173 , No. 19 Locus Transcriptional Organization of the rfaGBIJ of Salmonella typhimurium R. BRAZAS , t E. DAVIE , A. FAREWELL , t AND L. I. ROTHFIELD * Department of Microbiology , University of Connecticut Health Center , Farmington , Connecticut 06032 Received 22 April 1991/Accepted 19 July 1991 The transcriptional organization of the rfaGBIJ gene cluster of Salmonella typhimurium was studied by using lacZ and cat transcriptional probes .
The results indicated that the leftward end of the gene cluster ( rfaG-zfaB-ifal ) is an operon that is transcribed from one or more promoters that lie upstream of rfaG .
The results further indicated that the product of the rfafH ( sfrB ) gene acts as a positive regulator of transcription of the entire rfaGBIJ cluster .
At least one site required for the RfaH-mediated transcriptional regulation lies within or very close to the upstream promoter .
The lipopolysaccharide of gram-negative bacteria contains a core region that is similar in all gram-negative bacteria ( 16 ) .
The lipopolysaccharide core provides the attachment site for the 0-antigenic polysaccharide that is a major determinant of virulence in gram-negative organisms .
Loss of the ability to synthesize the complete core lipopolysaccharide is associated with a significant decrease or loss of virulence .
is synthesized by series of glycosyltransferases The core a that catalyze the stepwise transfer of sugars from their nucleotide-linked precursors ( 16 ) .
The genes coding for the glycosyltransferases are located in the rfaGBIJ gene cluster at approximately 79.5 units on the Salmonella typhimurium and Escherichia coli genetic maps ( 2 , 19 ) .
The relative positions of several of the genes within the cluster ( Fig. 1 ) for typhimurium by using genetic have been established S. of cloned methods ( 4 , 19 ) and by studying the ability known rfaG , rfaB , chromosomal fragments to complement rfaI , and rfaJ mutations ( 12 ) .
cluster is thought Expression of the genes of the rfaGBIJ the of the rfaH to be positively regulated by product gene is 84 ( formerly the sfrB gene in E. coli ) , which located at units on the genetic map ( 19 ) .
The evidence for the positive regulation of the rfa genes by RfaH is indirect .
It is based on the loss of specific glycosyltransferase activities and a marked decrease in the amount of one of the transferase proteins in rfaH amber mutant ( 5 ) .
The existence of the an rfaH amber that rfaH acts via its protein mutant suggests product .
RfaH is also a positive regulator of expression of the tra Y-Z operon of the F factor .
Beutin et al. ( 3 ) showed that truncated transcripts accumulate in rfaH mutant tra Y-Z cells ; this led them to suggest that the RfaH protein acts to prevent premature termination of transcription within the tra Y-Z transcriptional unit .
Support for this view has also from studies involving lacZ transcriptional probes ( 9 ) .
come It is not known whether rfaGBIJ gene expression is transcriptionally regulated by a similar antitermination mechanism .
In this paper we describe studies of the transcriptional t Present address : Department of Cellular , Viral and Molecular Biology. , University of Utah , Salt Lake City , UT 84132 .
i Present address : Department of Microbiology , University of Michigan Medical School , Ann Arbor , MI 48104 .
organization of the rfaGBIJ gene cluster and the effects of RfaH on the rfaGBIJ transcriptional-pattern , by using lacZ and cat probes as transcriptional monitors .
MATERIALS AND METHODS Bacteriological procedures .
Culture conditions and micro-were as described by Miller ( 15 ) .
Cells biological procedures and harvested in the were grown at 37 °C in L-broth ( LB ) Resistance to chlorampheni-midexponential-growth-phase .
form colonies after col was defined as the ability to overnight LB 100 of chloramgrowth at 37 °C on plates containing , ug phenicol per ml .
When isopropyl-p3-D-thiogalactoside ( IPTG ) was used , it was present at a concentration of 1 mM for three generations before the cells were harvested .
E. coli M1170 ( rfaH Alac ) and M1174 ( rfaH + Alac ) are isogenic E. coli K-12 strains provided by M. Achtman .
S. typhimurium AF21 ( rfaI : : Mudlac rfaH + ) was constructed as described below .
Strain AF21-106 ( rfaH487 rfaI : : Mudlac metE : : TnJO ) was constructed ES18-medi-by from ated cotransduction of rfaH487 and metE : : TnJO AF106 ( rfaH487 metE : : TnJO ) into AF21 , making use of the close metE transductional linkage of the rfaH and genes ( 19 ) .
