173 , No. 3 The nadI Region of Salmonella typhimurium Encodes a Bifunctional Regulatory Protein NING ZHU AND JOHN R. ROTH * Department of Biology , University of Utah , Salt Lake City , Utah 84112 Received 21 August 1990/Accepted 29 November 1990 Mutants of the nadI and pnuA genes were independently isolated on the basis of defects in repression of NAD biosynthetic genes and defects in transport nicotinamide mononucleotide ( NMN ) .
The mutations map at min 99 on the SalmoneUla chromosome , and the affected regions appear to be cotranscribed .
Some pairs of nadI and pnuA mutations complement , suggesting the existence of independent functions .
However , cis/trans tests with particular mutations provide evidence that both repressor and transport functions are actually performed by a single bifunctional protein .
( This result confirms sequencing data of Foster and coworkers [ J. W. Foster , Y. K. Park , T. Fenger , and M. P. Spector , J. Bacteriol .
We have designated the gene for this bifunctional protein nadI and distinguish the regulatory and transport defects with phenotypic designations ( R and T ) .
When a nad ( R-T ' ) mutation ( eliminating only repression function ) is placed cis to a superrepressor mutation , nadIf ( RS T - ) , the superrepression phenotype is lost .
In contrast , placement of R-and RS T-mutations in trans allows full superrepression .
This result suggests that the transport function ( eliminated by the R ' T-mutation ) and the repression function are provided by the same protein .
Insertion mutations in the promoter-proximal repressor region of the nadl gene eliminate transport function unless the inserted element can provide both for both transcription and translation start signals ; this finding suggests that there is no transcriptional or translational start between the regions encoding repression and transport functions .
In Salmonella typhimurium , NAD is synthesized both by a de novo pathway and by two salvage pathways .
These pathways are diagrammed in Fig. 1 of the accompanying report ( 34 ) .
Transcriptional regulation of this pathway is accomplished by a repressor , encoded by the nadl gene ; this repressor controls transcription of genes for the first two biosynthetic enzymes , nadB and nadA ( 10 , 14 , 17 , 32 ) .
The nadI repressor gene maps at min 99 of the Salmonella chromosome , very near a gene involved in nicotinamide mononucleotide ( NMN ) transport , pnuA ( 10 , 14 , 15 , 17 , 20 , 29 , 32 ) .
Previous work on the functional and genetic relationship between the nadI and pnuA genes led to the conclusion that the region contains either a single bifunctional protein or perhaps two independent genes in one operon ( 10 , 14 , 32 ) .
Foster et al. ( who designate this region nadR ) have found that the DNA sequence for this region includes a single open reading frame ( 15a ) .
We present here genetic data supporting their conclusion , providing in-vivo evidence for the existence of a single bifunctional gene and further characterizing the genetic structure of this locus .
We have designated the gene for the inferred bifunctional protein nadl and distinguish the regulatory and transport defects with phenotypic designations ( R and T ) .
In the accompanying paper ( 34 ) , we present evidence that both functions of this protein are regulatory .
One function ( R ) controls transcription of two biosynthetic genes ; the other ( T ) serves to regulate the activity of the NMN transport function .
Both functions appear to exert their control in response to internal NAD ( or NADP ) levels .
MATERIALS AND METHODS Bacterial strains .
All strains used in this study are derived from S. typhimurium LT2 and are listed in Table 1 .
Mu dA refers to a conditionally transposition-defective derivative ( 18 ) of the original Mu dl ( Lac Apr ) phage of Casadaban and Cohen which forms operon fusions ( 4 ) .
Mu dJ refers to a transposition-defective mini-Mu phage , Mu dl-1734 ( Lac Kmi ) , constructed by Castilho et al. ( 5 ) ; this phage lacks transposition functions and carries kanamycin resistance .
TnJOd ( Tc ) refers to a small transposition-defective derivative of TnJO ( TnJO Dell6 Dell7 constructed Tet ) by Way et al. ( 31 ) .
TnJOd ( Cm ) refers to a transposition-defective deriv-ative of transposon TnJO constructed by Elliott and Roth ( 13 ) .
The E medium of Vogel and Bonner ( 30 ) , supplemented with 0.2 % glucose , was used as the minimal-medium .
Difco nutrient broth ( NB ; 8 g/liter with 5 g of NaCl per liter ) was used as the rich-medium .
Difco agar was added at a final concentration of 1.5 % for solid medium .
Nutrients to feed auxotrophs were included in minimal media at final concentrations described by Davis et al. ( 11 ) ; exceptions are indicated in the text .
