strains TL1081 through TL1092 ( osmX proU1844 : : Mu d11734 zfi-9 : : TnlO ) were used to transduce strain TL456 ( proUI844 : : Mu dl B : : Tn9 ) to Tetr .
b In each case , 100 transductants were scored .
The linkage of the osmX mutation was scored on BTB plates .
The recipient strain carried the proUI844 : : Mu dl B : : Tn9 ( Ampr Cmlr ) allele .
Consequently the Kan ' transductants ( that did not inherit the proUl844 : : Mu d11734 insertion ) retained the Ampr phenotype due to the proUl844 : : Mu dl B : : Tn9 insertion .
The chloramphenicol resistance of the transductants was not checked tests , the osmX mutations were transduced into strain TABLE 5 .
Linkage of the osmX mutations to proU and zfi-9 : : TnlO loci in phage P22 transductionsa TABLE 6 .
Determination of the position of the osmX mutations Donor strain TL819 , diploid for the osmX + proU + region due to a tandem chromosomal duplication ( see Materials and Methods ) .
We measured the 3-galactosidase activities of the osmX 4 ( proU-lacZ ) / osmX + proU + diploid strains .
The results obtained with the diploid strains grown in K medium without osmotic-stress are shown in Table 7 .
The 3-galactosidase activities of the diploid strains were very nearly the same as those of the corresponding osmX haploid mutants .
The,-galactosidase specific activity of diploid strains was similar to that of the corresponding haploids in K medium containing 0.5 M NaCl ( data not shown ) .
In particular , the osmX3 , osmX9 , and osmXlO alleles , which resulted in low , uninducible Igalactosidase activity in haploid strains ( Table 3 ) , could not be complemented by the wild-type osmX + allele in diploid strains to bring about induction of , B-galactosidase by 0.5 M NaCl .
Thus , our P-galactosidase assay results for diploid stains indicate that the osmXI through osmX12 mutations are cis-dominant over the osmX + allele .
We also determined whether any of the osmX mutations might be trans-dominant over the osmX + allele .
This was possible because the recipient strain in the above constructions also carried AproBA putP proP mutations , and if the osmX mutations could activate transcription of the proU + operon located in trans , then the osmX I. ( proU-lacZ ) / osmX + proU + diploids might be able to grow on medium 63 containing 0.1 mM proline .
We found that none of the diploid strains were able to grow on this medium , indicating that none of the osmX mutations are transdominant .
We constructed a diploid strain in which the osmXl mutation was cis to the proU + operon and osmX + was cis to proUI844 : : Mu d11734 .
This strain did not show elevated levels of P-galactosidase in K medium ( data not shown ) , confirming that the osmXl allele is not transdominant to osmX + .
Selection of spontaneous proU constitutive mutations .
The osmX2 through osmX12 mutations were obtained in a selection that entailed TnlO mutagenesis of a 4 ( proU-lacZ ) strain , followed by screening for derivatives that expressed the lac operon at an elevated level in the absence of osmotic-stress .
We found that the TnlO element in the mutants could be separated by osmX mutations without affecting the constitutive phenotype , implying that the TnlO insertions are not necessary for the ProU constitutivity .
However , the ISIO elements of TnJO can transpose independently of the whole transposon , and possibly the ISIO elements have inserted into the N-terminal region of the proU operon , thus provided a constitutive promoter for elevated expression of the lac operon ( 8 ) .
Such ISJO-mediated activation of lacZ operon transcription could be the reason that all of the osmX mutations were cis-dominant .
To assess whether the TnlO mutagenesis might have biased our selection of proU constitutive mutants , we isolated 48 additional proU constitutive strains that arose spontaneously .
Strain TL671 [ t ( proU-lacZ ) ] was streaked to single colonies on BTB-lactose medium .
This medium is relatively nutrient poor , and any mutants that can utilize lactose will have a growth advantage .
After 48 h of incubation , every slow-growing blue colony ( Lac - ) of strain TL671 contained yellow ( Lac ' ) sectors or papillae .
Cells from the yellow sectors were restreaked on LB-kanamycin plates and then retested on BTB-lactose plates .
In this manner , we obtained 48 additional mutants that expressed f ( proU-lacZ ) at an elevated level .