Recombinants that had received metE : : TnJO were selected on the basis of their tetracycline resistance .
Within this group , transductants that had acquired the rfaH487 allele were identified on the basis of their resistance to bacteriophage P1 ( 22 ) .
Strain AF21-107 ( rfaH + rfaI : : Mudlac metE : : TnJO ) was obtained from the same transduction and was identified as a tetracycline-resistant recombinant that the P1-sensitive of AF21 .
Strain had retained phenotype LE392 ( hsdR supE44 supF58 lac YJ galK2 gal722 metBI trpR55 ) was obtained from W. Nunn .
pKZ26 ( provided by K. Sanderson ) is a pBR322 derivative that contains 7.5 kb of chromosomal DNA from S. typhimurium ( 12 ) .
The genes within the insert ( Fig. 1 ) were identified on the basis of genetic complementation of known rfa mutants .
The rfaGBIJ gene order in pKZ26 was deduced from the gene order of the chromosomal genes as obtained by standard genetic transduction and conjugation analysis ( 19 ) .
The gene order was partially confirmed by complementation studies with plasmids containing fragments of the chromosomal insert of pKZ26 ( 12 ) and by sequencing of the rfaI and rfaJ genes ( 4 ) .
( ii ) pRB18 to pRB85 .
The lacZ minitransposon from X1048 616 L. Lo .
Locations and orientations of lacZ insertions .
The locations and orientations of 15 rfa : : lacZ insertions are shown under the physical map of the chromosomal insert in pKZ26 .
Each number corresponds to the plasmid that carries the indicated insertion ( e.g. , 85 represents the lacZ insertion in pRB85 ) .
The approximate locations of the rfa genes is taken from Kadam et al. ( 12 ) .
In the case of rfaG , rfaB , and rfal , the positions also correspond to the genes inactivated by specific lacZ insertions in the present study ( Table 1 ) .
Coordinates are given in kilobases from the Hindill site in the rfaG end of the insert .
Abbreviations : C , ClaI ; H , HindIlI ; E , EcoRI ; B , BamHI ; Hp , HpaI ; P , PstI ; Pv , PvuI ; N , NdeI .
was inserted at random locations within pKZ26 by IPTG induction of LE392 ( X1048 ) / pKZ26 , essentially as described by Way et al. ( 20 ) .
The location and orientation of the lacZ sequences in each of the derivative plasmids are shown in Fig. 1 .
The 5.4-kb lacZ insert in each case contains a BamHI site at either end of the inserted sequences .
The numbers assigned to the inserts ( Fig. 1 ) are also used to identify the plasmids .
Thus , pRB18 contains the 7.5-kb chromosomal rfa segment containing insert 18 ( rfa : : lacZ18 ) .
In the pSL plasmids ( provided by C. Squires ) , the cat gene serves as a transcriptional reporter for transcripts exiting from a polycloning site ( 13 ) .
Transcriptional terminators prevent transcription from the plasmid into the reporter gene , and translational terminators prevent formation of a fusion protein between the cat gene product and the protein product of genes cloned into the polycloning site .
Transcription into cat was monitored either by measurement of chloramphenicol acetyltransferase ( CAT ) enzyme activity or by the appearance of resistance to chlor-amphenicol .
In pSL100 there is no promoter between the transcriptional terminators and the polycloning site ( Fig. 2 ) .
Therefore , cat expression requires that the inserted fragment contain a promoter oriented toward the cat gene .
In pSL102 , a ribosomal P2 promoter is present that provides an exogenous promoter for inserted into the sequences polycloning site ( Fig. 3 ) .
pSL140 is similar to pSL102 , except that the UV5 Plac promoter is present instead of the P2 promoter of pSL102 ( Fig. 3 ) .
( iv ) pRB101 to pRB188 .
The rfaGBIJ fragments shown in Fig. 2 and 3 were cloned into the polycloning sites of pSL100 , pSL102 , or pSL140 by standard methods .
The donor fragments were obtained from pRB56 , pRB66 , or pRB78 by using the BamHI site at the left end of the lacZ insert ( Fig. 1 ; Table 1 ) as the downstream end of the subcloned fragment , except for pRB179 , for which the donor fragment was obtained from pRB178 ( Fig. 3 ) .
