172 , No. 8 Molecular Cloning and Physical and Functional Characterization o the Salmonella typhimurium and Salmonella typhi Galactose Utilization Operons f HUO-SHU H. HOUNG , * DENNIS J. KOPECKO , AND LOUIS S. BARON Department of Bacterial Immunology , Walter Reed Army Institute of Research , Washington , D.C. 20307 Received 10 November 1989/Accepted 4 June 1990 The chromosomally encoded galactose utilization ( gal ) operons of Salmonella typhimuium and S. typhi were each cloned on similar 5.5-kflobase HindmI fragments into pBR322 and were identified by conpiementation of Gal-Escherichia coli strain .
Restriction endonuclease analyses indicated that these Salmonelae operons share considerable homology , but some heterogeneities in restriction sites were observed .
Subd ng d exonuclease mapping experiments showed that both operons have the same genetic organization as that establised for the E. coligal operon ( i.e. , 5 ' end , promoter , epimerase , transferase , kinase , and 3 ' end ) .
Two gal operator regions ( OE and ox ) of S. typkimurum , identified by repressor titration in an E. col superrepressor [ galR ( Sup ) ] mutant , were sequenced and found to flank the promoter region .
This promoter region is identical to the -10 and -35 regions of the E. coli gal operon .
Miniceil studies demonstrated that the three gal structural genes of S. typhimurium encode separate polypeptides of 39 kilodaltons ( kDa ) ( epimerase , 337 amino-acids [ aa 's ] ) , 41 kDa ( transferase , 348 aa 's ) , and 43 kDa ( kinase , 380 aa 's ) .
Despite functional and organizational similarites , DNA sequence analysis revealed that the S. typhimurium gal genes show less than 70 % homology to the E. colU gal operon .
Because of codon degeneracy , the deduced amino-acid sequences of these polypeptides are highly conserved ( > 90 % homology ) as compared with those of the E. coil gal enzymes .
These stdies have defined basic genetic parameters of the gal genes of two medally important Salmonela species , and our findings support the hypothesized divergent evolution of E. coil and Salmonella spp. .
from a common ancestral parent bacterium .
Salmonella typhimurium is a facultative intracellular pathogen of mice and the leading cause of gastroenteritis in humans in the United States ( 4 , 6 ) .
Parenteral and attenuated live vaccines against salmonellosis have been tested in experimental animals and human volunteers .
The low efficacy and unpleasant side effects obtained through parent-eral immunization with heavy suspensions of killed bacteria have prompted investigations into the use of live attenuated oral vaccines against salmonellosis ( 2 , 9 , 27 , 30 ) .
A galE mutant of Salmonella typhi , Ty2la , has been shown to be effective in protecting against typhoid fever without producing any side effects .
However , a protection efficacy of only 60 % was observed in S. typhi Ty2la in a field trial in Chile ( 19 ) .
In addition to the weak immunogenicity of galE mutants , the galactose sensitivity of galE mutants is lost in the presence of galactose .
This fact poses an even greater problem for the immunogenicity of galE mutants because galactose is required to synthesize the complete antigenic lipopolysaccharide , i.e. , smooth lipopolysaccharide .
Furthermore , all the galE mutants , including Ty2la , used as vaccine strains have been derived by nonspecific mutagenesis , i.e. , UV light or nitrosoguanidine mutagen treatment .
Thus , in addition to the specific galactose mutation , galE , these mutant strains also contain other undesirable characteristics , such as slow-growth and a poor survival rate because of the lyophilization process ( 10 ) .
The gal operon of Escherichia coli is known to consist of three structurally continuous genes which specify the enzymes required for the metabolism of galactose : galE ( UDPgalactose 4-epimerase ) , galT ( galactose-l-phosphate uridylyltransferase ) , and galK ( galactokinase ) .
These genes are expressed from a polycistronic mRNA in the order galE , galT , and galK ( 1 , 24 ) .
The genetic structure of the genes encoding these enzymes in S. typhimurium was partially characterized by Hone and colleagues ( 14 , 15 ) .
They utilized the available information to construct defined galE mutants of S. typhimurium and S. typhi .
These mutants were found to be avirulent and protective in mice ( 15 ) .
However , the galE mutant of S. typhi Ty2 derived by Hone et al. was found to be virulent in humans ( 14 ) .
To help in elucidating the complete genetic structure of the gal operons of S. typhimu-rium and S. typhi , we cloned the gal genes from these organisms .