As was the case with the previous osmX mutants , these ProU constitutive mutants fell into two groups : 19 of the 48 retained the ability to increase expression of the proU operon in response to osmotic-stress , and the remaining 27 had lost this osmotic control .
The 3-galactosidase activity of the former group ranged from 8 to 12 nmol/min per mg of protein in K medium and exhibited a range of 31-to 58-fold induction in the presence of an additional 0.5 M NaCl .
The P-galactosidase activity of the latter group was in the range from 2 to 110 nmol/min per mg of protein in K medium and was the same or slightly reduced in the presence of an additional 0.5 M NaCl .
Under these conditions , the control osmX + strain had a basal level of 3-galactosidase of 0.6 nmol/min per mg of protein , which was induced approximately 230-fold by 0.5 M NaCl .
We transduced the spontaneous osmX mutations into diploid strain TL1095 ( osmX + losmX + ) and found all of them to be cis-dominant ( data not shown ) .
Thus , the mutations obtained spontaneously appear to be similar to the osmX2 through osmX12 alleles .
Therefore , the latter alleles probably are representative of the spontaneous ProU constitutive mutations , even though they were obtained in a strain that had been mutagenized by TnlO .
Since all of the proU constitutive mutations , except osmXl were obtained on BTB-lactose medium , the frequency of their formation is difficult to assess .
Strain TL671 [ osmX + 1 ( proU-lacZ ) ] was unable to grow on medium 63 with 5 mM lactose , but of the 48 spontaneous proU-constitutive derivatives , 40 were able to do so .
To obtain an estimate of the mutation frequency to proU constitutivity , we isolated spontaneous mutants of strain TL671 that could grow on medium 63 with 5 mM lactose .
Such derivatives arose at a frequency of 9.3 x 10-6 per cell .
Ten Lac ' derivatives were used to transduce strain TL1095 ( osmX + / osmX + ) to Kanr , and nearly 100 % of the transductants obtained with each donor expressed 4 ( proU-lacZ ) constitutively .
This result indicates that the mutations obtained in this last selection were representative of all other osmX mutations , and the frequency of appearance of these latter mutants ( -1 x 10-5 per cell ) is a valid estimate of the mutation rate to constitutive expression of the proU operon .
DISCUSSION We isolated 60 independent mutations that resulted in increased transcription of the proU operon in media of lo osmolarity .
One of these mutations , osmXI , was obtained in a selection that demanded that the resultant strain have increased proline transport activity via the ProU system in the absence of osmotic-stress .
The osmXl mutation apparently enables strain TL395 ( osmXl proU + AproBA putP proP ) to take up proline via the ProU system at a sufficient rate to support a growth-rate of 0.87 generations per h in medium containing 0.1 mM proline ( Table 2 ) .
Assuming that the proline content of S. typhimurium proteins is the same as that of E. coli B/r , which is 382 nmol of prolipe per mg of protein ( 21 ) , we can calculate that the rate of uptake of proline by the ProU system is at least 3.8 nmol/min per mg to support the observed growth-rate .
Cairney et al. ( 5 ) reported that the ProU system has a high affinity for glycine betaine ( Km = 1 , uM ) .
Because they were unable to detect significant uptake of proline via the ProU system , they concluded that proline is not an important substrate for this system .
Previously , using flow dialysis to mneasure proline transport , we demonstrated that the ProU system can accumulate proline in cells grown in media of elevated osmolarity ( 13 ) at rates comparable to the minimum rate we calculated above .
Although we have not carried out extensive measurements of the affinity of the ProU system for proline , we found it to be relatively poor and estimated the Km for proline to be > 0.02 mM ( 13 ) .
The ProU system involves a periplasmic binding protein with a high affinity for glycine betaine and little or no affinity for proline ( 27 ) .
It is possible that the proU operon encodes a second binding protein specific for proline .
An alternative possibility is that transport of proline by the ProU system might proceed without the involvement of a binding protein .
In this case , the greater affinity of the system for glycine betaine than for proline might derive from the fact that the periplasmic binding protein could channel glycine betaine to the inner membrane transport proteins of the ProU system , while proline would have to be bound directly by the inner membrane transport protein .