Other restriction sites used for the cloning are indicated in Fig. 2 and 3 .
( v ) pRBD14 and pRBD19 .
Plasmids pRBD14 and pRBD19 contain partial deletions of the left end of the chromosomal insert of pRB161 ( Fig. 2 ) and were prepared as described below .
Construction of partial deletions of rfaGBIJ sequences .
pRB161 ( Fig. 2 ) was linearized by treatment with BglII and then treated with Bal 31 exonuclease for approximately 15 min at 30 °C .
BamHI linkers were ligated to the ends of the partially digested linear fragments .
The preparation was then treated with BamHI to generate BamHI-BamHI fragments from which a variable amount of DNA had been removed from the upstream end of the pRB161 chromosomal insert .
Fragments of approximately 3 to 3.5 kb were eluted from an agarose gel and ligated into the BamHI site of pSL100 to yield a series of plasmids ( including pRBD14 and pRBD19 ) .
The structure , orientation , and length of the chromosomal fragments in the resulting plasmids were determined by restriction analysis .
Plasmids pRB18 to pRB85 ( Table 1 ) and pKZ26 were transformed into the following rfa indicator strains : S. typhimurium SL1032 ( rfaG ) , SL4805 ( rfaB ) , TV148 ( rfaf ) , and SL3750 ( rfaJ ) .
Correction of the mutant phenotypes was indicated by correction of the Fo-resistant phage resistance pattern of the test strains ( 11 ) .
Isolation of chromosomal rfa : : Mudlac insertions .
Random insertions of Mudl ( Apr lac ) into the chromosome of S. typhimurium TV119 ( rib ) were generated by coinfection with phages Mudl ( Apr lac ) and Mucts62hPl-1 as described by Csonka et al. ( 6 ) .
Ampicillin-resistant clones in which Mudl ( Apr lac ) was inserted into rfa genes were selected on the basis of their resistance to the rfb-specific bacteriophage 6SR .
The insertion mutants isolated in this way were further characterized and assigned to specific rfa genes on the basis of their phage resistance patterns ( 11 ) and the characteristic electrophoretic patterns of lipopolysaccharides isolated from the mutant strains ( 10 ) .
About 50 % of the rfa : : Mudlac mutants were Lac ' on indicator plates .
This result is consistent with the expectation that half of the insertions would be in the correct orientation to permit transcription to proceed from the mutated gene into the lacZ probe .
One of the Lac ' mutants ( AF21 [ rfaI : : Mudlac ] ) was used in the studies described here .
3-Galactosidase activity from the chromosomal rfaI : : MudlacZ insertion was approximately 10 % of the activity of a fully induced chromosomal copy of lacZ under control of Plac ' Enzyme measurements in rfaH + and rfall strains .
The rfaH + and rfaH strains used were M1174 and M1170 , respectively .
When P-galactosidase was assayed , the strains also contained F ' 481 ( lacIq lac : : TnS ) .
Assays were performed on parallel samples of cells harvested in mid-expo-nential growth .
, B-Galactosidase , P-lactamase , and CAT assays were performed as previously described ( 13 , 15 , 18 ) .
- ` Units are defined as originally described except for CAT , in which 1 unit is defined as 1 , umol of chloramophenicol acetylated in 10 min .
The results are expressed as the ratio of the activity of P-galactosidase or CAT to the activity of,-lactamase , thereby normalizing for possible differences in plasmid copy number in the different samples .
Unless otherwise described , procedures were performed as described by Miller ( 15 ) or Maniatis et al. ( 14 ) .
Physical maps and orientations of the plasmids described in Fig. 1 to 3 were determined by restriction analysis .
3 5 7 1 HEaPv EC P Pv R Hp I I I I I cH .1 I EN P N BHp H H !
I I * I 0 Loo 85 LO Lo Lb .
Lo0 18 67 71 66 55 81 44 78 i a ldagl ldall , a B ILa/Bla H dig + dadl q 1118 215 C M PV d a. I .
-0 Cam-resistance dad d & ¬ Cat/Bla pSL1 00 < 0.1 < 0.1 S S R pRB161 21.0 4.7 S R R pRB162 31.0 24.0 pRB187 0.8 0.6 S S pRB157 1.6 1.8 S S pRB1I1 3.3 nd nd S pRB103 3.9 nd nd S R pRB160 30.0 nd S pRBD19 0,65 17.5 5.0 R S pRBD14 9.7 2.0 nd s s FIG. 2 .