Antibiotics were added to media at the following final concentrations : ampicillin ( sodium salt ) , 30 , ug/ml in NB and 15 , ug/ml in E medium ; tetracycline hydrochloride and chloramphenicol , 20 , ug/ml in NB and 10 , ug/ml in E medium ; and kanamycin sulfate , 50 , ug/ml in NB and 125 jig/ml in E medium .
All antibiotics were obtained from Sigma Chemical Co. .
Media containing ampicillin were prepared fresh before use .
The chromogenic 3-galactosidase substrate 5-bromo-4-chloro-3-indolyl-i-D-galactoside ( XGal ) was dissolved in N,N-dimethyl formamide ( 20 mg/ml ) and added to media at a final concentration of 25 , g/ml .
Similarly , the chromogenic substrate for alkaline phosphate , 5-bromo-4-chloro-3-indolylphosphate , toluidine salt ( X-P ) , was dissolved in N,N-dimethyl formamide ( 20 mg/ml ) and added to media to a final concentration of 50 , ug/ml .
The high-frequency , generalized transducing bacteriophage P22 mutant HT105/1 int-201 was used for all transductional crosses .
This phage was derived by Roberts ( 25a ) from the P22 HT105/1 phage of Schmieger ( 27 ) .
To select for the inheritance of the Kmr marker of Mu dJ and the Cmr marker of TnJOd ( Cm ) , the transduction mixture of cells and phages was spread on NB plates and incubated overnight before replica printing to selective plates .
In all other crosses , selective plates were spread directly with 2 x 108 cells and 108 to 109 phages .
Phage-free transductant clones were identified by nonselective streaking on green indicator plates ( 6 ) .
Phage-free clones form light-colored colonies on green indicator plates ; phage sensitivity was confirmed by cross-streaking with P22 H5 phage , a clear-plaque mutant of P22 .
Insertion mutations generated by Mu dJ and Mu dK were isolated by the cis complementation method described previously ( 19 ) .
Deletion mapping was done by phage P22-mediated transductional crosses .
A fresh overnight culture of a strain carrying deletion from the serB to nadI region ( 109 cells per ml ) was washed , concentrated 10-fold by centrifugation , and then infected at a multiplicity of 10 with transducing lysates grown on serB + nadI ( T - ) mutant strains .
Before mapping , nadI ( R T + ) insertion mutants were converted to nadcI ( R-T - ) deletions ( see text ) .
Transductants that are Ser + and able to transport NMN were selected on minimal-medium with i0 ' M NMN .
After 2 days of incubation on selective medium , a set of recombinant colonies was patched and all markers were scored by replica printing .
A wild-type donor typically yields more than 105 recombinant colonies ( SerB + NMN + ) per plate under these conditions .
A cross was scored as negative if no recombinants were seen on any of five plates .
Thus , the resolution of the map is 2 x 10-6 .
P-Galactosidase activity was determined as described by Miller ( 23 ) , with sodium dodecyl sulfate-chlo-roform-permeabilized cells .
The j-galactosidase activity i reported as nanomoles per minute per optical density unit ( at 650 nm ) of cells .
b Nomenclature is as described by Demerec et al. ( 12 ) , Chumley et al. ( 7 ) , and Schmid and Roth ( 26 ) .
Construction of nad ( R-T + ) and nadl ( RS T - ) double mutants .
The double mutant used in the cisltrans test was constructed by transducing a serB nadIS ] ( RS T - ) strain ( TT16032 ) with phage grown on the nadI260 ( R-T ) mutant ( TT10091 ) ; Ser + transductants were scored for inability to use NMN ( nadISIl ; RS T - ) and constitutive expression of the nadB : : lac fusion ( inheritance of nadI260 ) .
Recombinants with both phenotypes were expected to be nadI260 nadPS1 double mutants ; this structure was confirmed by crosses which recovered each of the individual mutations from the double mutant .
RESULTS Isolation of nadil insertion mutations that form lac operon and protein fusions .
Insertions of elements Mu dJ and Mu dK in the nadI gene were identified by a defect in repressor function .
These insertions were isolated as prototrophic revertants of nadI509 ( RS T - ) mutant strains ( TT13278 and TT11420 ) .
The parental nadI509 ( RS T - ) mutation causes a Nad-auxotrophic phenotype by tight repression of NAD synthetic genes ( 32 ) .
Revertant mutations permitting pyri-dine-independent growth map at the nadI locus and have lost repressor function .
These mutants were screened for their ability to transport NMN .
Since the parent nadI ( Rs ) mutation itself causes a transport-defective phenotype , the new nadI ( R - ) insertions had to be genetically separated before their transport phenotype could be checked .