We obtained an additional 59 osmX mutations that resulted in elevated proU transcription by selecting derivatives of a f ( proU-lacZ ) strain that express,-galactosidase at an elevated level on BTB-lactose medium of low-osmolarity , We estimate that such mutations can arise spontaneously at a frequency of -10 -5 per cell ( see Results ) .
The mutations in all of the proU constitutive strains , including the osnmXI mutant , were closely linked to the proU operon and were cis-dominant over the osmX + allele in osmX + pro U+I osmX FD ( proU-lacZ ) diploids .
Thus , in this extensive search , we were unable to uncover an osmoregulatory mutation that was recessive or unlinked to the proU operon .
Our inability to obtain unlinked or recessive osmoregulatory mutations suggests that the osmotic regulation of the proU operon is probably not under negative control , because in that case mutations inactivating the repressor protein and thereby causing constitutive expression of pro U would be frequent .
However , it is possible that proU is subject to negative control but that the repressor protein is involved in some other central osmoregulatory function , and mutations inactivating the repressor might be deleterious under the growth-conditions we used .
At present , we do not have sufficient data to propose specific models for the transcriptional regulation of the proU operon , and we can not rule out a number of alternative mutations that might have generated the alterations resulting in elevated proU expression .
It is possible that in some of the mutants the increased basal transcription of the pro U-lacZ fusion might be the result of acquisition of a new promoter due to a point mutation , deletion , chromosomal duplication ( 2 ) , or integration of an IS element ( 8 , 33 ) .
P-Galactosidase activity of osmX ft ( proU-1acZ ) / osmX + proU + diploid strainsa , B-Galacto-1-Galacto-sidase sp actb .
r. sidase sp act ( nmol/min per Diploid strain ( nmol/min per mg of protein ) mg of protein ) os + , TL837 0.2 TL1113 0.2 osmXl TL1081 9 TL11O1 9 osmX2 TL1082 14 TL1102 8 osmX3 TL1083 5 TL1103 2 osmX4 TL1084 41 TL1104 44 osmX5 TL1085 74 TL1105 70 osmX6 TL1086 19 TL1106 11 osmX7 TL1087 22 TL1107 12 osmX8 TL1088 0.3 TL1108 0.1 osmX9 TL1089 0.6 TL1109 0.1 osmX10 TL1090 0.1 TL1110 0.2 osmXll TL1091 5 TL1111 2 osmXl2 TL1092 6 TL1112 6 a Haploid strains were osmX I ( proU-lacZ ) ; diploid strains were osmX F ( proU-lacZ ) IosmX + proU + .
b Determined on cells growing exponentially in K medium containing 10 mM glucose , 2 mM proline , and tetracycline .
osmX Haploid strain allele Also , there are a few examples of cis-acting regulatory proteins ( 28 ) , and the protein mediating the osmoregulation of transcription of proU might be of this type .
We have cloned the proU region from several strains carrying proU : : Mu dl-8 alleles ( D. G. Overdier and L. N. Csonka , unpublished data ) , and a comparison of the restriction map and nucleotide sequence of the osmX pro U region of osmX + and osmX strains will aid in defining the sites that are important for transcriptional control of the operon .
It should be noted that of the 60 strains with elevated basal expression of the proU operon , 24 retained the ability to induce it in response to osmotic-stress .
This result suggests that there is site within the transcriptional control or promoter region ofpro U , such as an operator or attenuator , that can be mutated readily to result in about an 8-to 12-fold increase in the basal level of transcription without abolishing completely the osmotic control of transcription of the operon .
In enteric bacteria , transcription of the kdp operon , encoding components of a high-affinity potassium transport system , is under osmotic control ( 24 ) .
However , expression of this operon follows a pattern of regulation different from that of the proU operon .
Osmotic stress brings about only a burst of transcription of the kdp operon , after which transcription returns to prestress levels ( 24 ) , whereas it results in a steady-state increase of transcription of the proU operon that is maintained as long as the osmotic-stress persists ( 5 , 12 , 16 ) .
This difference in the expression of the kdp and proU operons suggests that the two sets of genes respond to different inducing signals .
Laimins and Epstein ( 24 ) proposed that potassium is one of the important solutes that maintain cell turgor , and loss of turgor is the signal for the induction of the kdp operon .