Using endonuclease restriction mapping , subcloning techniques , and minicell labeling experiments , we found three clustered structural genes encoding the galactokinase , galactose-1-phosphate uridylyltransferase , and UDPgalactose 4-epimerase enzymes to be confined within a 5.50-kilobase ( kb ) HindIII fragment of S. typhimurium .
The information obtained from this study will provide the basis for constructing specific , stable galactose mutants of S. typhimurium for use as improved attenuated oral vaccines .
MATERIAL AND METHODS Bacterial strains and plasmids .
S. typhimurium LT2 and S. typhi Ty2 were the sources of the chromosomal DNA for the cloning of the galactose genes .
The E. coli galactose mutants HB101 ( galK ) , LE392 ( galT ) , CGSC4498 ( galE ) , CGSC5693 [ A ( gal-bio ) ] , and SA1796 [ galR ( Sup ) ; superrepressor ] were used as the recipients for complementation tests to identify the gal genes encoded by recombinant plasmids .
LE392 was described as a galTK double mutant ( 21 ) , but the results of this study showed that this strain actually is a-galT single mutant .
Plasmids pBR322 , pACYC184 , and pHC79 were used as cloning vectors ( 3 , 7 , 13 ) .
439 RESULTS dodecyl sulfate-polyacrylamide gel electrophoresis gel .
A `` 4C-methylated protein mixture of myosin , phosphorylase b , bovine serum albumin , ovalbumin , carbonic anhydrase , and lysozyme ( 5 , uCi/ml ; Amersham ) was used for protein mo-lecular weight marking .
Cesium chloride gradient-purified plasmid DNA was denatured with alkali as described by Hattori and Sakaki ( 11 ) and subjected to the dideoxynucleotide termination sequencing reaction ( 28 ) .
[ 35S ] dATP was used as the radioactive nucleotide .
DNA sequences obtained Were analyzed by using sequence analysis software from the University of Wisconsin , Madison , Genetics Computer Group and the DNA-protein sequence analysis system of International Biotechnologies .
Media , chemicals , and enzymes .
Bacterial stocks were maintained frozen at -70 °C in Luria broth with 20 % glycerol .
L broth was used as a complete medium for the routine growth of both Salmonella strains and E. coli strains .
MacConkey agar-galactose medium ( MacConkey base agar [ Difco Laboratories , Detroit , Mich. ] containing 0.5 % galactose ) was used to determine the galactose phenotypes of the strains .
Restriction nuclease enzymes , Bal31 exonuclease , and T4 DNA ligase were used as recommended by the suppliers ( New England BioLabs , Inc. , Beverly , Mass. ; Bethesda Research Laboratories , Inc. , Gaithersburg , Md. ; and International Biotechnologies , Inc , New Haven , Conn. ) .
All the chemicals used were reagent grade .
Construction of Hindml genomic libraries of Salmonella DNA .
Salmonella chromosomal DNA was prepared by the method of Marmur ( 22 ) .
HindIII-digested chromosomal DNA was ligated into the HindIII site of pBR322 at a ratio of 5:1 ( inserted DNA versus vector ) .
The ligated DNA was introduced into the E. coli HB101 ( galK ) strain by calcium chloride transformation ( 21 ) .
The transformation mixture was directly plated on MacConkey agar-galactose medium supplemented with 50 , g of ampicillin per ml .
Galactosepositive recombinants were identified after overnight incubation at 37 °C .
The recombinant plasmids obtained from galactose-positive transformants contained the Salmonella galK locus .
To determine whether the recombinant plasmids code for a particular gal locus , we used different gal mutants of E. coli as recipients for complementation tests .
Plasmids were transformed into competent E. coli mutants , and the galactose phenotypes of the transformants were identified after overnight incubation with MacConkey agar-galactose medium .
Exonuclease IH mapping of the gal genes of S. typhimurium .
The 5.5-kb HindIII insert of pHH2 coding for the entire gal operon of S. typhimurium was blunt ended by Klenow enzyme repairing and subcloned into the SmaI site of pUC19 ( 17 , 23 ) .
Both orientations of the inserted DNA were subcloned into pUC19 .
One of the resultant recombinant plasmids , pHH21 , has two unique restriction sites , SphI and BamHI , which lie between the sequencing primer-binding site and the end of the insert .
