No. 9 Nucleotide Sequence of the Transcriptional Control Region of the Osmotically Regulated proU Operon of Salmonella typhimurium and Identification of the 5 ' Endpoint of the proU mRNA DAVID G. OVERDIER , ' ERIC R. OLSON ,2 BRUCE D. ERICKSON , ' MARTINA M. EDERER , ' AND LASZLO N. CSONKAl * Department of Biological Sciences , Pu-rdiue University , West Lafcayette , Indiana 479061 ; Molecullar Biology Research , The Upjohn Company , Kalamazoo , Michigan 490012 ; and McArdle Laboratoryfvo/Canc er Resear ( ch , University of Wisc onsin , Madison , Wisceonsin 533706 Southern blot analysis of 15 proU transposon insertions in Salmonella typhimurium indicated that this operon is at least 3 kilobase pairs in length .
The nucleotide sequence of a 1.5-kilobase-pair fragment that contains the transcriptional control region of the proU operon and the coding sequences specifying 290 amino-acids of the first structural gene of the operon was determined .
The predicted amino-acid sequence of the product of this gene shows extensive similarity to the HisP , MaIK , and other proteins that are inner membrane-associated components of binding protein-dependent transport systems .
SI mapping and primer-extension analysis of the proU mRNAs revealed several species with different 5 ' ends .
Two of these endpoints are sufficiently close to sequences that have weak similarities to the consensus -35 and -10 promoter sequences that they are likely to define two transcription start sites .
However , we can not rule out the possibility that some or all of the 5 ' endpoints detected arose as a result of the degradation of a longer mRNA .
The expression of proU-lacZ operon fusions located on plasmids was normal in S. typhimurium regardless of the plasmid copy number .
The sequences mediating normal , osmoregulated expression of the proU operon were shown by subcloning to be contained on an 815-base-pair fragment .
A 350-base-pair subclone of this fragment placed onto a lacZ expression vector directed a high-level constitutive expression of , i-galactosidase , suggesting that there is a site for negative regulation in the proU transcriptional control region which has been deleted in the construction of this plasmid .
The ability to adapt to changes in the osmotic strength of the environment is an important trait for organisms .
To withstand fluctuations in the osmolarity of the environment , enteric bacteria accumulate low-molecular-weight solutes such as potassium , glutamate , proline , and glycine betaine ( for reviews , see references 12 and 32 ) .
These compatible solutes are thought to balance the internal osmolarity with the outside or to protect proteins from the deleterious effects of high ion concentration ( 23 , 48 ) .
In Sallmonella typhimia-rilum and Escherichia coli , the uptake of proline and glycine betaine is mediated by the ProP and the ProU systems ( 4 , 8 , 9 , 11 , 17 , 24 ) .
The activities of these two transport systems are enhanced by exposure of the cells to osmotic-stress .
In the ProP system , the osmotic control is primarily the result of a posttranscriptional modification of the transport protein ( s ) ( 8 , 17 , 38 ) , whereas in the ProU system , it is brought about by a several-hundred-fold increase in transcription of the proU operon ( 4 , 9 , 17 , 24 ) .
The ProU system of enteric bacteria belongs to a family of related transport systems which utilize a periplasmic binding protein ( 3 , 28 , 35 ) .
the nucleotide Analysis of the protein products ( 14 , 26 ) and sequence ( 25 ) of the proU operon of E. coli indicated that V , proW , and proX , this locus contains three genes , pro arranged in one operon .
The predicted amino-acid sequence showed high degree of similarity the of the proV gene a to inner membrane-associated components of number of a other binding protein-dependent transport systems , and the predicted amino-acid sequence of the proX gene matched the partial amino-acid sequence determined for the glycine betaine-binding protein by Barron et al. ( 3 ) .
The mechanism for sensing changes in osmolarity of the environment is poorly understood .
Epstein ( 19 ) and Sutherland et al. ( 51 ) suggested that the signal for transcription of proU is an increase in intracellular potassium ions .
In support of this hypothesis , Ramirez et al. ( 42 ) and Jovanov-ich et al. ( 33 ) demonstrated that 0.1 to 0.3 M potassium glutamate stimulated the expression of the proU operon in cell-free coupled transcription-translation systems .
DrugerLiotta et al. ( 16 ) isolated 60 regulatory mutations in S. tvphimuriimn which caused the ProU system to be expressed at elevated levels in the absence of hyperosmotic conditions .
Each of these mutations proved to be closely linked to the proU promoter and to be cis dominant in diploid strains , indicating that these mutations are most likely alterations of the promoter-transcriptional control region of the proU operon .
There are a number of conceivable reasons for the absence of tr-an .
s-acting regulatory mutations for proU , such the as possibility that the phenotypes of these mutants would be different from those expected or that the mutations might be lethal .
However , an interesting alternative reason for the be that the osmotic inability to obtain such mutations might control of proU transcription is mediated entirely by cisacting information contained in the proU promoter region ( 16 ) .
Higgins et al. ( 27 ) demonstrated that growth in elevated osmolarity increased the in-vivo supercoiling of reporter plasmids in E. ( oli and S. typhinuriuim .
These researchers isolated tr ( ins-acting regulatory mutations in E. coli which proved to be the topA and osinZ loci that encode proteins governing the supercoiling of DNA .
Because of these observed effects of supercoiling on the expression of the proU operon , Higgins et al. ( 27 ) proposed that the osmotic control of the transcription of proU is exerted by changes in the supercoiling of the proU promoter region .
To gain insights into the mechanism of the transcriptional control of the proU operon , we determined the nucleotide sequence of the S. typhimurium promoter region .
469 TL1487 TL1491 proUl872 : : Mu dl-8 recAl srl-2 : : TnlO TL1497 TL1510 proU1697 : : TnlO proP675 AputPA557 galE zcc-678 : : Tn5 Azjd-27 : : TnJO metA22 metE55 ilv ( ?
) xyl404 rpsL120 H1-bnml H2-enx hsdL6 hsdSA29 Fels - ( pRS415 , pDO10 , pDO40 , pDO41 , pDO70 , pDO81 , pDO83 ) recAl srl-2 : : TnJO ( pRS415 , pDO10 , pDO40 , pDO41 , pDO70 , pDO81 , pDO83 ) metA22 metE55 ilv ( ?
) xyl404 rpsL120 H1-bnml H2-enx hsdL6 hsdSA29 Fels - ( pDO101 ) recAl srl-2 : : TnJO ( pDO101 ) hisD9953 : : Mu dl-8 hisD9950 : : Mu dl-8 hisD9953 : : Mu d11734 Transform TL155 with pRS415 , pDO10 , pDO40 , pDO41 , pDO70 , pDO81 , pDO83 = Ampr Transform TL1463 with DNA from TL1519 to TL1525 = Ampr Transform TL155 with pDO101 = Ampr TL1519 to TL1525 TL1526 to TL1532 TL1533 TL1539 Transform TL1463 with DNA from TL1533 = Ampr J. Roth 30 31 TT7689 TT7692 TT10286 MATERIALS AND METHODS Media and growth-conditions .
Growth conditions for the strains , the composition of the rich-medium LB , and minimal-medium 63 ( M63 ) have been described previously ( 16 ) .
K medium was employed as the low-osmolarity medium .
Antibiotics were used at the following concentrations ( milli-grams per liter ) : sodium ampicillin , 25 for Mu dl lysogens and 100 for strains with multiple copies of the bla + gene ; chloramphenicol , 12.5 ; tetracycline , 15 ; spectinomycin hydrochloride , 1,000 for S. typhimurium and 50 for E. coli .
Strains were grown at 37 °C , except for lysogens carrying bacteriophage Mu dl , which were grown at 300C .
The genotypes and derivation of the S. typhimurium strains are presented in Table 1 .
E. coli host , 6 proUJ884 : : Mu dl ( B : : Tn9 ) proPJ654 AputPA557 zcc-628 : : Tn5 proU1884 : : Mu d11734 proUJ884 : : Mu dl-8 zfi-8 : : TnlO proU1884 : : Mu d11734 proU1697 : : TnJO proU1884 : : Mu d11734 proUJ872 : : Mu dl proP675 AputPA557 galE zcc-678 : : Tn5 A ; jd-27 : : TnlO metA22 metE55 ilv ( ?