Promoter activities of rfa fragments in pSL100-derived plasmids .
The indicated plasmids , in which cat expression requires the presence of a promoter in the chromosomnal insert , were introduced into rfaH + and rfaH strains .
Activities of P-galactosidase ( LacZ ) and CAT and the resistance ( R ) or sensitivity ( S ) of each of the plasmid-containing strains to chloramphenicol are shown on the right of the figure .
In the diagram for the pSL100 vector , the solid line indicates the polycloning site .
The arrows indicate the location and orientation of the chromosomal insertions inserted into the polycloning site upstream of the cat reporter gene , relative to the physical map of the rfa locus shown in Fig. 1 .
pRB19 and pRB14 are 2 of the 22 different Bal 31 deletion mutant plasmids that were examined in this study , as described in the text .
Bg , BgII ; nd , not determined .
For other abbreviations , see the legend to Fig. 1 aC Hp Cat/Bla rfaH + 0 5 pSL1 02 pRB1 88 299.0 ( 1.0 ) 57.0 ( 0.19 ) 296.0 58.0 0 pRB1 07 pRB165 pRB1 09 pRB1 78 27.0 ( 0.09 ) 29.0 ( 0.1 ) 80.0 ( 0.27 ) 66.0 ( 0.22 ) nd m ¬ 22.0 nd 66.0 Q. 7 0 % L1 .4 .
a7Q71  - ¶ jJ 0 pS A n ¬ 45.8 30.9 4.6 5.3 0.7 * 27.7 18.8 5.0 4.1 FIG. 3 .
Effects of promoter-distal rfa fragments on transcription from exogenous promoters .
The indicated plasmids were introduced into rfaH + and rfaH strains .
Asterisks indicate that the cells were grown in the presence of IPTG .
The individual chromosomal inserts are shown below their locations in the physical map of the rfa locus ( Fig. 1 ) .
In the diagram for the pSL102 and pSL140 vectors , the solid line indicates the polycloning site .
For other details , see Fig. 1 and 2 .
pRB179 I RESULTS Organization of the rfa cluster .
To monitor transcription from different regions of the rfaGBIJ gene cluster , a lacZ transcriptional probe was inserted at random locations within plasmid pKZ26 , whose chromosomal insert include 44 3.5 CCW 85 178 0.5 64 4.8 CCW 111 215 0.51 59 5.5 CCW 96 156 0.61 61 5.7 CCW 72 110 0.65 78 7.35 CCW 403 582 0.09 a Plasmids pRB18 to pRB85 , containing the indicated rfa : : lacZ-transcriptional-fusions , were introduced into rfaH ' and rfaH strains , and activities of,-galactosidase were determined .
b The locations of the lacZ insertions are given from the left end of the chromosomal insert in pKZ26 ( Fig. 1 ) .
c CW ( clockwise ) and CCW ( counterclockwise ) refer to the orientation of lacZ relative to the genetic map of pBR322 ( the parent of pKZ26 ) .
d The indicated plasmid-encoded gene was inactivated by the lacZ insertion .
Where no gene is indicated , the rfa : : IacZ plasmid corrected all of the chromosomal mutations tested .
e Values are expressed relative to P-lactamase ( see Materials and Methods ) .
the rfaG , rfaB , rfaI , and rfaJ genes .
Fifteen independent insertions that were distributed throughout the 7.5-kb rfaGBIJ insert were selected for further study ( Fig. 1 ) .
Several of the lacZ insertions resulted in inactivation of the rfaG , rfaB , or rfaI genes of the parental plasmid , as shown by loss of the ability of the rfa : : lacZ plasmids to complement chromosomal rfaG , rfaB , or rfaI mutations ( Table 1 ) .
These presumably were located within coding sequences or essential upstream regions of the affected genes .
None of the lacZ insertions affected the ability of the other known rfa genes of pKZ26 to complement their corresponding chromosomal mutant alleles .
All of the probes that were oriented in the clockwise direction ( relative to the pBR322 genetic map ) showed significant levels of lacZ expression in rfaH + cells , as monitored by , B-galactosidase activity ( Table 1 ) .