The galE end of pHH21 is located adjacent to the BamHI site of the cloning vector , pUC19 .
Plasmid pHH21 ( 5 , ug ) was subjected to BamHI and SphI digestions to leave a 3 ' protrusion ( SphI cut ) adjacent to the sequencing primer-binding site and a 5 ' protrusion ( BamHI cut ) adjacent to the gal sequences .
The digested DNA was subjected to exonuclease III digestion as described in the Erase-base kit manual ( Promega Biotech , Madison , Wis. ) ( 12 ) .
After exonuclease III treatment , DNA samples taken at different times were subjected to Klenow enzyme repairing followed by T4 DNA ligase incubation .
The ligated DNA mixtures were transformed into E. coli HB101 to select for ampicillin-resistant transformants .
Recombinant plasmids were isolated and screened for gal genotypes by complementation tests .
[ 35SJmethionine labeling of plasmid-encoded proteins in a miniceli system .
Proteins encoded by recombinant plasmids were detected in a minicell strain of E. coli , DS410 ( 26 ) .
In a typical experiment , minicells were isolated from 200 ml of an overnight Penassay broth culture of E. coli DS410 harboring the desired plasmids .
These were labeled with [ 35S ] methionine ( 5 , jCi ; specific activity , 600 Ci/mmol ; Amersham Corp. , Arlington Heights , Ill. ) essentially as described by Newland et al. ( 26 ) .
Each of the radiolabeled lysates ( 106 cpm ) was loaded onto a lane and analyzed on a 12 % sodium Cloning of Salmonella gal genes .
HindIlI libraries of S. typhimurium and S. typhi DNA in pBR322 were initially screened for galactose-positive phenotypes .
The recombinant plasmids that code for the functional galK gene in these strains were analyzed for their content .
Plasmids pHH2 and pHH6 , derived from S. typhimurium and S. typhi , respectively , contained 5.5-kb inserts .
Restriction analyses showed that the 5.5-kb inserts in these two recombinant plasmids were similar but not identical ( Fig. 1 and 2 ) .
Complementation tests showed that both pHH2 and pHH6 were capable of complementing various gal mutations in E. coli strains , including the one that has its entire gal operon sequences deleted , i.e. , E. coli CGSC5693 [ A ( gal-bio ) ] .
None of the subcloned fragments of parental pHH2 and pHH6 were capable of complementing CGSC5693 to Gal ' .
This result suggests that pHH2 and pHH6 code for the entire gal sequence of S. typhimurium and S. typhi , respectively .
Subcloning and mapping of the gal operons of S. typhimu-rium and S. typhi .
Various recombinant plasmids were constructed by subcloning the restriction fragments of pHH2 and pHH6 as shown in Fig. 1 and 2 .
The gal genotypes of these plasmids were determined through complementation tests with E. coli gal mutants as recipients .
The results showed that the gal gene orders for S. typhimurium and S. typhi were the same as in the E. coli gal operon ( galE , galT , and galK ) .
Plasmids pHH11 to pHH14 were capable of complementing LE392 to Gal ' .
Plasmid pHH11 could not complement either the galE ( strain CGSC4498 ) or the galK ( strain HB101 ) single mutation .
It is unlikely that all the Gal ' transformants of LE392 were the result of recombinations between pHH11 and the chromosomal gal sequences .
Thus , LE392 should be a galT single mutant , even though it was previously described as a galTK double mutant ( 21 ) .
Exo-nuclease III mapping also confirmed that the S. typhimurium gal genes were progressively deleted in the order galE , galT , and galK .
Once deletions go beyond the galE locus , the remainder of the galT and galK genes can still be expressed in the absence of the gal promoter by utilization of the lac promoter of pUC19 .
The gal repressor of E. coli binds to the gal operator region ( 16 , 20 ) .
The multiple-copy recombinant plasmids carrying the gal operator sequences are capable of titrating out the limited number of gal repressor molecules produced by the E. coli galR ( Sup ) mutant SA1796 .
Thus , the chromosomal gal activities of the galR - ( Sup ) mutant SA1796 can be derepressed by multiple-copy plasmids encoding the gal operator sequences .
Titration of the gal repressor by plasmids was performed by transforming the E. coli galR ( Sup ) mutant SA1796 ( provided by S pHHl 4 GalE BglI Cial pHH28 GalE GalT I BglI PVUI 7I M1UI GalK IF  BglII PVUI M1UI PVUI AccI PVUI LiCaI .