) xyl-404 rpsL120 H1-bnml H2-enx hsdL6 hsdSA29 Fels - ( pDO4 , pDO5 , pDO39 ) recAl srl-2 : : TnlO proUJ884 : : Mu dl-8 recAl srl-2 : : TnlO recAl srl-2 : : TnlO ( pDO39 , pDO5 , pDO4 ) This laboratory ( 17 ) This laboratory ( 16 ) P22 ( TT7692 ) TL671 = Ampr [ His ' Kans ] P22 ( KS179 ) TL671 = Tetr [ Kanr ] P22 ( CH710 ) TL671 = Tetr [ Kanr ] TL334 made Tets on Bochner plates ( 7 ) Transform TL155 with pDO4 , pDO5 , pDO39 = Ampr P22 ( TL154 ) - TL1 = Tetr [ UV , ] P22 ( TL154 ) - TL1311 = Tetr [ UVS ] Transform TL1463 with DNA from TL1460 , TL1459 , TL1458 = Ampr Transform TL155 with pHJS21 , pDO57 = Spcr TL1458 to TL1460 TL1463 TL1467 TL1476 to TL1478 TL1479 , TL1480 TL1481 , TL1482 metA22 metE55 ilv ( ?
) xyl404 rpsL120 H1-bnml H2-enx hsdL6 hsdSA29 Fels - ( pHJS21 , pDO57 ) recAl srl-2 : : TnlO ( pHJS21 , pDO57 ) hisD9953 : : Mu dI1734 ( IacZ : : TnlO ) Transform TL1463 with DNA from TL1479 , TL1480 = Spcr P22 ( MS1202 ) TT10286 = Tetr [ Amps Kanr Lac-HisP22 ( TL1483 ) - TL1402 = Tetr [ Amps His ' Lac-P22 ( TL1485 ) - TL1 = Tetr [ Kanr ] P22 ( TT7689 ) TL1487 = Ampr [ Tets Kans His ' Lac ' ] P22 ( TL154 ) - TL1491 = Tetr [ UVS ] P22 ( CH710 ) TL1402 = Tetr [ Amps ] TL1483 TL1485 proU1872 : : Mu dI1734 ( IacZ : : TnlO ) proP675 AputPA557 galE zcc-678 : : TnS Aljd-27 : : TnJO proU1872 : Mu dI1734 ( IacZ : : TnlO ) pro UJ872 : : Mu dl-8 TL1487 TL1491 proUl872 : : Mu dl-8 recAl srl-2 : : TnlO TL1497 TL1510 proU1697 : : TnlO proP675 AputPA557 galE zcc-678 : : Tn5 Azjd-27 : : TnJO metA22 metE55 ilv ( ?
) xyl404 rpsL120 H1-bnml H2-enx hsdL6 hsdSA29 Fels - ( pRS415 , pDO10 , pDO40 , pDO41 , pDO70 , pDO81 , pDO83 ) recAl srl-2 : : TnJO ( pRS415 , pDO10 , pDO40 , pDO41 , pDO70 , pDO81 , pDO83 ) metA22 metE55 ilv ( ?
) xyl404 rpsL120 H1-bnml H2-enx hsdL6 hsdSA29 Fels - ( pDO101 ) recAl srl-2 : : TnJO ( pDO101 ) hisD9953 : : Mu dl-8 hisD9950 : : Mu dl-8 hisD9953 : : Mu d11734 Transform TL155 with pRS415 , pDO10 , pDO40 , pDO41 , pDO70 , pDO81 , pDO83 = Ampr Transform TL1463 with DNA from TL1519 to TL1525 = Ampr Transform TL155 with pDO101 = Ampr TL1519 to TL1525 TL1526 to TL1532 TL1533 TL1539 Transform TL1463 with DNA from TL1533 = Ampr J. Roth 30 31 TT7689 TT7692 TT10286 HB101 and JM103 were used for all in-vitro plasmid constructions .
Before plasmids were introduced into S. typhinlnriium strains of interest , they were passed through restrictiondeficient , modification-proficient S. typhimturilim TL155 .
All S. typhimurium strains were derivatives of LT2 ( Table 1 ) .
Phage P22 HT104 int3-mediated transductions were done as described by Davis et al. ( 15 ) .
The construction of plasmids is outlined in Table 2 .
Plasmid vector pBW2 ( 52 ) was used to clone the proUl884 : : Mu dl and proUl872 : : Mu dl fusions ; the plasmids derived from pBW2 are pDO1 , pDO4 , pDO5 , pDO10 , and pDO39 ( Fig. 1 ) .
Plasmids pDO1 and pDO39 were constructed by digesting with PstI chromosomal DNA from strains TL393 ( proU1884 : : Mu dl ) and TL334 ( proU1872 : : Mu dl ) , respectively , and ligating the inserts into the PstI site of pBW2 .
Selection for Ampr ensured that the 5 ' end of the bla gene in the Mu dl DNA was fused to the 3 ' end of the bla gene in pBW2 ( 52 ) .
These two plasmids contained at least 8 kilobase pairs ( kbp ) of S. typhimurium DNA in addition to the 9 kbp of DNA derived from the end of Mu dl that carries the lac genes .
Plasmid pDO1 contained two extra PstI fragments in addition to the fragment carrying the Mu dl fusion ; it was digested with PstI and religated , resulting in pDO4 , which had only one PstI insert .
Plasmid pRS415 ( 49 ) is derived from pBR322 and contains cloning sites in front of a promoterless lacZ + gene .
The plasmids derived from pRS415 are pDO40 , pDO41 , pDO70 , pDO81 , pDO83 , and pDO101 ( Fig. 1 ) .
Plasmid pHJS21 ( A. Sonenshein , personal communication ) is a low-copy lacZ expression vector derived from pSC101 ( 10 ) which contains a HindIll site in front of a promoterless lacZ + gene and a gene conferring spectinomy-cin resistance ( Spcr ) .
We constructed one derivative of pHJS21 , pDO57 , by inserting the 2.0-kbp HindIll fragment from pDO39 into the HindIll site of pHJS21 ( Fig. 1 ) .
Isolation of plasmid DNA , agarose gel electrophoresis , DNA ligations , and restriction enzyme digestions were done as described previously ( 34 ) .
Restriction enzymes were obtained from Bethesda Research Laboratories , Inc. ( Gaith-ersburg , Md. ) , Boehringer Mannheim Biochemicals ( India-napolis , Ind. ) , or New England BioLabs , Inc. ( Beverly , Mass. ) .
T4 DNA ligase and the Klenow fragment of DNA polymerase I were obtained from Bethesda Research Laboratories .
3-galactosidase assays were done with exponentially growing cells cultured in the indicated media , as described by Miller ( 37 ) .
The nonradioactive BlueGENE Kit was obtained from Bethesda Research Laboratories , and the directions were followed according to the manufacturer .
The probe used was biotin-labeled pDO10 .
Various fragments of the chromosomal DNA cloned in pDO39 were subcloned into M13mplO , M13mpll , or pUC19 and were used to determine the nucle-otide sequence of both strands of the proU transcriptional control region .
Sequencing of both double-stranded and single-stranded templates was done by the method of Sanger et al. ( 47 ) as modified by Zagursky et al. ( 54 ) for reverse transcriptase .
Oligonucleotides used for sequencing and primer-extension were synthesized by the methoxy-phos-phoramide triester coupling approach ( 6 ) and supplied by N. Thearault ( The Upjohn Co. ) .
RNA was prepared by the hot phenol method ( 46 ) from strain TL1 grown exponentially at 37 °C in K medium ( 0.07 osM ) or in K medium plus 0.5 M NaCl ( 1.0 osM ) .
For the proU probes , AvaI-or SstIl-digested pDO42 was labeled with [ y-32P ] ATP by polynucleotide kinase , cut with Hindlll , and gel purified .
A probe carrying the rpsU-dnaG-rpoD operon of S. typhimrurium was used as a control for the expression of a promoter which is not subject to osmotic control .
Plasmid pKKW10 carrying this operon ( 21 ) was labeled at a HindlIl site in the rpsU gene and cut with EcoRI .