According to this model , the increased synthesis of the high-affinity potassium transport system encoded by the kdp operon results in enhanced potassium uptake , which restores turgor and eventually shuts off transcription of the kdp operon .
Epstein ( 14 ) suggested that the increased intracellular potassium concentration in turn might be the signal for induction of the proU operon , and Sutherland et al. ( 38 ) recently published results supporting this model .
Epstein ( 14 ) and Sutherland et al. ( 38 ) proposed that a regulatory protein is involved in the transcriptional control of the proU operon and that this protein assumes different conformations in response to the potassium concentration or ionic strength of the cytoplasm and thereby modulates transcription of the gene .
Our selection procedures , which resulted in a large number of mutants with alterations in osmoregulation of the proU operon , failed to yield any mutations that were recessive to the wild-type allele .
Thus , we did not obtain any mutants that had an obvious defect in a trans-acting protein that might mediate the transcriptional control of the proU operon .
There are two possible reasons for our inability to obtain mutations in the structural gene of such a regulatory protein .
The selection procedure we used demanded increased transcription of the proU operon in the absence of osmotic-stress .
If the proU operon were regulated by a positive control mechanism , then mutations that caused the regulatory protein to be able to activate transcription of the proU operon in the absence of the regulatory signal might be rare in comparison to cis-acting mutations in the promoteroperator region of the operon .
Alternatively , if the protein regulating the expression of the proU were also involved in some other fundamental cellular process , then mutations in the regulatory protein might be deleterious .
We isolated a number of mutants that were unable to induce proU normally in media of elevated osmolarity and found that the had growth defects in a number of media of low-osmolarity ( M. T. Ederer and L. N. Csonka , unpublished observations ) , in accord with the possibility that mutations in the regulatory protein for the proU operon might have pleiotropic effects .
DiBlasio and Vinopal ( Abstr .
1986 , K123 , p. 214 ) reported the isolation of a temperature-sensitive lethal mutation resulting in elevated expression of the proU-lacZ fusion in E. coli .
This mutation lies in or very near the topA locus , which encodes topoisomerase I. Thus , DNA supercoiling , which is known to be involved in the transcriptional regulation of several operons ( 29 , 37 ) , might also be involved in regulation of the proU operon .
Since DNA supercoiling and the formation of DNA secondary structures are sensitive to ionic strength ( 1 ) , the osmotic control of transcription of the proU promoter could be exerted solely by the effects of the cytoplasmic ionic strength on the superhelical density or secondary structure of chromosome near the proU promoter , and conceivably there may not be a specific protein involved in the osmotic regulation of transcription of the proU gene .
ACJkNOWLEDGMENTS We thank John Roth , T. Silhavy , and R. Vinopal for helpful discussion and S. B. Gelvin and H. E. Umbarger for constructive criticism of the manuscript .
The work was funded by Public Health Service grant 1-RO1-GM-3194401 from the National Institutes of Health .
Anderson , P. , and W. Bauer .
Supercoiling in closed circular DNA : dependence upon ion type and concentration .
Anderson , R. P. , and J. R. Roth .
Tandem chromosomal duplications in Salmonella typhimurium : fusion of the histidine genes to novel promoters .
Boro , H. , and J. E. Brenchley .
A new generalized transducing phage for Salmonella typhimurium LT2 .
Cairney , J. , I. R. Booth , and C. F. Higgins .
Salmonella typhimurium proP gene encodes a transport system for the osmoprotectant betaine .
Cairney , J. , I. R. Booth , and C. F. Higgins .
Osmoregulation of gene expression in Salmonella typhimurium : proU encodes an osmotically induced betaine transport system .
Castilho , B. A. , P. Olfson , and M. J. Casadaban .
Plasmid insertion mutagenesis and lac gene fusion with mini-Mu bacteriophage transposons .
Christian , J. H. B. 1955 .
The influence of nutrition on the water relations of Salmonella oranienburg .
Ciampi , M. S. , M. B. Schmid , and J. R. Roth .
Transposon TnlO provides a promoter for transcription of adjacent genes .
Proline over-production results in enhanced osmotolerance in Salmonella typhimurium .