I PvuI PVUI HindIII PstI M1UI I I IIZCR HindIIIDraI CRI EcoRV Cial DraI BstEII 1.0 kb FIG. 1 .
Mapping of galE , galT , and galK loci of S. typhimurium by subcloning of the restriction fragments of pHH2 , which codes for the entire gal oppron of S. typhimurium .
Only the inserted portions of these recombinant plasmids are shown .
The gal genotypes of each recombinant plasmid were determined by complementation tests with E. coli gal mutants as recipients .
Plasmid pHH28 was generated through Bal3l exonuclease digestion to delete the nucleotides ( shown as a broken line ) coding for the galT and galK loci of S. typhimurium .
Adhya ) with plasmids and plating the transformation mixture onto MacConkey agar-galactose plates supplemented with appropriate antibiotics .
After overnight incubation at 37 °C , transformants bearing plasmids unable to bind the gal repressor formed white colonies .
Those bearing plasmids coding for functional gal operator sequences formed bright red colonies .
Only the nucleotide sequences associated with galE activity , i.e. , plasmids pHH2 , pHH12 , pHH13 , pHH14 , pHH28 , pHI6 , pHH62 , and pHH64 , were capable of derepressing the ` galR ( Sup ) , mutation .
This result demonstrated that the gal sequences of S. typhimurium have a binding capacity for the gal repressor similar to that of E. coli .
Identficato of polypeptides encoded by the gal genes of S. typhimuriwm .
Minicell labeling experiments showed that plasmid pHH11 , coding for the single galT gene of S. typhimurium , produced a 41-kilodalton ( kDa ) polypeptide ( Fig. 3 ) .
Plasmid pHH12 produced an additional 39-kDa polypeptide .
This 39-kDa polypeptide should be the product of the galE locus .
Parental plasmid pHH2 coded for a third polypeptide of 43 kDa .
Even though pHH13 did not code for the functional galK activity , it carried 1,000 base pairs ( bp ) of the galK locus .
It is possible that the 43-kDa peptide of pHH13 is a hybrid product between the truncated galK gene and other sequences derived from the cloning vector , pHC79 .
Usually hybrid proteins are less stable than regular peptides ; this may explain why the intensity of the 43-kDa band was lower than that of the parental plasmid , pHH2 .
All the recombinant plasmids , except for pHH13 , used in the minicell experiments were directly derived from the pBR322 vector .
Plasmid pHH13 was constructed by subcloning the EcoRl-BglII frnment of pHH2 into the EcoRl-BamHI-digested pHC79 vector .
The extra peptide band below the 29-kDa marker probably was the product of the 2.2 kb of pHC79 which was missing from the pBR322 vector , which was used to construct the other recombinant plasmids used in the minicell experiments .
By deducing the polypeptides encoded by the galE and galT loci from their parental plasmids , we concluded that the third polypeptide , of 43 kDa , is the galK product .
Nucleotide sequencing of the promoter region of the S. typhimurinum gal operon .
The 2.3-kb ClaI insert of pHH14 codes for galE and galT functions as well as for gal repres-sor-binding activity .
In minicell labeling experiments it was found that two separate polypeptides , of 39 and 41 kDa , were the products of galE and galT .
At least 1.1 and 1.2 kb are required to code for the galE and galTproducts of S. typhimurium .
It is likely that the 2.3-kb ClaI insert of pHH14 contains just enough nucleotides to code for these two gal genes and their promoter region and for gal repressor-binding sequences .
gal repressor-binding activity is known to be associated with the galE locus of S. typhimurium .
Thus , the gal regulatory region should be located adjacent to one of the ClaI boundaries , at the galE end of pHH14 .
The ClaI sites of pHH14 were flanked by the EcoRI and HindIII sites of the pBR322 vector .
Two commercially available primers , EcoRI and HindIII primers ( Pharmacia ) , were directly used to determine the nucleotide sequences adjacent to the ClaI sites of pHH14 ( 31 ) .
The nucleotide sequences obtained were compared with the E. coli gal regulatory sequences ( 5 , 25 ) by using the sequence analysis software package of the Genetics Computer Group ( University of Wisconsin ) .
A stretch of DNA derived from the galE end of pHH14 was found to be highly homologous to the E. coli gal promoter .
The results indicated perfectly matched se quences between the -10 and -35 regions for these organisms ( Fig. 4 ) .