Nuclease S1 analysis was performed by a procedure modified from that of Barry et al. ( 5 ) .
p.g of DNA probe were hybridized at 450C for 4 h , followed by digestion with 300 U of SI nuclease for 30 min at 370C .
The products were electrophoresed on a 5 % polyacrylamide sequencing gel .
Polynucleotide kinase and Si nuclease were obtained from New England BioLabs and Boehringer Mann-heim Biochemicals , respectively .
Primer extension of RNA was done by following the first-strand cDNA synthesis procedure of Polites and Marotti ( 41 ) .
Total cellular RNA was isolated from strain TL1311 that had been grown at steady state in M63 with 0.3 M NaCl ( 0.83 osM ) and was hybridized to 0.2 pLg of an oligonucleotide complementary to nucleotides 919 to 938 4696 OVERDIER ET AL. .
A HindlIl linker was attached , and this end was joined to the HindIlI site in the vector EcoRI-HindIII 0.8-kbp fragment of pDO88 in the EcoRI-HindllI sites of pDO86 ; see Fig. 1 A. Sonenshein , personal communication ; see Materials and Methods Reference 49 pDO67 pDO70 pDO81 pDO83 pDO86 pDO88 pDO101 pHJS21 pRS415 pDO41 H S V PvH I ~ ~ ~ ~ ~ ~ ~ ~ ~ I I I pDO57 H T S V pDO70 I I I I k 0 .
b 0.4 , 1.2 0.8 i S C pDO1 01 T T iI pDO81 ,83 FIG. 1 .
Restriction nuclease map of DNA fragments containing the transcriptional control region of the proU operon of S. tvphilinirilum .
The names of the plasmids carrying these inserts are indicated in the right-hand column .
and their construction is presented in Materials and Methods and Table 2 .
The open rectangles represent cloned S. tvphimiiiiriuiin DNA .
and the shaded rectangles represent an -0.5-kbp portion of phage Mu d DNA from the right end to an internal Hindlll site .
The origin for the distances in the restriction map is a Hindlll site ( 0.0 kbp ) .
The order of the two leftmost Pstl fragments of pDO1 is not known .
Note that the scale has been expanded for the inserts in plasmids pDO70 , pDO101 , pDO81 , and pDO83 .
Restriction sites are denoted as : C , Hincil ; E , EcoRI ; H , Hinidlll ; P , PstI ; Pv , Pv ` iull ; S , Sstll ; T , Taiql ; and V , EcoRV .
( see Fig. 3 ) that had been end labeled with [ -32 P ] ATP .
The hybridization mix ( final volume , 20 [ l ) was heated to 65 °C and allowed to cool slowly to 37 °C .
The hybridized products were extended with reverse transcriptase ( 41 ) and electro-phoresed on a 6 % sequencing gel .
Strains were grown overnight in the indicated media .
A sample was removed , and the cells were extracted with CHCI3 ( 2 ) .
The proteins in the resultant extract were precipitated with trichloroacetic acid ( 10 % ) at room temperature and then extracted with ether .
The pellet was taken up in 10 [ 1I of a buffer containing 8 M urea , 2 % sodium dodecyl sulfate , 10 mM Tris ( pH 6.8 ) , 10 % glycerol pDO5 From proU1872 : : Mud : E E PvH V SH I L I P V S V I I I pDO39 E V V S H S V Pv pDO40 I EV V S V Pv pDO41 H S V PvH I ~ ~ ~ ~ ~ ~ ~ ~ ~ I I I pDO57 H T S V pDO70 I I I I k 0 .
b 0.4 , 1.2 0.8 i S C pDO1 01 T T iI pDO81 ,83 FIG. 1 .
Restriction nuclease map of DNA fragments containing the transcriptional control region of the proU operon of S. tvphilinirilum .
The names of the plasmids carrying these inserts are indicated in the right-hand column .
and their construction is presented in Materials and Methods and Table 2 .
The open rectangles represent cloned S. tvphimiiiiriuiin DNA .
and the shaded rectangles represent an -0.5-kbp portion of phage Mu d DNA from the right end to an internal Hindlll site .
The origin for the distances in the restriction map is a Hindlll site ( 0.0 kbp ) .
The order of the two leftmost Pstl fragments of pDO1 is not known .
Note that the scale has been expanded for the inserts in plasmids pDO70 , pDO101 , pDO81 , and pDO83 .
Restriction sites are denoted as : C , Hincil ; E , EcoRI ; H , Hinidlll ; P , PstI ; Pv , Pv ` iull ; S , Sstll ; T , Taiql ; and V , EcoRV .
4 % 3-mercaptoethanol , and bromthymol blue and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis in a 12 % polyacrylamide gel .
RESULTS Physical mapping of proU locus .
We first cloned the proU1884 : : Mu dl insertion from strain TL393 into plasmid pBW2 ( 52 ) , generating pDO1 and its two subclones , pDO4 and pDO10 ( Table 2 ; Fig. 1 ) .
Plasmid pDO10 , which contains a 4-kbp HindlIl fragment spanning the junction of the Mu dl and S. typhimiuriiim DNA ( Fig. 1 ) , was used as a probe in Southern analysis to map the positions of 11 other pro U-lacZ-fusions generated by Mu dl ( 17 ) .
Chromosomal DNA was prepared from the strains , digested with HinidlIl , blotted .
and probed as described in Materials and Methods .
The results of this analysis are shown in Fig. 2 .
The proU : : Mu dl insertions proved to be located in a 2-kbp region , with the proU1872 insertion being closest to the proU promoter and proU1884 being the farthest away from the promoter .
Three TnlO insertions were also mapped , proU1655 : : TnlO and proU1697 : : TnlO , which confer a ProU-phenotype ( 9 , 11 ) , and zfi-8 : : TnIO ( 50 ) , which does not affect the ProU system ( M. M. Ederer , unpublished data ) .
The analysis ( Fig. 2 ) indicated that proUl697 : : TnlO is located upstream of all the other proU insertions tested .
When combined with the proU1884-lacZ fusion ( which is the most distal from the promoter ) , proUI697 : : TnlO conferred Lac-phenotype a ( data not shown ) , indicating that the 2.5-kbp region between these two insertions is transcribed as a single operon .
The Southern analysis also revealed that the zfi-8 : : TnlO insertion is located about 0.9 kbp upstream of the site of insertion of pro Ul697 : : TnlO .
Since this insertion does not affect the expression of the proU operon , the promoter for this operon is between the zfi-8 and proU1697 : : TnIO insertions ( Fig. 2 ) .
Deletion analysis of proU-IacZ plasmids .
Various fragments of plasmids carrying the proU1884-lacZ or proU1872-l1auZ tusion were subcloned into loi expression vectors , and the 3-galactosidase activities of strains carrying the resultant plasmids and grown in K medium and K medium plus 0.5 M NaCI were determined ( Table 3 ) .
Strains harboring plasmids which contain the region extending from the HinclI site ( 0.1 kbp , Fig. 1 ) up to or beyond the SstII site ( 0.9 kbp , Fig. 1 ) ( i.e. , pDO4 , pDO10 , pDO39 , pDO40 , pDO57 , pDO70 , and pDO101 ) can express 3-galactosidase , with at least a 30-fold increase in the synthesis of this enzyme upon osmotic-stress .
Plasmid pDO5 , which was made from pDO4 by deleting the region between the HindlIl sites at 0.0 and 4.3 kbp , and plasmid pDO41 , which was made from pDO40 by deleting the region between the SstII sites at -0.3 and 0.9 kbp ( Fig. 1 ) , can direct only a low-level synthesis of 3-galactosidase , which is not inducible by osmotic-stress .
These results suggest that the cis-acting transcriptional regulatory regions of the proU operon are contained between the HinclI site at 0.1 kbp and the SstII site at 0.9 kbp ( Fig. 1 ) .
The expression of the lacZ gene on plasmids pDO5 and pDO41 is most likely due to a promoter unrelated to the proU operon located in the region upstream of the SstII site at -0.3 kbp or in the vector .
These conclusions concerning the position of the proU promoter are in agreement with the mapping of the insertions ( Fig. 2 ) .