In contrast,,-galactosidase activity from most of the counterclockwise inserts was very low , ranging from 8 to 15 % of the average activity from the clockwise inserts .
The only exceptions were inserts lacZSS and lacZ78 , which are located at the extreme ends of the chromosomal fragment .
Evidence that the rfaH gene product affects transcription of the genes of the rfaGBIJ cluster was obtained by comparing lacZ expression from the rfa : : lacZ plasmids in isogenic rfaH + and rfaH mutant cells .
In the case of the clockwiseoriented lacZ insertions , lacZ expression was significantly higher in rfaH + than in rfaH cells ( Table 1 ) .
In contrast , for the counterclockwise probes , P-galactosidase activities in rfaH + cells were actually lower than the activities in rfaHcells .
These results indicate that all of the genes within the locus are transcribed in the same direction .
The results also show that RfaH positively regulates rfaGBIJ expression at the transcriptional level .
Evidence that transcription of the rfaGBIJ cluster was driven by a promoter that lies upstream of rfaG was obtained by inserting fragments of the cloned rfa region into pSL100 , which lacks a promoter capable of driving transcription of the cat gene ( Fig. 2 ) .
Expression of the downstream cat gene therefore requires an intrinsic promoter within the cloned insert .
Insertion into pSL100 of fragments that included the leftward 1.1 kb of the chromosomal insert resulted in significant cat expression , as shown by resistance to chloramphenicol killing and/or by increased activity of CAT ( Fig. 2 , pRB161 , pRB160 , and pRBD19 ) .
In contrast , cells containing plasmids with rfa sequences that lay downstream of this region ( pRB187 , pRB157 , pRB101 , pRB103 , and pRBD14 ) remained sensitive to chloramphenicol and showed only 5 to 15 % of the CAT activity induced by plasmids containing the more leftward region of the locus .
These fragments therefore lacked significant intrinsic promoter activity .
The upstream promoter was more precisely mapped by constructing a series of deletions that extended into the chromosomal sequences from the left .
These revealed significant promoter activity from all of the several fragments whose left ends were located at or before 0.65 units on the map of the chromosomal insert ( illustrated by pRBD19 in Fig. 2 ) , as monitored by chloramophenicol resistance and by CAT assay .
Fragments that began at or after 0.7 map unit ( illustrated by pRBD14 in Fig. 2 ) lacked significant promoter activity by these criteria .
These results indicate that the in major promoter ( s ) associated with the rfaGBIJ locus ( Pfa Fig. 4 ) lies upstream of rfaG , between 0.65 and 0.7 kb from the left end of the original chromosomal insert .
When overlapping promoter-distal fragments that covered the 0.7-to 3.9-kb region were interposed between the ribosomal P2 promoter and the cat transcriptional monitor of pSL102 ( pRB188 , pRB107 , pRB165 , and pRB178 ) , there was a 78 to 91 % decrease in CAT expression as compared with that for the unsubstituted pSL102 vector .
A significant decrease in CAT expression was also seen with the nonoverlapping fragments present in pRB188 and pRB109 .
This suggests that one or more transcriptional terminators may be located within this 3.2-kb region of the rfaGBIJ locus .
Effect of rfaff on transcription of cloned rfa genes .
To further localize the sites responsible for the RfaH-mediated transcriptional regulation , we studied rfaGBIJ fragments in the pSL100 vector , where possible artifacts due to transcription into the insert from the vector are avoided .
This confirmed that all elements required for the rfaH-mediated transcriptional regulation were present in the leftward end of the rfa locus .
This was shown by the four-to fivefold-higher expression of the lacZ ( pRB158 ) and cat ( pRB161 ) reporter genes and by the chloramphenicol resistance associated with the presence of the plasmids in rfaH + but not rfaH cells ( pRB161 , pRB160 , and pRBD19 ) ( Fig. 2 ) .
As expected , when the 3.9-kb chromosomal fragment of pRB161 was inserted into pSL100 in the opposite orientation ( in pRB162 ) , there was no significant difference in cat expression between rfaH + and rfaH cells .
These results established that at least one site that was required for the rfaH-mediated transcriptional regulation was located between 0.65 unit ( the start of the insert in pRBD19 ) and 1.05 units ( the downstream end of the insert in pRB160 ) on the rfaGBIJ map .