Sequencing the entire gal operon of S. typhimurium .
Exo-nuclease III was used to create nested deletions of the 5.5-kb HindIII insert of pHH2 , which codes for the entire gal operon of S. typhimurium .
Twenty-six independent deletion plasmids were isolated and identified among the pool of recombinant plasmids of the pHH2 derivative previously treated with exonuclease III ( see Materials and Methods for details ) .
These plasmids were found to have deletions ranging from 100 to 300 bp and spanning the entire gal operon of S. typhimurium .
Thus , the entire nucleotide sequence was determined with these plasmids as DNA templates by means of the dideoxynucleotide termination reactions of Sanger et al. ( 28 ) .
A Biosearch DNA synthesizer was used to synthesize oligonucleotide primers complementing the available nucleotide sequences .
These synthetic primers were used to sequence the opposite DNA strand .
A total of 5,490 bp of nucleotide sequence was determined for the 5.5-kb HindIII insert of pHH2 .
The first nucleotide of the start codon ( ATG ) of the galE gene was assigned as the beginning of the gal sequence for S. typhimurium ( Fig. 4 ) .
A single open reading frame which codes for three separate polypeptides of 337 amino-acids ( bp 1 to 1011 ) , 348 amino-acids ( bp 1027 to 2070 ) , and 380 amino-acids ( bp 2077 to 3216 ) was found .
There was less than 70 % homology between the nucleotide sequences coding for the gal operons of S. typhimurium and E. coli ( 8 , 18 ) .
The deduced amino-acid sequences for the gal open reading frames of these two organisms appeared to be highly conservative , with more than 90 % homology .
The mismatched amino-acids between the S. typhimurium and E. coli gal enzymes were usually found to have similar hydr-opathy characteristics , e.g. , valine versus isoleucine , glycine versus glutamic-acid , and serine versus proline .
Thus , th PVUI M1UI I ilI IFEol H4ndiIIDraI EC laI EcoRV DraI BstEII pHHl1 GalT I I M1UI PVUI I I I M1UI PVUI ClaI pHH12 EcoRI AglI ClaI P Miul PvuI P PatI Pvul VUI UI V pHH1 3 GalE GalT I I I EcoRIClI MIUI BglI PVuI I ci'a 4 Miul ' Bg M1UIPVfT g1III Acc PVUI PVUI + PstI GalT l M1UI PvuI I M1UI M1UI ClaI pHHl 4 GalE BglI Cial pHH28 GalE GalT I BglI PVUI 7I M1UI GalK IF  BglII PVUI M1UI PVUI AccI PVUI LiCaI .
I PvuI PVUI HindIII PstI M1UI I I IIZCR HindIIIDraI CRI EcoRV Cial DraI BstEII 1.0 kb FIG. 1 .
Mapping of galE , galT , and galK loci of S. typhimurium by subcloning of the restriction fragments of pHH2 , which codes for the entire gal oppron of S. typhimurium .
Only the inserted portions of these recombinant plasmids are shown .
The gal genotypes of each recombinant plasmid were determined by complementation tests with E. coli gal mutants as recipients .
Plasmid pHH28 was generated through Bal3l exonuclease digestion to delete the nucleotides ( shown as a broken line ) coding for the galT and galK loci of S. typhimurium .
pHH6 GalE BglI GalK GalT Mu 1  0 P MII M ; JI ssIIi 1vuI ClaI PvuI AccI PvuI HminIII PIuI MluI HiH D Eco ndiIIiIi1lraIDraIRI BstEII EcoRV pHH62 GalT + GalE l I ¬ PvuI Milu HindI , LDraI DraI I EcoRI BstEII EcoRV KluI ClaI BglI pHH63 GalT + ClaI PvuI PvuI MluI pHH64 GalT GalK -- I-i-Bg II MluI PvuI tl , lI ClaI MluI AccI PvuI GalE I I ` III ' I HindIII Drai EcoRV BglI DraI EcoRI BstEII I IHindIII PvuI Mll PvuI pHH65 GalK PvuI PvI MluI -RCB- BglI I ClaI MluI AccI PvuI HindI I I 1.0 kb FIG. 2 .
Mapping of galE , galT , and galK loci of S. typhi by subcloning of the restriction fragments of pHH6 , which codes for the entire gal operon of S. typhi .
Only the inserted portions of these recombinant plasmids are shown .