Regulation of proU-lacZ-fusions on plasmids .
The base level of expression of the lacZ gene in S. typhimurium strains carrying proU-lacZ-fusions on high-copy derivatives of pBR322 ( pDO4 , pDO10 , pDO39 , pDO40 , pDO70 , pDO101 ) was approximately 50 times higher than that seen with the fusions carried in single copy on the chromosome ( Table 3 ) .
Nevertheless,,-galactosidase could be induced quite normally by osmotic-stress when the proU-lacZ-fusions were carried on the high-copy plasmids , such that the ratio of induced to basal levels of P-galactosidase in the plasmidcarrying strains was similar to that of strains carrying the proU-lacZ-fusions on the chromosome .
The strain carrying the proU-lacZ fusion on pDO57 ( which has the origin of replication of the low-copy plasmid pSC101 ) had a basal level of,-galactosidase that was about twice as high as that of its parental chromosomal fusion , but its induction ratio was identical to those of the other strains ( Table 3 ) .
Sequence of proU promoter region .
The nucleotide sequence analysis of 1.5 kbp of DNA containing the proU promoter region is presented in Fig. 3 .
There is a short open reading frame ( ORF1 ) extending from the HindlIl site ( position 1 ) to position 303 , which could encode a protein of at least 101 amino-acids .
At the 3 ' end of ORF1 , there is a perfect inverted repeat of 12 nucleotides separated by 8 nucleotides ( positions 313 to 344 ) , followed by a run of AT base pairs .
This inverted repeat could be the terminator for the transcription of the mRNA for ORF1 , but the possibility that it is involved in the transcriptional regulation of the proU operon has not been completely ruled out .
There is a second long ORF extending from position 660 through the entire known sequence which could encode a protein of at least 290 amino-acids .
The predicted amino-acid sequence encoded by this long ORF has about 50 % amino-acid homology with the products of the hisP , malK , pstB , and oppD genes of S. typhimurium and E. coli ( Fig. 4 ) , with a conservation of sites believed to be involved in ATP binding ( 1 ) .
The observation that this putative gene product has such strong similarity to other inner membrane-associated transport proteins suggests that this gene is the structural gene for an analogous component of the ProU system .
In agreement with Dattananda and Gowrishankar ( 14 ) , we use the symbol proU to designate the entire operon that encodes the components of the S. typhimurium ProU system , and proV , proW , and proX to designate the genes in the respective order of location on the operon .
Analysis of the intergenic region between ORFi and the proV gene ( Fig. 3 ) from positions 300 to 680 revealed several imperfect direct or inverted repeat sequences at positions 417 to 447 , 568 to 580 , 591 to 612 , and 650 to 672 ( Fig. 3 ) .
The computer search also identified a region between nucleotides 443 to 465 showing similarity to the DNA gyrase-binding site within the oriC locus of S. typhimiiritim and E. coli ( 36 ) .
Analysis of this region with a program written to detect sequence-directed bending structures ( 18 ) revealed a region centered at position 450 which would be considered a bend .
The biological significance of any these above sequences is unknown .
Si nuclease mapping was used to determine the 5 ' endpoints of the proU mRNA ( s ) isolated from the wild-type strain TL1 grown in K medium or K medium plus 0.5 M NaCl .
The RNA samples were allowed to hybridize with two probes , both having an unlabeled endpoint at the HindlIl site ( position 1 , Fig. 3 ) and a labeled endpoint at the AvaI site ( position 855 , Fig. 3 ) or the SstII site ( position 903 , Fig. 3 ) , and the Si-resistant hybrid mole-cules were analyzed as described in Materials and Methods .
We detected hybridization to both proU probes with RNA isolated from cells grown in K medium plus 0.5 M NaCl but not with RNA from cells grown in K medium ( Fig. 5 ) .
The osmolarity of the medium did not have notable effects on the expression of the rpsU-dnaG-rpoD operon , used as a control .
The largest RNA species protected by the proU probe labeled at the SstII site ( position 903 ) was 306 to 310 nucleotides long , and the longest RNA species protected by the probe labeled at the AvaI site ( position 855 ) was 265 to 266 nucleotides long ( Fig. 5 ) .
These results are consistent with each other within the limits of resolution of Si mapping , and they indicate that the 5 ' endpoint of a proU mRNA is located around positions 594 to 598 ( denoted Si in Fig. 3 ) .
Several shorter proU mRNA species were resolved by the Si mapping .
The most prominent of these shorter RNA hybrids was one of -88 nucleotides , detected with the probe labeled at the SstII site ( denoted S2 in Fig. 3 ; see Fig. 5 ) .
The 5 ' endpoint of this RNA corresponds to position -815 , which is well within the proV gene .
It is not certain whether this position represents a processing point of the mRNA or a promoter within the proU operon .
( This short RNA was not detected with the AvaI probe , perhaps because the RNA-DNA duplexes formed with this probe were too short to be stable during the hybridization and nuclease Si treatment .
) The proU mRNA was also analyzed by primer-extension .
In this case , RNA was isolated from strain TL1311 ( proU1884 : : Mu dl-8 ) grown in M63 and M63 with 0.3 M NaCl and hybridized to an oligonucleotide primer complementary to nucleotides 919 to 938 ( Fig. 3 ) .
We detected five cDNAs ( Fig. 6 ) with apparent endpoints around positions-544 , 586 , 596 , 605 , and 651 ( denoted E3 through E7 in Fig. 3 ) .
Two of these endpoints , namely , at positions 586 and 596 , are in excellent agreement with the approximate mRNA endpoint ( s ) resolved by Si mapping , so that the composite results of the two techniques are consistent with one transcription start site for the proU mRNA being around position 596 and possibly a second one around position 586 .
However , the possibility that some or perhaps all of the observed 5 ' endpoints are due to degradation of a longer mRNA rather than transcription initiation can not be excluded .
The 5 ' endpoint of the longest proU mRNA detected by primer-extension ( position 544 , Fig. 6 ) was not seen in the Si nuclease analysis ( Fig. 5 ) , and therefore one ( perhaps the only ) transcription start site for the proU operon may be at or upstream of position 544 .
However , if this position corresponds to the 5 ' endpoint of a species of proU mRNA , it is not clear why we did not detect it with the Si nuclease analysis .
An overestimate of the length of an mRNA in primer-extension could result from the formation of hairpin structures at the 3 ' end of the cDNA serving as primers for the synthesis of a second DNA strand .
Promoters in opposite directions .
To examine further the cis-acting sequences that are required for osmotic control of transcription of the proU operon , we subcloned the 275 base-pair ( bp ) HindIII-TaqI fragment ( positions 1 to 275 , Fig. 3 ) , the 350-bp TaqI-TaqI fragment ( positions 275 to 625 , Fig. 3 ) , and the 122-bp TaqI-TaqI fragment ( positions 625 to 747 , Fig. 3 ) into plasmid pRS415 .
Strains carrying the plasmids with the HindIII-TaqI fragment and the 122-bp J. BACTERIOL .
C ' ) 0 0 0 0 0 1 ¬ U ) 0 0 0 0 0 0 rco proU : : Mud 0 0 0 0 0 X r. u ) U ) 40 0 ) 0D proU : : TnlO proU + N v I1 v v v v v v I , , t : : : .
: : : : : - : : - : : : I : : : : - t t Mud E s H E S E S proV kb :O I 2 3 4 FIG. 2 .
Location of proU : : TnlO orproU : : Mu dl insertions .
The locations of these insertions were determined by Southern blot analysis of chromosomal DNA of strains carrying the given proU insertion mutations , with plasmid pDO10 as the probe ( see Materials and Methods ) .
The restriction map represents the proU region from strain TL346 , which carries the proU1884 : : Mu dl insertion .
The reference point for the restriction map and the abbreviations for restriction sites are described in the legend to Fig. 1 .
The top set of numbers above the restriction map are the allele numbers of independent proU : : Mu dl insertions , the middle set are the allele numbers of pro U : : Tn1O insertions , and the lowest number refers to a TnlO insertion outside the proU locus .