Some , but not all , of the separated nadI ( R - ) insertion mutations eliminated NMN transport ; the mutations were designated nadI : : Mu d ( R-T ) ( repressor deficient and transport proficient ) or nadI : : Mu d ( R-T - ) ( repressor deficient and transport deficient ) , respectively .
The nadI : : Mu d ( R-T ) insertions that form blue colonies on medium containing X-Gal have formed lac operon fusions or lacZ protein fusions to the nadI gene .
Simple transport-deficient Mu dJ insertions were isolated from a nadI + nadB pncA parent ( TT13120 ) as derivatives that fail to use NMN as a pyridine source .
All such insertions are either in the pnuC gene ( located near the nadA gene at min 17 ) or in the nadl gene near serB at min 99 .
The nadI ( T - ) insertions were all found to be deficient in the repressor function ; i.e. , they are all nadI : : Mu dJ ( R-T - ) insertions .
The insertion mutants that formed blue colonies on medium containing X-Gal were presumed to be transcribed by the nadI promoter ; Expression of the nadI ( R-T - ) and nad ( R-T + ) fusions is not influenced by exogenous nicotinic acid , nicotinamide , or NMN ( see below ) .
By using a method described previously ( 7 , 22 ) , an F ' lac plasmid was inserted into the chromosome by recombination between a chromosomal nadI : : Mu d ( lac ) element and the lac sequences of the plasmid .
The direction of chromosome transfer by the resulting Hfr strain ( data not shown ) , indicated that both nadI : : Mu dJ ( R-T ) and nadI : : Mu dJ ( R-T - ) fusions were transcribed in a clockwise direction ( serB -- nadI-thr ) that is , transcription of the nadI region starts at the end nearest the serB gene and proceeds toward the thr locus .
The orientation of transcription was confirmed by studying spontaneous deletion mutations in the region as described below .
Map order of nal ( R-T + ) and nad !
Repressor mutations , nadl ( R - ) , and NMN transport mutations , nadI ( T - ) , previously designated pnuA , both map between the serB and thr genes ( 10 ) .
However , their map order with respect to each other is uncertain .
We determined their order by the phenotypes of deletion mutations and by three-factor crosses .
The results ( see below ) support the map order ( read clockwise ) serB-nadI ( R ) - nadI ( T ) - thr .
The nadI : : Mu dJ ( R-T + ) insertion mutants can use NMN as a pyridine source .
Deletions extending from serB to a nadI : : Mu dJ ( R-T + ) insertion ( e.g. , TT16214 ) remove the material between serB and the nadI : : Mu dJ insertion , including the nadI promoter , but leave the Kmr determinant of Mu dJ and all of the nadI material downstream of the Mu dJ insertion site .
These deletion strains still express the nadI transport function , since they can grow on lo-M NMN as does the parent insertion mutant .
Deletion mutants generated by recombination between the Mu dJ sequences of mutations nadI : : Mu dJ ( R-T + ) and thr : : Mu dA ( e.g. , TT16156 and TT16176 ) were tested for growth on 1o-4 M NMN .
These mutants have lost material between nadl : : Mu dJ and thr : : Mu dA , including the distal end of the nadI gene ; they retain nadI material upstream from the Mu dJ insertion site .
Consistent with these findings , the nadI : : Mu dJ-thr : : Mu dA deletion mutants have become NMN - .
The order of functions in this region has also been tested by three-factor crosses involving a nadI : : Mu dJ ( R-T + ) insertion mutant ( TT16205 ) and a nadI ( R-T + ) point mutant ( TT16234 ) used as recipients and a serB : : TnJOd ( Tc ) nadI ( R + T - ) strain ( TT16176 ) used as the donor .
The order of these mutations supports the order serB-nadI ( R ) - nadI ( T ) - thr .
The nadl ( RS ) mutations directly affect the NMN transport function .
The nadI ( RS ) mutants were isolated because they show constitutive repression of the synthetic genes nadA and nadB ; this repression is sufficient to cause an auxo-trophic phenotype .
All nadI ( Rs ) mutants also prove to be defective in NMN utilization ( 32 ) .
( In the accompanying paper [ 34 ] , we show that the transport deficiency is not due to a failure to express the regulated transport function pnuC .
) The transport defect of nadI ( RS ) mutations suggested that mutations affecting repression in this way might directly affect the transport function of a bifunctional NadI protein .
In the course of mutant isolation ( described above ) , it was found that nadI : : Mu dK ( R-T + ) mutations suppress the auxotrophic phenotype of nadI ( Rs ) mutations , presumably by preventing superrepression .