The gal genotypes of each plasmid were identified and confirmed by complementation tests with E. coli gal mutants as recipients .
The broken line in the pHH64 map shows the MluI deletion region generated through MluI endonuclease digestion and T4 DNA self-ligation .
Plasmid-encoded polypeptides in E. coli minicell strain DS410 .
Lanes : A , pBR322 ; B , pHH2 ( galETK ) ; C , pHH13 ( galET ) ; D , pHH14 ( galET ) ; E , pHH11 ( galT ) ; F , protein markers ( myosin , 200 kDa ; phosphorylase b , 92.5 kDa ; bovine serum albumin , 69 kDa ; ovalbumin , 46 kDa ; carbonic anhydrase , 30 kDa ; and lysozyme , 14.3 kDa .
The 39 - , 41 - , and 43-kDa polypeptides were the products of the galE , galT , and galK of the S. typhimurium gal genes operon , respectively .
TTT TCT CCC CAC AGT AAA ACG CTG CCG GAG TTG AGC CTG CCC GCG CTG ACG GAA Sir Lou Ph. Ser Pro Asp His Ser Lys Thr Lou Pro Glu Lou Pro Ala Lou Thr Glu Ala Val ATC GTC AGA ACC TGG CAG ACG CAG ACC GCC GAG TTA GGC AAA ACC TAT CCG TGG GTG I1-Val Arg Thr Trp Gln Thr Gln Thr Ala Glu Leu Gly Lys Thr Tyr Pro Trp Val Lys Slu 1,500 CCG CAT GGA CAG GTC TTT GAA AAT AAA GGC GCG GCG ATG GGC TGT TCG AAC CCG CAT Gln Vai Ph. Glu Asn Lys Gly Ala Ala Met Gly Cys Sir Asn Pro His Pro His Gly CAG GTC TGG GCC AAC AGC TTC CTG CCG AAT GAG GCG GCC GAA CGC GAA GAT CGT TTA Gln Vii Trp Ala Asn Ser Phe Lou Pro Asn Glu Ala Ala Glu Arg Glu Asp Arg Leu rie * CAA AAA GCC TAC TTC GCC GAG CAG CGC TCG CCA ATG CTG GTG GAT TAT GTC CAG CGA Gln Lys Ala Tyr Phe Ala Glu Gln Arg Ser Pro Met Leu Val Asp Tyr Vai Gln Arg Slu Lys FIG. 4 .
Nucleotide and deduced amino-acid sequences of the gal coding region of S. typhimurium .
The first nucleotide of the galE start codon ( ATG ) is the beginning of the 3,219-bp gal coding sequence ( GenBank accession number M33681 ) .
The upstream sequences of the S. typhimurium galE locus show the -10 and -35 regions perfectly matched to those of the E. coli gal operon .
The differences between the S. typhimurium and the E. coli gal polypeptide sequences are shown as italicized amino-acids of E. coli ; asterisks indicate the missing residues of E. coli It z. DISCUSSION In this study , the gal operons of S. typhimurium LT2 and S. typhi Ty2 were isolated from their Hindlll chromosomal libraries .
Two similar but not identical 5.5-kb HindIII inserts in pHH2 and pHH6 , derived from these two organisms , were found to code for the entire gal operon sequence , including the galE , galT , and galK structural genes and the gal promoter region .
Most of the restriction sites found on these two HindIII fragments matched each other , with a few exceptions .
The ClaI , PvuI , EcoRI , EcoRV , BglII , and MluIl sites were almost superimposable between these two fragments , whereas the PstI , Sacl , and DraI sites were different ( Fig. 1 and 2 ) .
Subcloning of restriction fragments of pHH6 and pHH2 was performed in a similar fashion , and similar recombinant plasmids coding for identical gal activities were derived from these two parental plasmids .
The galT genes of S. typhi and S. typhimurium were encoded by similar 1.3-kb PvuI fragments of pHH6 and pHH2 .
The gal regulatory region and galE locus were both located upstream from galT in both pHH2 and pHH6 .
The third gal gene , galK , was located downstream from galT in both pHH2 and pHH6 .
Too many endonuclease restriction sites were located within the galK structural genes of S. typhimurium and S. typhi ( Fig. 1 and 2 ) to enable construction of a recombinant plasmid coding for the single galK gene by subcloning of available restriction fragments of these two organisms .
On the basis of these results , it is likely that S. typhimurium and S. typhi contain very similar gal coding sequences .