Expression of , B-galactosidase in V ( proU-lacZ ) strains,3-Galactosidase ( nmol/min per mg of protein ) ' in K medium containing : Ratio No NaCI 0.5 M NaCI TL1526 pRS415 None 2 2 1 TL1481 pHJS21 None 1 1 1 TL1467 None proU1884 0.5 160 320 TL1478 pDO4 pooU1884 24 720 30 TL1527 pDO10 pr-oU1884 38 4,960 130 TL1477 pDO5 prooU1884 28 45 2 TL1497 None pi-oU1872 2 470 235 TL1476 pDO39 proU1872 35 7,490 210 proU1872 105 14,000 130 TL1528 pDO40 TL1529 pDO41 pi-oU1872 4 4 1 TL1533 pDO101 proU1872 100 7,700 77 TL1530 pDO70 pr-oU1872 375 12,100 32 TL1531 pDO81 proU1872 1,490 2,490 1.7 TL1532 pDO83 pr-oU1872 3,240 12,900 4 TL1482 pDO57 pr-oU1872 4 1,750 440 ' The,-galactosidasc activities of the strains growing exponentially for at least six generations in the indicated media were measured as described in reference 37 Strain Plasmid Fusion OVERDIER ET AL. .
VR HisP 2 MSEN JLIIVIDJIIKIRYIGIGHEVL KGVSLQARAG VISWIIGISSGSGKS T F-LR MaIK 1 MAS L QNVT -- K AW GIEVVVSK INLD I HEGEFVVFVIGIPSG [ GKSTL-LR PstB 6 T A P SK I VRNVR N FYYGKFHALLKNINLLJDIJAKN VTAFIGPSG GKS TL-LR OppD 16 LLEVNDLRVTFATPDGDVTAVNLDLNFTLRA G JTLGWVGESGSGKS S R L R RbsA ( N ) 1 MEALLQLKGIDKAFPGVKALS-GAAjNVYPIG V M A L V G E N GAG K STM-MK RbsA ( C ) 2S2 A PG DIR L K V D N L C G PGVN DVS -- FT L R K I L G V S G L M GA R T E L-M K ProV L R [ LIEPWR [ ] GQ .
-- KISDAE.LREV [ RRKKIAMVFQSFA IlisP C IJFtLIEKPSEIGIAI IVNGQNWINLVR [ KDGQLKVADKNQLRLLRITRLT-MVFQIIFN MaIK MI AG LETISGDLF I-GE K -- RMNDTP -- PAE ------- RG.VG.MVFQSYA PstB TF KMFELYPEQ .
EEGDN I LTNSQD -- I ALLRAKVG-MVFQKPT OppD-CGLLM ALATN .
- RE ILNLPERELNT -- R R A E Q I S-MIFQL DPM RbsA VL ITGIYTRDAIG TLLWL GKETTFTGPKSSQEA - .
GIGIIIIQE -- LNLIP Q-LT RbsA VYGA LPR2SGYVTLDGIlEVVTRS PQDGLAN -- G IVY I SEDEKRDG-LEjLG-MS llisP LWHMJT -- VLENVMEAPIVL-IGALSKH-DAR RALKY AKV YGIDERAQGA -- ViE MalK LYtrLS -- VAEN-MSFGLPA - [ GJAKKE-VINQRVNQ-VAEVJLQLAIILLDRK PJKA PstB P F PMS I-Y D N I A F G V R L F KE-SR A DM D ER V QIT K A A L WN E T K D K LII-Q S G Y S OppD TSLNPYMREGEQLMEVLMIIK-SGMSKAEAFEWSVRMLDAVKMPEARKRMKMY IIE RbSA IAENIF -- LG -- REFVN-RF-GKIDWKTMY AEEDK L [ AKLNLRFKSD .
KLVGD RbsA V K E N S -- L TA L R Y F SR-AG-ES L K HAD EQ Q A VS D F IL R F N V K T PS ME-Q AI G L Site B ProV L S G GMQR VGLARALAINNDI MDEAFSALDPILIRT [ FEIM Q D E-L V KLQ A K 11 .
Qj lfisP LSGGQ QRVS IAR LAMEIP  DVLLFDEPTSALDPELVGWVLRI-MQQILI-AEE-GK MaIK L S G G QERQ R V A I RT A E  P  S V F L L D E  P LS [ ] L D A AL R V Q-MR I E I SRIL  - IK R LG PsIB LSGGGQ QRLC I ARIA I RPIEVLLLDEPC ALD P ISTGRIEEL-ITEIL-KQDY -- OppD FSG GM QR VMIAMALLCRLPJKLLIADEPTTALD VTVQAQIMTL LNEILI-KREFNT RbsA SWjGD QMVE I AKVLS FESKV I IMDE  PTDALTDTETES LFRV-I R E [ LJ-K S Q-GRI RbSA LSGGN KAI GLMTRKV ILDEP TRGV VGAKKEI YQL-I NQ F-KA D-G L ProV T I V F I S LDEAMRW1G [ ERWA IJQN-GEVVQVGTPD EI LNN -- ] AN DYV R T llisP TMV I MG ARIIVSSHVI FLHQGKI EEEIGDPEQVFGN -- PQSPRLQQ MaIK TMIYVTII QV AIMTLADKWVVLDAGRVAQVGJK [ JLAVPLSG -- RPFCRR IYR PstB T VVI V TIfNM Q QARCSIDIHTAF [ YLGELI-EFSNTDDLFT -- KP-A K K Q T E OppDAI IM ITfLGVVA GICIDKVLV [ YAGRTM-EYGKARDVFYQPVIlPYSIGLLNA RbsA G IVY I SHRMKI FEWICCDDVTV FRDGQF I A ER EVA S LTEDS LIEMMVGRKLED RbsA S I L VS EMP VLGMSDRE I VRIIlEJHLSGEFTREQATQEVLMAAAVG-KPNR FIG. 4 .
The proV gene product resembles inner membrane-associated components of other binding protein-dependent transport proteins .
The predicted sequence of the first 290 amino-acids of the proV gene product is compared with the amino-acid sequences of other transport proteins ( 1 ) , starting with the residues indicated in the top set of lines .
Boxes are drawn if the proV gene product is identical at that position to at least two other gene products .
The residues at sites A and B have been proposed to be involved in ATP binding ( 1 ) .
TaqI-TaqI fragment were white on MacConkey-lactose agar containing 0.2 M NaCl , indicating that these two fragments did not have a promoter ( data not shown ) , and they were not studied further .
However , the strain carrying the plasmid with the 350-bp TaqI-TaqI fragment ( designated pDO83 ) proU mRNA .
RNA was isolated from the wild-type strain TL1 , and nuclease S1 protection experiments were done as described in Materials and Methods .
( A ) Probe was the HindIll-Av ` aI proU fragment ( positions 2 to 855 ) labeled at the Av ` aI site ; ( B ) probe was the HindIII-Sstll proU fragment ( positions 2 to 903 ) labeled at the SstI site ; ( C ) control for a promoter not subject to osmoregulation with a probe carrying the rpsU-dnaG-rpoD operon of S. tphimi-rium ( 21 ) .
In each panel , the lanes are results with probe only .
without RNA or nuclease SI ( lane a ) , probe without RNA and treated with nuclease Si ( lane b ) , probe hybridized with RNA isolated from cells grown in K medium and treated with nuclease Si ( lane c ) , and probe hybridized with RNA isolated from cells grown containing NaCI and treated with nuclease Si in K medium 0.5 M ( lane d ) .
The three most abundant proU mRNA species are highlighted with arrowheads .
MW , Molecular weight standards ; the lengths of fragments in base pairs are indicated .
exhibited a Lac ' phenotype on MacConkey-lactose agar .
The 3-galactosidase level conferred by this plasmid in cells grown in M63 was at least 10-fold greater than that seen with the parental plasmid pDO101 ( Table 3 ) , and it was not increased further by osmotic-stress .
Interestingly , when the same TaqI-TaqI fragment was inserted into plasmid pRS415 in the direction opposite to the transcription of IlacZ from the proU promoter ( pDO81 ) , it could still direct the synthesis of high and unregulated levels of P3-galactosidase ( Table 3 ) .