We then attempted to separate the RfaH-responsive site from the rfaGBIJ promoter region ( Prfa ) by examining the effect of rfaH + and rfaH alleles on expression from promot-erless rfa fragments that were transcribed from the ribosomal P2 promoter of pSL102 ( Fig. 3 ) .
In all cases , transcripts that originated from the exogenous promoter wer TABLE 1 .
lacZ expression from rfa : : lacZ gene fusions ' 3-Galactosidase Lo ( kb ) ion activity for : Fusion Orientationc Gened rfaH-rfaH rfaH - / rfaH 85 2.1 CW rfaG 951 345 2.8 18 3.5 CW 1,217 454 2.7 56 3.9 CW 785 155 5.1 67 4.4 CW 652 130 5.0 53 4.8 CW 679 159 4.3 71 5.5 CW 753 237 3.2 73 5.7 CW 804 191 4.2 66 6.17 CW 639 146 4.4 rfaB rfaB rfaI rfal rfaI 55 0.2 CCW 380 531 0.72 81 2.2 CCW 130 254 0.51 rfaG rfaB rfaI rfaI 0 2 4 6 Strain Relevant genotype 1ac3t-iGvailtayct ( ousniidtass ) e ' H EC P Pv R Hp EN P N BHp H - , l li -- l a , D a S. typhimurium AF21-106 rfaI : : Mudlac rfaH 1.9 S. typhimurium AF21-107 rfal : : Mudlac rfaH + 25.9 S. yphimurium AF21 rfaI : : Mudlac rfaH930 .3 Pfa I ----- D -- I E. coli JK268 ( + IPTG ) 97.0 E. coli JK268 ( - IPTG ) 1.2 a P-galactosidase units are as defined by Miller ( 15 ) .
Diagram of the general organization of the rfaGBIJ region .
Only the major promoter activities ( P ) are indicated .
Terminators ( t ) are shown within the region in which their presence was suggested from the experiments shown in Fig. 3 ; the question mark indicates that their presence is speculative .
insensitive to RfaH-mediated regulation , as shown by the unchanged levels of CAT in rfaH + and rfaH hosts ( pRB188 , pRB165 , and pRB178 in Fig. 3 ) .
Because of the possibility that the high level of transcription from the strong P2 promoter might have masked the presence of RfaH-responsive sites in the promoterless inserts , we inserted the promoterless 3.2-kb fragment that lies immediately distal to Prfa into pSL140 , thereby placing it downstream of Plac ( pRB179 in Fig. 3 ) .
When pRB179 was studied under conditions where the level of cat expression in rfaH + cells was similar to expression from Prfa , there was no significant difference in the levels of CAT between rfaH + and rfaH cells ( Fig. 3 ) .
The inability to obtain promoterless fragments that retained their sensitivity to RfaH-mediated regulation indicates that at least one site needed for the transcriptional regulation was located in or very close to Prfa .
Effect of rfiaH on transcription of chromosomal rfal gene .
To confirm that the effect of rfaH expression on transcription of the plasmid-borne rfa genes was a valid representation of the behavior of the chromosomal genes , we inserted a lacZ transcriptional probe into the chromosomal rfaI gene by lysogenization with phage Mudl ( Apr lac ) .
P-Galactosi-dase activity from the chromosomal rfal : : MudlacZ insertion was approximately 5 % of the activity from the plasmidborne rfaI : : lacZ probes .
The difference presumably reflects the increased number of cellular copies of rfal : : lacZ with the plasmid-bome probes .
As shown in Table 2 , lacZ expression from the chromosomal rfaI : : lacZ probe was 14-fold higher in rfaH + than in rfaH cells .
This increase was significantly greater than the approximately fourfold stimulation of the corresponding plasmid-borne rfal : : lacZ constructs ( rfaI : : lacZ71 , rfaI : : lacZ73 , and rfaI : : lacZ66 [ Table 1 ] ) .
This may indicate that the protein responsible for the RfaH-mediated transcriptional regulation is present in amounts that are insufficient to activate all of the plasmid-borne targets .
DISCUSSION The present results suggest the model of transcriptional organization of the rfaGBIJ locus of S. typhimurium that is shown in Fig. 4 .
The entire leftward portion of the locus , including rfaG , rfaB , and rfaI , appears to be driven by one or more promoters located upstream of rfaG ( Pifa in Fig. 4 ) .