Perhaps only a few nucleotide differences exist between the gal coding sequences of these two closely related organisms .
Exonuclease III progressive deletion of the pHH21 plasmid further confirmed the physical layout of the S. typhimurium gal operon as being in the order galE , galT , and galK .
DNA sequences derived from the 2.3-kb ClaI insert of pHH14 revealed that the promoter sequence from -10 to -35 was identical to the promoter sequence of the E. coli gal operon .
The existence of homologous promoter sequences between E. coli and S. typhimurium has also been reported for the bio operons of these two organisms ( 29 ) .
The striking aspect of the organization of the regulatory loci of the gal operon in E. coli is the existence of two essential operators which are separated by more than 90 bp and which flank , but do not overlap , the promoter region ( 16 , 20 ) .
Dimeric gal repressors bound to external and internal operators ( OE and o , ) generate a DNA loop structure and prevent RNA polymerase binding to the promoter region .
The binding capacity of the gal sequences of S. typhimurium suggests that similar or identical gal operator sequences ( both OE and ox ) should be found in close proximity to the promoter sequences .
Nucleotide sequence analysis of the 5,490-bp HindIII insert of pHH2 showed that the -10 and -35 consensus sequences are located 32 bp upstream from the galE transcription start site ( Fig. 4 ) .
Nucleotide sequence analysis also showed that the gal operon of S. typhimurium contains two operator loci , OE ( 5 ' - GTGGAATCGTlTTACAC-3 ' ) and o0 ( 5 ' - GTGGTA GCGGTTACAT-3 ' ) , having sequences perfectly matched to those of the E. coli gal operon ( 16 , 20 ) .
The former is located around -79 , i.e. , upstream of the promoter sequences , and the latter is located in the +20 region , i.e. , downstream of the promoter sequences and inside the galE structural gene .
The precise locations of S. typhimurium gal operators are slightly different from those of E. coli gal operators , located in the -52 and +45 regions ( 16 ) .
However , the distances between two gal operators for these two organisms are approximately the same , 99 versus 97 bp .
The homologous gal operator sequences explain the binding capacity of the S. typhimurium gal operators encoded by the multicopy plasmid that can titrate out the gal repressor activity of the galR ( Sup ) mutation in the E. coli background .
This explanation is consistent with the results of the derepression experiments , indicating that the repressor-binding activities are always associated with the nucleotides coding for the galE gene but not the other two distal genes , galT and galK .
The nucleotide sequence data and gal repressor-binding experiments of this study further confirm the unique regulatory loci of the E. coli gal operon that had been previously established through extensive genetic and biochemical studies .
A Southern blot study under highly stringent conditions ( 50 % formamide and 65 °C wash temperature ) with the S. typhimurium galT gene on the 1.3-kb PvuI fragment as a radioactive probe revealed that homologous nucleotide sequences exist among E. coli , S. typhimurium , S. typhi , and Shigella flexneri .
However , the restriction maps for the gal coding sequences of E. coli and S. typhimurium are totally different .
It is surprising that the nucleotide sequences of the S. typhimurium gal structural genes ( galE , galT , and galK ) have less than 70 % homology with the corresponding genes of E. coli , whereas the promoter and operator sequences for these two operons are perfectly matched .
Because of codon degeneracy , the deduced polypeptide sequences of these two operons have more than 90 % homology .
The number of amino-acid-residues for the galactose enzymes of S. typhi-murium and E. coli are slightly different : 337 versus 338 for galE , 348 versus 347 for galT , and 380 versus 382 for galK .
The longest mismatched amino-acid sequences are found near the carboxy terminus of UDPgalactose 4-epimerase .
It is therefore likely that the carboxy end of the epimerase does not play an essential role in enzymatic functioning .
Adhya , S. , and J. A. Shapiro .
The galactose operon of E. coli K-12 .
I. Structural and pleiotropic mutations of the operon .
Ashcroft , M. T. , J. Morrison-Ritchie , and C. C. Nicholson .
Controlled field trial in British Guiana school children of heat-killed-phenolized and acetone-killed lyophilized typhoid vaccines .
Balbas , P. , X. Soberon , E. Merino , M. Zurita , H. Lomeli , F. Valie , N. Flores , and F. Bolivar .
Plasmid vector pBR322 and its special-purpose derivatives-a review .