These results suggest that there is a promoter in the TaqITaqI fragment ( from positions 275 to 625 ) pointing away from the proU operon which can be recognized on a highcopy plasmid .
Analysis of periplasmic proteins .
Barron et al. ( 3 ) and Faatz et al. ( 22 ) reported that in E. coli , the structural gene for the glycine betaine-binding protein is the first gene of the proU operon .
Our DNA sequencing revealed that the first gene of the proU operon in S. tvyphimiurilum encodes an inner mem-brane-associated protein rather than a periplasmic binding protein and therefore seems to contradict this conclusion .
To settle this discrepancy , we analyzed periplasmic protei llisP LWHMJT -- VLENVMEAPIVL-IGALSKH-DAR RALKY AKV YGIDERAQGA -- ViE MalK LYtrLS -- VAEN-MSFGLPA - [ GJAKKE-VINQRVNQ-VAEVJLQLAIILLDRK PJKA PstB P F PMS I-Y D N I A F G V R L F KE-SR A DM D ER V QIT K A A L WN E T K D K LII-Q S G Y S OppD TSLNPYMREGEQLMEVLMIIK-SGMSKAEAFEWSVRMLDAVKMPEARKRMKMY IIE RbSA IAENIF -- LG -- REFVN-RF-GKIDWKTMY AEEDK L [ AKLNLRFKSD .
KLVGD RbsA V K E N S -- L TA L R Y F SR-AG-ES L K HAD EQ Q A VS D F IL R F N V K T PS ME-Q AI G L Site B ProV L S G GMQR VGLARALAINNDI MDEAFSALDPILIRT [ FEIM Q D E-L V KLQ A K 11 .
Qj lfisP LSGGQ QRVS IAR LAMEIP  DVLLFDEPTSALDPELVGWVLRI-MQQILI-AEE-GK MaIK L S G G QERQ R V A I RT A E  P  S V F L L D E  P LS [ ] L D A AL R V Q-MR I E I SRIL  - IK R LG PsIB LSGGGQ QRLC I ARIA I RPIEVLLLDEPC ALD P ISTGRIEEL-ITEIL-KQDY -- OppD FSG GM QR VMIAMALLCRLPJKLLIADEPTTALD VTVQAQIMTL LNEILI-KREFNT RbsA SWjGD QMVE I AKVLS FESKV I IMDE  PTDALTDTETES LFRV-I R E [ LJ-K S Q-GRI RbSA LSGGN KAI GLMTRKV ILDEP TRGV VGAKKEI YQL-I NQ F-KA D-G L ProV T I V F I S LDEAMRW1G [ ERWA IJQN-GEVVQVGTPD EI LNN -- ] AN DYV R T llisP TMV I MG ARIIVSSHVI FLHQGKI EEEIGDPEQVFGN -- PQSPRLQQ MaIK TMIYVTII QV AIMTLADKWVVLDAGRVAQVGJK [ JLAVPLSG -- RPFCRR IYR PstB T VVI V TIfNM Q QARCSIDIHTAF [ YLGELI-EFSNTDDLFT -- KP-A K K Q T E OppDAI IM ITfLGVVA GICIDKVLV [ YAGRTM-EYGKARDVFYQPVIlPYSIGLLNA RbsA G IVY I SHRMKI FEWICCDDVTV FRDGQF I A ER EVA S LTEDS LIEMMVGRKLED RbsA S I L VS EMP VLGMSDRE I VRIIlEJHLSGEFTREQATQEVLMAAAVG-KPNR FIG. 4 .
The proV gene product resembles inner membrane-associated components of other binding protein-dependent transport proteins .
The predicted sequence of the first 290 amino-acids of the proV gene product is compared with the amino-acid sequences of other transport proteins ( 1 ) , starting with the residues indicated in the top set of lines .
Boxes are drawn if the proV gene product is identical at that position to at least two other gene products .
The residues at sites A and B have been proposed to be involved in ATP binding ( 1 ) .
Nucleotide sequence of the proU transcriptional control region .
The nucleotide sequence of the strand containing the information corresponding to the putative proU mRNA is presented .
The origin for the numbering of the nucleotide position is at the HindIll site ( base pair 1 ) , which corresponds to the HindIlI site at 0.0 kbp in Fig. 1 and 2 .
Perfect or near perfect inverted repeats are highlighted by the inverted arrows below the sequence .
The boxes Si and S2 ( positions -596 and 815 , respectively ) denote the 5 ' endpoints of the two most abundant mRNA species detected by nuclease Si mapping ( see Fig. 5 ) , and the boxes E3 through E7 ( positions -544 , -586 , -596 , 605 , -651 , respectively ) denote the 5 ' endpoints of five mRNA species detected by primer-extension ( see Fig. 6 ) .
There is a sequence showing similarities to the proposed gyrase-binding site ( 36 ) at positions 443 to 465 , highlighted by a line above the nucleotide sequences .
The probable translation start site of the proV gene is at the ATG at positions 660 to 662 .
The predicted amino-acid sequence of the first 290 amino-acids of the proV gene product and the last 101 amino-acids of a protein of unknown function ( ORFi ) upstream of the proU operon are shown with the single-letter code below the nucleotide sequence 4702 OVERDIER ET AL. .
S. \ AAstt1 Ava VIW Ai t ld ; J. h t id : 1 bl ( d1 00 632-527 ¬ 404 ¬ - i 309-238 ¬ 201-180 ¬ 160 ¬ e 'S * !
' p ' p ' p ' p 4 ) ( ) 1-123 ¬ 9 ¬ - ~ ~ ~ ~ 4 * 40 61-0 - / FIG. 5 .
Nuclease Si determination of the 5 ' endpoints of the FIG. 6 .
Primer extension of the 5 ' of analysis endpoints proU mRNA .
RNA was isolated from strain TL1311 dl-8 ) ( proUJ844 : : Mu in M63 containing 0.3 M NaCI , and primer-extension was grown done as described in Materials and Methods .
The was an primer oligonucleotide complementary to nucleotides 919 to 938 ( Fig. 3 ) .
Five mRNA that were resolved are highlighted with arrows .
species The results of the DNA determination are in sequence presented lanes G , A , T , and C ; this sequence is for the strand complementary to that given in Fig. 3 .
The left-hand column of numbers indicates the nucleotide position as shown in Fig. 3 .
, se we * S ; 5Cjj F .
'S ' > w -- * '' * 42kDa - s cents , xe , ; iL .
jLjji Niil $ ; iEiSiVe ikiZ .
a ` ge `` F 2lkDa t : RtAg!s * , , ev .
Periplasmic proteins of proU insertion mutants .
The periplasmic proteins of the indicated strains were isolated and analyzed as described in Materials and Methods .
The strains were grown in M63 ( - ) or M63 plus 0.3 M NaCl ( + ) .
The 33-kilodalton ( kDa ) glycine betaine-binding protein ( GBBP ) is present only in extracts of strain TL1 grown in M63 plus 0.3 M NaCI .
extracts from 24 S. typhimurium strains carrying various proU insertion mutations .
Each of them proved to lack the periplasmic glycine betaine-binding protein .
The results obtained with representative strains TL334 ( proU1872 : : Mu dl ) , TL346 ( proU1884 : : Mu dl ) , and TL1510 ( proU1697 : : TnJO ) grown in M63 with 0.3 M NaCl are shown in Fig. 7 .
A 33-kilodalton periplasmic protein was detectable in the wildtype strain TL1 grown in the presence of 0.3 M NaCl but not in any of the other strains .
Because the insertions in the strains used are separated by 2.5 kbp , we conclude that the structural gene for the glycine betaine-binding protein is not the first gene of the proU operon in S. typhimurium but rather is located in a region at least 2.5 kbp downstream from the promoter .
DISCUSSION There is excellent agreement between our nucleotide sequence results for the proU promoter region of S. typhi-murium and those obtained by Gowrishankar ( 25 ) for the E. coli counterpart .
There is a 79 and an 87 % nucleotide sequence identity and an 85 and a 98 % amino-acid sequence identity for ORF1 and the proV gene , respectively , in the regions where the sequences can be compared .
Gowrishan-kar ( 25 ) concluded that the gene for the glycine betaine-binding protein , proX , is the last gene of the proU operon , and he noted , as we have , that the first gene of the operon , proV , encodes an inner membrane-associated protein .