A third L-proline permease in Salmonella typhimurium which functions in media of elevated osmotic strength .
Davis , R. W. , D. Botstein , and J. R. Roth .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 12 .
Dunlap , V. J. , and L. N. Csonka .
Regulation of the osmotically stimulated transport of proline and glycinebetaine in Salmonella typhimurium , p. 115-1128 .
In J. L. Key and T. Kosuge ( ed .
) , Cellular and molecular biology of plant stress .
Alan R. Liss , Inc. , New York .
Dunlap , V. J. , and L. N. Csonka .
Osmotic regulation of L-proline transport in Salmonella typhimurium .
Osmoregulation by potassium transport in Escherichia coli .
Flowers , R. J. , P. F. Troke , and A. R. Yeo .
1977 , The mechanism of salt tolerance in Annu .
Identification of osmoresponsive genes in Escherichia coli : evidence for the participation of potassium and proline transport systems in osmoregulation .
proP-mediated proline transport also plays a role in Escherichia coli osmoregulation .
18 , Gowrishankar , J. , P. Jayashree , and K. Rajkumari .
Mo-lecular cloning of an osmoregulatory locus in Escherichia coli : increased proU gene dosage results in enhanced osmotolerance .
Grothe , S. , R. L. Krogsrad , D. J. McClellan , J. L. Milner , and J. M. Wood .
Proline transport and osmotic-stress response in Escherichia coli K-12 .
Hughes , K. T. , and J. R. Roth .
Directed formation of deletions and duplications using Mud ( Ap , lac ) .
Maale , and F. N. Neidhardt .
Growth of the bacterial cell , p. 110 .
Sinauer Associates , Inc. , Sunderland , Mass. 22 .
Itikawa , H. , S. Baumberg , and H. Vogel .
Enzymatic basis for genetic suppression : accumulation and deacylation of N-acetylglutamic y-semialdehyde in enterobacterial mutants .
Osmotic regulation and the biosynthesis of membrane-derived oligosaccharides in Escherichia coli .
Laimins , L. A. , D. B. Rhoads , and W. Epstein .
Osmotic control of kdp operon expression in Escherichia coli .
Le Rudulier , D. , and L. Bouillard .
Glycine betaine , an osmotic effector in Klebsiella pneuntoniae and other members of the Enterobacteriaceae .
H. , N. Rosebrough , A. L. Farr , and R. J. Randall .
Protein measurement with the Folin phenol reagent .
May , G. , E. Faatz , M. Villarejo , and E. Bremer .
Binding protein dependent transport of glycine betaine and its osmotic regulation in Escherichia coli K12 .
Menzel , R. , and M. Gellert .
Regulation of the genes for Escherichia coli DNA gyrase : homeostatic control of DNA supercoiling .
Menzel , R. , and J. Roth .
Identification and mapping of a second proline permease in Salmonella typhimurium .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 32 .
Perroud , B. , and D. Le Rudulier .
Glycine betaine transport in Escherichia coli : osmotic modulation .
Reynolds , A. E. , J. Felton , and A. Wright .
Insertion of DNA activates the cryptic bgl operon in E. coli K12 .
Sanderson , K. E. , and J. R. Roth .
Linkage map of Salmonella tryphimurium , edition VI .
Shkedy-Vinkler , C. , and Y. Avi-Dor .
Betaine-induced stimulation of respiration at high osmolarities in a halotolerant bacterium .
Somers , J. M. , G. L , Sweet ; and W. W. Kay .
Fluocitrate resistant tricarboxylate transport mutants of Salmonella typhi-murium .
Sternglanz , R. , S. DiNardo , K. Voelkel , Y. Nishimura , Y. Hirota , K. Beecher , L. Zumstein , and J. C. Wang .
1981t Mutations in the gene coding for Escherichia coli DNA topoisomerase I affect transpcription and transposition .
Sutherland , L. , J. Cairney , M. J. Ehnore , I. R. Booth , and C. F. Higgins .
Osmotic regulation of transcription : induction of the proU betaine transport gene is dependent on accumulation of intracellular potassium .
Yancey , P. H. , M. E. Clark , S. C. HIand , R. D. Bowlus , andG .
Living with water stress evolution of osmolyte systems .