However , these double mutants ( like the single RS parental mutant ) fail to use NMN .
Thus , the nadl ( RS ) mutation continues to shows its effect on NMN transport when in combination with a mutation that abolishes the repressor function and restores expression of nadI biosynthetic genes .
This indicates that the nadI ( RS ) mutations do not lose transport function because of superrepression but may directly affect the transport region of the nadI protein .
Three-point crosses and a deletion map of the nadI region ( see below ) clearly demonstrate that nadIs mutations map at the promoter-distal ( transport ) end of the nadI gene .
The existence of nadl ( RS ) mutations with a transport defect supports the idea of a single NadI ( repression-transport ) protein but does not eliminate the possibility of a complex of separate repressor and transport proteins .
Deletion map of the nadl-pnuA region .
A deletion map of the nadI region was constructed by using spontaneous deletion mutants , isolated as Tcs revertants of a serB : : TnJOd ( Tc ) strain ( TT15490 ) .
Mutants selected as Tcs were screened for the ability to use NMN ( transport ) and for constitutive expression of the nadB499 : : Mu dJ fusion ( repression ) .
All serB-nadI deletion mutants are ( R-T - ) .
The nadI insertion or point mutation strains used in map construction are all serB + .
Point mutants with an NMN phenotype ( either R + T - , R-T - , or Rs T - ) were directly crossed with deletion mutants , selecting for growth on NMN as a pyridine source and SerB + prototrophy .
Mapping of mutations with an NMN + phenotype ( R-T + ) was more difficult .
This was done only for the Mu dJ and Mu dK insertion mutations which retained an NMN + phenotype .
For each such mutation , a nadI-thr deletion mutation was constructed by recombination between these Mu d elements and thr : : Mu dA insertions .
These deletions leave the promoter end of the nadI gene unchanged up to the insertion site but make cells phenotypically NMN-and Thr - ; the new phenotypes make it possible to map the insertion site .
Mapping crosses were done by P22 transduction crosses , using serB-nadI deletion mutants ( R-T - ) as recipients and the nadl-thr deletion mutants as donors ; transductants with a Ser + NMN + Thr + phenotype were selected .
This selection demands an exchange between the donor and recipient deletions .
( This method does not allow mapping of R-T + point mutations .
) The deletion map is shown in Fig. 1 .
All R-T + insertion mutations map at the left end of the nadI gene ( nearest to the serB gene ) .
The R + T-and RS T-mutations map at the right end of nadI ( closest to thr ) .
The R-T-mutations are scattered throughout the gene .
The deletion map confirms the map order serB-nadI ( R ) - nadI ( T ) - thr and the conclusion that nadr mutations are located within the region encoding transport function .
The nadI ( R - ) and nadl ( T - ) mutations show complementation .
Three distinct recessive phenotypes have been seen for nadI mutations : R-T + , R + T - , and R-T - .
This suggests that the transport and repression functions are distinct and each can be present without the other , but null mutations for the protein destroy both functions .
If this is the case , one should see complementation between R-T + and R + Tmutations .
The result of a complementation test is presented in Table 2 .
The merodiploid ( TT16005 ) carries both the nadI260 ( R-T ) and the nadI299 ( R + T - ) alleles and is proficient for both regulation and NMN transport , while each haploid with a single allele is defective for one or the other function .
This suggests that the two functions are independent .
Below we present evidence that these two independent functions are provided by a single protein .
Dominance test of the two ' phenotypes of nadlP mutations ; superrepression is dominant and NMN utilization deficiency is recessive .
Both phenotypes of nadI ( Rs T - ) mutations have been tested for dominance in a strain carrying a tandem duplication of the nadI region , as described previously ( 32 ) .
The nadI ( Rs T - ) InadI ( R + T + ) merodiploid strain is phenotypically RS T + ; that is , it shows superrepression of a nadB : : lac fusion but is able to transport NMN ( 32 ; Table 3 ) .
Thus , the superrepression phenotype of the nadf mutation is dominant to the wild-type allele , while its NMN transport deficiency phenotype is recessive .
cisltrans complementatio n'tests indicate a single nadI gene .
The cisltrans complementation tests were done by using tandem nadl duplication strains which also carry a nadB227 : : Mu dA insertion ( fusing lac to the nadB gene promoter ) and a pncA278 : : TnJOd ( Cm ) insertion ( to block alternative routes of NMN utilization ) .
These'outside mutations allow scoring of repression and NMN transport phenotypes .