Last , Gowrishankar ( 25 ) placed the 5 ' endpoints of two proU mRNAs at 52 and 60 nucleotides upstream of the probable translation start of the proV gene .
These positions correspond to nucleotide positions 607 and 599 of the S. typhimu-rium sequence ( Fig. 3 ) , and therefore the endpoints E5 and E6 we determined by primer-extension are in close agreement with the results of Gowrishankar ( 25 ) .
In several respects , there are differences between our results and those of Gowrishankar .
Gowrishankar ( 25 ) found one set of additional mRNA endpoints , 178 to 180 and 213 to 215 nucleotides upstream of the probable translation start of the proU gene , which we did not detect .
More important , we found that cloning of the S. typhimurium proU operon on high-copy plasmids did not affect the osmotic control of its expression in S. typhimurium hosts ( Table 3 ) , whereas the E. coli pro U operon no longer responded to the normal osmotic control when placed on high-copy vectors in E. coli strains growing exponentially in media of high-osmolarity ( 14 ) .
Last , Dattananda and Gowrishankar ( 14 ) reported that in E. coli , plasmids which carry the proV gene and only portions of the pro W or proX genes resulted in an osmosensitive phenotype .
We did not observe any osmosensitivity with any of the S. typhimurium clones we tested in either E. coli or S. typhi-muritum hosts .
The high degree of similarity between the E. coli and S. typhimurium nucleotide sequences is not entirely conserved for the intergenic region between ORF1 and the proV gene .
Figure 8 presents a comparison of the nucleotide sequences of this region from the two organisms .
There is only a 68 % nucleotide sequence identity in the region corresponding to positions 304 to 662 of the S. typhimurium sequence .
Interestingly , the sequence conservation between the organisms is stronger in the vicinity of the proU promoter region , as there is an 85 % nucleotide sequence identity for the nucle-otides corresponding to positions 461 to 655 of the S. typhimurium sequence , but only a 60 % identity for positions 307 to 460 .
Because the 12-nucleotide inverted repeat structure at positions 313 to 324 and 333 to 344 of the S. typhimurium sequence is lacking from E. coli , we conclude that this structure is not important for the transcriptional control of the proU operon .
A scan of the nucleotide sequences upstream of the first structural gene with the algorithms of Mulligan et al. ( 40 ) revealed a number of sequences that may serve as possible promoters for the proU operon .
Thus , the sequences resembling the E. coli consensus promoter ( 29 ) are ( i ) TTGTCT ( -35 , positions 564 to 569 , Fig. 2 ) and TAGGGT ( -10 positions 586 to 591 ) and ( ii ) TTCAGG ( -35 , positions 557 to 562 ) and TATGTT ( -10 , positions 581 to 586 ) .
These might be possible promoters for the mRNAs detected by SI mapping and primer-extension at positions 597 and 585 , respectively .
Likewise , the sequences ATCACA ( positions 514 to 519 ) and AATATT ( positions 536 to 541 ) might be the -35 and -10 sequences for an mRNA resolved by primer-extension to have a 5 ' endpoint at position 544 .
However , the differences between these sequences and the consensus E. coli -35 and -10 sequences are sufficiently great to suggest that they do not constitute strong promoters in the absence of a positive regulatory protein .
The TaqI-TaqI fragment from positions 275 to 625 apparently contains at least one promoter for mRNA transcribed in the orientation opposite that of the proU operon .
There does not appear to be a strong promoter in this direction either ; the computer analysis revealed the hexamers TAGAAA ( positions 612 to 607 ) and TAACAT ( positions 587 to 582 ) , which may be the -35 and -10 sequences of a weak promoter .
Both the basal and the induced levels of the expression of the proU regulatory sequences fused to the lacZ gene on both high-and low-copy plasmids were positively correlated with the copy number of the vectors used .
The induction ratio , however , remained nearly constant for all the plasmids that carried the entire proU promoter region .
The regulation of DNA supercoiling is not understood sufficiently to be able to predict it for a given plasmid , and it is possible that the localized supercoiling of the proU insert is not influenced by the various vectors used .
However , because supercoiling is determined by a number of factors , including the number of active promoters on a replicon ( 53 ) , it is unlikely that the supercoiling of the proU region on the chromosome is the same as on the low-copy plasmid pSC101 or the high-copy plasmid pBR322 .
Thus , the nearly invariant induction ratios of the proU operon when it is carried on the various replicons do not seem to be readily consistent with the proposal that supercoiling is the main regulatory signal for the transcriptional control of the proU operon ( 27 ) .
2lkDa t : RtAg!s * , , ev .
Periplasmic proteins of proU insertion mutants .
The periplasmic proteins of the indicated strains were isolated and analyzed as described in Materials and Methods .
The strains were grown in M63 ( - ) or M63 plus 0.3 M NaCl ( + ) .
The 33-kilodalton ( kDa ) glycine betaine-binding protein ( GBBP ) is present only in extracts of strain TL1 grown in M63 plus 0.3 M NaCI .
The observation that the proU operon can be expressed efficiently under conditions of hyperosmotic-stress despite the lack of a promoter with strong similarity to consensus -35 and -10 sequences raises some questions concerning the factors involved in the recognition of the proU promoter by RNA polymerase .
There are no concrete data bearing on this point , but it is possible that some positively acting transcriptional regulatory factor or the degree of supercoiling directs the RNA polymerase to the proU promoter under conditions of high-osmolarity .
The result that the TaqI-TaqI fragment from positions 275 to 625 can confer a relatively high level of constitutive expression of the lacZ gene in pDO83 ( Table 3 ) suggests that there is some negative cisacting site for proU transcription ( e.g. , a target for a repressor protein , a site for transcription attenuation or termination , or a site of mRNA processing ) upstream of position 275 or downstream of position 625 .
However , we have no data to rule out the possibility that an inadvertent promoter had been generated as a result of fortuitous juxtaposition of insert and vector sequences in pDO83 .
Therefore , more experiments are needed to verify the existence of the proposed negatively acting control site .
Because there is no direct evidence for positive or negative transcriptional control factors for the proU operon other than RNA polymerase , it is possible that the transcriptional regulation of the proU operon is effected only by the interaction of RNA polymerase with the promoter region .
Both the affinity and the rates of binding of proteins to their target sites on DNA are extremely sensitive in-vitro to the electrolyte concentration of the buffers used ( 39 , 43-45 ) .
Since the intracellular K + concentration can vary from 0.2 to 0.9 M depending on the external osmolarity ( 20 , 44 ) , one might expect that fluctuations in the extracellular osmolarity would have dramatic effects on the in-vivo transcription of nearly all genes .
Different DNA-protein interactions can have very different sensitivities to the electrolyte concentration , and therefore changes in the electrolyte concentration may have very complex effects on proteins that bind to th same target site .
Because there may be opposing promoters in the TaqI-TaqI fragment from positions 275 to 625 , RNA polymerase ( alone or together with transcriptional regulatory proteins ) might bind preferentially to the promoter read in the direction opposite to the transcription of the proU operon and thereby impair the binding of RNA polymerase to the proU promoter .
Osmotic control of transcription of the proU operon might be the result of the preferential weakening of the binding of RNA polymerase to the anti-proU promoter in comparison with the proU promoter and thereby result in increased transcription of the proU operon in media of high-osmolarity .
There is no concrete experimental proof that the proposed anti-proU promoter is recognized or is subject to osmotic control when present at its normal chromosomal location , and until such evidence is available , the above model should be regarded merely as a starting point for the next set of experiments .
Comparison of the nucleotide sequence of the promoter region of S. typhitni-iutin and E. c oli K-12 .
The sequences correspond to nucleotide positions 304 to 662 of the S. typhimuriuin sequence ( Fig. 3 ) .
The E. coli K-12 data are from reference 25 .
We thank J. Hamer for critical reading of the manuscript , R. Burgess and M. T. Record , Jr. , for helpful discussion , and J. Livingstone for carrying out the computer search for matches to the consensus promoter .
This work was funded by Public Health Service grant 1-ROI-GM 3194401 ( to L.N.C. ) .