This conclusion is based on the absence of significant transcription into the lacZ and cat transcriptional probes from fragments that included the rfaGBI region but lacked the upstream sequences .
The conclusion is supported by the observation that significant expression was seen only with clockwise-oriented probes and that these were the only probes that showed the expected response to the presence or absence of the wild-type rfaH allele .
This suggests that the gene cluster forms an operon that is transcribed in the order rfaG-rfaB-rfaI ( - rfaJ ) .
Although rfaJ is likely to be part of the same unit , the possibility remains that rfaJ is transcribed from another promoter oriented in the counterclockwise direction , since none of the available transcriptional probes monitored clockwise transcripts within the rfaJ structural gene .
This is made less likely by the failure of lacZ78 to respond to the presence or absence of the rfaH + allele ( Table 1 ) , but it still can not be excluded .
It has been suggested that in E. coli the rfa gene cluster is also transcribed unidirectionally , with promoters located both upstream and within the locus ( 1 ) .
The present results also demonstrate that the RfaH-medi-ated regulation of rfaGBIJ operates at the transcriptional level , as has previously been shown for tra genes .
The results of the tra studies also suggested that the RfaHmediated positive regulation occurred by an antitermination mechanism , in which the rfaH gene product permitted transcription to proceed past terminators located within the transcriptional unit ( 3 , 8 ) .
A suggestion that a similar mechanism may be responsible for the RfaH-mediated regulation of rfaGBIJ in S. typhimurium has come from the demonstration that certain mutations in the gene for transcriptional termination protein rho can suppress the abnormal phenotype of rfaH mutants ( 7 ) .
Although consistent with the idea that RfaH acts as a transcriptional antiterminator , the present study does not establish that such a mechanism is responsible for the RfaH-mediated regulation of the rfaGBIJ cluster since it was not possible to separate the site ( s ) required for RfaH regulation from the upstream promoter ( s ) .
Therefore , the possibility that the rfaH system regulates rfaGBIJ expression at the level of initiation of transcription rather than at the level of premature termination can not be excluded .
If antitermination is involved , at least one essential site must be located very close to the upstream promoter ( s ) .
In the well-studied antitermination systems mediated by the XN and XQ proteins , the site at which the antitermination protein engages the transcriptional complex is distinct from and lies upstream of the termination sites at which the antitermination effect is manifested ( 17 ) .
If the RfaH-medi-ated transcriptional regulation occurred by such a mechanism , the inability to separate an upstream RfaH-responsive site from the promoter would resemble the situation with the XQ protein , in which the qut site responsible for interaction with Q overlaps the relevant promoter ( 9 , 22 ) 11 .
Hudson , H. , A. A. Lindberg , and B. A. D. Stocker .
Lipopolysaccharide core defects in Salmonella typhimurium which are resistant to Felix 0 phage but retain smooth character .
Kadam , S. K. , A. Rehemtulla , and K. E. Sanderson .
Cloning of the rfaG , B , I , and J genes for glycosyl transferase enzymes for the synthesis of the lipopolysaccharide core of Salmonella typhimurium .
Li , S. C. , C. L. Squires , and C. Squires .
Antitermination of E. coli rRNA transcription is caused by a control region segment containing lambda nut-like sequences .
Maniatis , T. , E. F. Fritsch , and J. Sambrook .
Molecular a cloning : laboratory manual .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 15 .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 16 .
In F. C. Neidhardt , J. L. Ingraham , K. B. Low , B. Magasanik , M. Schaechter , and H. E. Umbarger ( ed .
) , American Society for Microbiology , Washington , D.C. 17 .
Phage lambda and the regulation of transcription termination .
Ross , G. , and C. O'Callaghan .
Sanderson , K. E. , and J. R. Roth .
Linkage map of Salmonella typhimurium , edition VII .
Way , J. C. , M. A. Davis , D. Morisato , D. E. Roberts , and N. Kleckner .
New TnWO derivatives for transposon mutagenesis and for construction of lacZ operon fusions by transposition .
Wilkinson , R. G. , and B. A. D. Stocker .
Nonsmooth mutants of S. typhimurium : differentiation by phage sensitivity and genetic mapping .
Yang , X. , and J. W. Roberts .
Gene Q antiterminator proteins of E. coli phages 82 and lambda suppress pausing by RNA polymerase at a rho-dependent terminator and at other sites .