Blanden , R. V. , G. B. Mackaness , and F. M. Collins .
Mechanisms of acquired resistance in mouse typhoid .
Mutations in Escherichia coli operon that define two promoters and the binding site of the cyclic-AMP-receptor-protein .
Centers for Disease Control .
Salmonella surveillance summary for 1980 .
Centers for Disease Control , Atlanta .
Chang , A. C. Y. , and S. N. Cohen .
Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the plSA cryptic miniplasmid .
DeBouck , C. , A. Ricco , D. Schumperli , K. McKenny , J. Jeffers , C. Hughes , and M. Rosenberg .
Structure of the galactokinase gene of Escherichia coli , the last ( ?
) gene of the gal operon .
DuPont , H. L. , R. B. Hornick , M. J. Snyder , J. P. Libonati , and T. E. Woodward .
Immunity in typhoid fever : evaluation of live streptomycin-dependent vaccine , p. 236-239 .
Germanier , R. , and E. Furer .
Immunity in experimental salmonellosis .
Basis for the avirulence and protective capacity of gal E mutants of Salmonella typhimurium .
Hattori , M. , and Y. Sakaki .
Dideoxy sequencing method using denatured plasmid templates .
Unidirectional digestion with exonuclease III creates targeted breakpoints for DNA sequencing .
Hohn , B. , and J. Collins .
A small cosmid for efficient cloning of large DNA fragments .
Hone , D. M. , S. R. Attridge , B. Forrest , R. Morona , D. Daniels , J. T. LaBrooy , R. C. A. Bartholomeusz , D. J. C. Shearman , and J. Hackett .
A galE via ( Vi antigen-negative ) mutant of Salmonella typhi Ty2 retains virulence in humans .
Hone , D. , R. Morona , S. Attridge , and J. Hackett .
Construction of defined galE mutants of Salmonella for use as vaccines .
Irani , M. , L. Orosz , and S. Adhya .
A control element within a structural gene : the gal operon of Escherichia coli .
Cleavage of structural proteins during the assembly of the head of bacteriophage T4 .
Lemaire , H.-G. , and M.-H .
Nucleotide sequences of the galE gene and the galTgene of E. coli .
Levine , M. M. , R. E. Black , D. Ferreccio , M. L. Clements , C. Lanata , J. Rooney , R. Germanier , and Chilean Typhoid Committee .
Field trial efficacy of attenuated Salmonella typhi oral vaccine strain Ty2la , p. 79-89 .
In J. Robbins ( ed .
) , Proceedings of the International Symposium on Bacterial Vaccines .
Praeger Publishers , New York .
Majumdar , A. , and S. Adhya .
Demonstration of two operator elements in gal : in-vitro repressor binding studies .
Maniatis , T. , E. F. Fritsch , and J. Sambrook .
Molecular cloning : a laboratory manual .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 22 .
A procedure for the isolation of deoxyribo-nucleic acid from microorganisms .
Marsh , J. L. , M. Erfle , and E. J. Wykes .
The pIC plasmid and phage vectors with versatile cloning sites for recombinant selection by insertional inactivation .
Michaelis , G. , and P. Starlinger .
Sequential appearance of the galactose enzymes in E. coli .
Musso , R. , R. Di Lauro , M. Rosenberg , and B. De Crombrugghe .
Nucleotide sequence of the operator-promoter region of the galactose operon of Escherichia coli .
Newland , J. W. , B. A. Green , J. Foulds , and R. K. Holmes .
Cloning of extracellular DNase and construction of a DNase-negative strain of Vibrio cholerae .
Evaluation of typhoid vaccines in the laboratory and in a controlled field trial in Poland .
Sanger , F. , S. Nicklen , and A. R. Coulson .
DNA sequencing with chain-terminating inhibitors .
Shiuan , D. , and A. Campbell .
Transcriptional regulation and gene arrangement of Escherichia coli , Citrobacterfreundii , and Salmonella typhimurium biotin operons .
Wahdam , M. H. , C. Serie , Y. Cerisier , S. Sallam , and R. Germanier .
A controlled field trial of live Salmonella typhi strain Ty2la oral vaccine against typhoid : three-year results .
Wallace , R. B. , M. J. Johnson , S. V. Suggs , K.-I .
Miyoshi , R. Bhatt , and K. Itakura .
A set of synthetic oligodeoxyribo-nucleotide primers for DNA sequencing in the plasmid vector pBR322 .