ADDENDUM IN PROOF The nucleotide sequence data reported in this paper have been submitted to the EMBL , GenBank , and DDBJ nucle-otide sequence data bases and have been assigned the accession number M26063 .
Bacterial periplasmic transport systems : structure , mechanism , and evolution .
Ames , G. F.-L. , C. Prody , and S. Kustu .
Simple , rapid , and quantitative release of periplasmic proteins by chloroform .
Barron , A. , J. U. Jung , and M. Villarejo .
Purification and characterization of a glycine betaine binding protein from Esch-erichia coli .
Barron , A. , G. May , E. Bremer , and M. Villarejo .
Regulation of envelope protein composition during adaptation to osmotic-stress in Escherichia coli .
Barry , G. , C. Squires , and C. L. Squires .
Attenuation and processing of RNA from the rplJL-rpoBC transcription unit of Escherichia coli .
Beaucage , S. L. , and M. H. Caruthers .
Deoxynucleoside phosphoramidites-a new class of key intermediates for deoxypolynucleotide synthesis .
Bochner , B. R. , H.-C .
Huang , G. L. Schieven , and B. N. Ames .
Positive selection for loss of tetracycline resistance .
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 .
Cohen , S. N. , A. C. Y. Chang , H. W. Boyer , and R. B. Helling .
Construction of biologically functional bacterial plasmids in-vitro .
A third L-proline permease in Salmonella typhimurium which functions in media of elevated osmotic strength .
Physiological and genetic responses of bacteria to osmotic-stress .
Csonka , L. N. , M. M. Howe , J. L. Ingraham , L. S. Pierson III , and C. L. Turnbough , Jr. 1981 .
Infection of Salmonella typhi-murium with coliphage Mu dl ( Apr lac ) : construction of pyr : lac gene fusions .
Dattananda , C. S. , and J. Gowrishankar .
Osmoregulation in Escherichia coli : complementation analysis and gene-protein relationships of the proU locus .
Davis , R. W. , D. Botstein , and J. R. Roth .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 16 .
Druger-Liotta , J. , V. J. Prange , D. G. Overdier , and L. N. Csonka .
Selection of mutations that alter the osmotic control of transcription of the Salmonella typhimurium proU operon .
Dunlap , V. J. , and L. N. Csonka .
Osmotic regulation of L-proline transport in Salmonella typhimuriium .
Eckdahl , T. T. , and J. N. Anderson .
Computer modelling of DNA structures involved in chromosome maintenance .
Osmoregulation by potassium transport in Escherichia coli .
Epstein , W. , and S. G. Schultz .
Cation transport in Escherichia coli .
V. Regulation of cation content .
Erickson , B. D. , Z. F. Burton , K. W. Watanabe , and R. R. Burgess .
Nucleotide sequence of the rpsU-dnaG-rpoD operon from Salmonella typhimurium and a comparison of this sequence with the homologous operon of Escherichia coli .
Faatz , E. , A. Middendorf , and E. Bremer .
Cloned structural for the genes osmotically regulated binding-protein-de-pendent glycine betaine transport system ( ProU ) of Escherichia coli K-12 .
Flowers , T. J. , P. F. Troke , and A. R. Yeo .
The mechanism of salt tolerance in halophytes .
Identification of osmoresponsive genes in Escherichia coli : evidence for participation of potassium and proline transport systems in osmoregulation .
Nucleotide sequence of the osmoregulatory proU operon of Escherichia coli .
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 .
Higgins , C. F. , C. F. Domain , L. Sutherland , D. L. Stirling , I. R. Booth , G. May , and E. Bremer .
A physiological role for DNA supercoiling in the osmotic control of gene expression in S. typhimurium and E. coli .
Higgins , C. F. , L. Sutherland , J. Cairney , and I. R. Booth .
The osmotically regulated proU locus of Salmonella typhimu-rium encodes a periplasmic betaine-binding protein .
Hoops , B. C. , and W. R. McClure .
Strategies in regulation of initiation , p. 1231-1240 .
In F. C. Neidhardt , J. L. Ingraham , K. B. Low , B. Magasanik , M. Schaechter , and H. E. Umbarger ( ed .
) , Escherichia coli and Salmonella typhimurium : cellular and molecular biology .
American Society for Microbiology , Washington , D.C. 30 .
Hughes , K. T. , and J. R. Roth .
Conditionally transposi-tion-defective derivative of Mu dl ( Amp Lac ) .
Hughes , K. T. , and J. R. Roth .
Transitory cis complementation : a method for providing transposition functions to defective transposons .
Osmoregulation and compatible solutes in eubacteria .
Jovanovich , S. , M. T. Record , Jr. , and R. R. Burgess .
In an Escherichia coli coupled transcription-translation system elevated potassium concentrations and by an extract from cells grown at high osmolality .
Maniatis , T. , E. F. Fritsch , and J. Sambrook .
Molecular cloning : a laboratory manual .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 35 .
May , G. , E. Faatz , M. Villarejo , and E. Bremer .
Binding protein dependent transport of glycine betaine and its osmotic regulation in Escherichia coli K12 .
McMacken , R. , L. Silver , and C. Georgopoulos .
In F. C. Neidhardt , J. L. Ingraham , K. B. Low , B. Magasanik , M. Schaechter , and H. E. Umbarger ( ed .
) , Escherichia coli and Salmonella typhimurium : cellular and molecular biology .
American Society for Microbiology , Washington , D.C. 37 .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 38 .
Milner , J. L. , S. Grothe , and J. M. Wood .
Proline porter II is activated by a hyperosmotic shift in both whole-cells and membrane vesicles of Escherichia coli K12 .
Mossing , M. C. , and M. T. Record , Jr. 1985 .
Thermodynamic origins of specificity in the lac repressor-operator interaction : adaptability in the recognition of mutant operator sites .
Mulligan , M. E. , D. K. Hawley , R. Entriken , and W. R. McClure .
Escherichia coli promoter sequences predict in-vitro RNA polymerase selectivity .
Polites , H. G. , and K. R. Marotti .
A step-wise protocol for cDNA synthesis .
Ramirez , R. M. , W. S. Prince , E. Bremer , and M. Villarejo .
In vitro reconstitution of osmoregulated expression of proU Escherichia coli .
Record , M. T. , Jr. , C. F. Anderson , P. Mills , M. Mossing , and J.-H .
Ions as regulators of protein-nucleic acid interactions in-vitro and in vivlo .
Richey , B. , D. S. Cayley , M. C. Mossing , C. Kolka , C. F. Anderson , T. C. Farrar , and M. T. Record , Jr. 1987 .
Variability of the intracellular ionic environment of Escherichia coli : differences between in l'itro and in viso effect of ion concentrations on protein-DNA interactions and gene expression .
Roe , J.-H. , R. R. Burgess , and M. T. Record , Jr. 1984 .
Kinetics and mechanism of the interaction of Escherichia coli RNA polymerase with A PR promoter .
Salser , W. , R. F. Gesteland , and A. Bolle .
In vitro synthesis of bacteriophage lysozyme .
Sanger , F. , S. Nicklen , and A. R. Coulson .
DNA sequencing with chain-terminating inhibitors .
Is there an osmotic regulatory mechanism in algae and higher plants ' ?
Simons , R. W. , F. Houman , and N. Kleckner .
Improved single and multicopy lac-based cloning vectors for protein and operon fusions .
Somers , J. M. , G. D. Sweet , and W. W. Kay .
Fluorocitrate resistant tricarboxylate transport mutants of Salmonella typhi-murium .
Sutherland , L. , J. Cairney , M. J. Elmore , 1 .
R. Booth , and C. F. Higgins .
Osmotic regulation of transcription : induction of the proU betaine transport gene is dependent on accumulation of intracellular potassium .
Molecular cloning of Mu d ( bla lacZ ) transcriptional and translational fusions .
Wu , H.-Y. , S. Shyy , J. C. Wang , and L. F. Liu .
Transcription generates positively and negatively supercoiled domains in the template .
Zagursky , R. J. , K. Baumeister , N. Lomax , and M. L. Berman .
Rapid and easy sequencing of large linear double-stranded DNA and supercoiled plasmid DNA .