The R-T + mutation chosen for this test is nadI260 ( R-T + ) ; this is a nonpolar point mutation by the criterion of having no polar effect on expression of a downstream nadl : : lac fusion ( data not shown ) .
Mutation nadISJJ ( RS T - ) was used to represent nadI ( RS ) mutations .
For the test of the trans configuration , strain TT15998 [ nadI260 ( R-T + ) InadP51i ( R ' T - ) ] was constructed by transducing the nadISJJ1 mutation into a serB9 nadI260O serB9 nadI260 homozygous duplication strain , selecting Ser + , and scoring for inheritance of the nadISJJ ( RS T - ) mutation .
The cis configuration strain TT16039 [ nadI260 ( R-T + ) nadISJJ ( Rs T - ) InadI + ] was constructed by first ` constructing a nadI260 nad1 ` 5ll double mutant ( see Materials and Methods ) .
This double mutant was used to transduce a serB9 nadI + pnuA + IserB9 nadI + pnuA + duplication strain ; Ser + transductants were checked for possession of both of the donor nadI mutations .
The structures of both cis and trans configuration strains were verified by segregation tests ; all alleles were shown to be arranged as expected .
Note that in haploid strains , the superrepression phenotype of the nadI ( RS ) mutation ( Table 3 , line 3 ) is substantially eliminated by an R-mutation placed cis to R ' in a haploid strain ( line 6 ) .
This is expected since superrepression requires the repressor function .
The cisltrans complementation test ( lines 1 and 4 ) demonstrates that the dominant superrepression phenotype is sixfold stronger if the Rs allele is located cis to a functional R + region ( trans to the Rmutation ) .
The residual repression seen by the R-RS double mutant allele ( line 6 ) is expected since the R-mutation used does not completely eliminate repressor function ( line 2 ) ; the residual dominant superrepression seen for the cis diploid ( Rs R - ) ( line 4 ) is presumably due to this residual repressor function of the R-Rs allele .
Some of this residual dominance could also be due to repressor subunit mixing .
Two types of nad !
: : Mu d ( R-T + ) insertions .
As described above , some of the R-insertions of Mu dl and Mu dK are phenotypically R-T .
If the nadl region encodes a single protein , the distal portion ( T ) of this protein must be produced in these mutants despite the upstream insertion mutation .
When examined more closely , the R-T + insertions fall into two classes ( types I and II ) .
Type I and type II insertions are similar in that they are able to suppress the auxotrophic phenotype ` of an nadl ( Rs ) mutation ( the basis for their isolation ) .
Both types map at the nadI locus and retain NMN transport ability .
Type I insertions cause only slight relief of the superrepression phenotype of the parental Rs mutation ; type II mutations appear nearly devoid of repressor function ( Table 4 ) .
Type I insertions all map in the left-most deletion interval at the promoter end of the nadI region , while type II insertions are in the second deletion interval ( Fig. 1 ) .
The NMN transport ability of type I insertions is expressed at high-temperatures , while that of type II insertion is not ( explained below ) ; polar insertions in the serB gene show a strong polar effect on lac expression from type I insertions and only a weak effect on type II insertions , indicating that the lac operon of a type I insertion is transcribed mainly from the serB promoter and that of a type II insertion is expressed mainly by a promoter outside of serB , probably the nadI promoter ( Table 5 , line 1 to 3 ) .
Our interpretation of these results ( explained in more detail below ) is that type I insertions actually lie outside of the nadI gene ( to the left of the nadI promoter ) and reduce nadI expression by blocking its transcription from serB , while type II insertions are within the nadI structural gene .
Transcription from serB extends through the nadI region .
A serB insertion mutation was found to reduce slightly the , B-galactosidase level of standard nadI : : lac fusions .
It seems likely that this decrease is due to termination of serB transcripts , which may normally extend through nadL .
Two insertion mutations serBJ463 : : TnJO and serB1466 : : TnJOd ( Tc ) were used in these tests .
The complete TnWO element provides a promoter for adjacent sequences , and the deletion derivative TnJOd ( Tc ) does not .
If the effect of a serB insertion on nadI is due to transcription termination , rather than a simple lack of the serB protein , the two types of TnWO insertions might have distinct effects on expression of a nadI : : lac fusion .
Strains were constructed with combinations of serB alleles and various lac fusion alleles of the nadI region .
The results for a standard nadI fusion ( Table 5 , lines 7 to 9 ) show that nadI transcription is reduced more by the serB : : TnJOd ( Tc ) insert ( TT16218 ) than it is in the strain with the serB : : TnJO insertion ( TT16222 ) .
For the type II fusion the same general effect is seen , but levels for this fusion are too low to permit evaluation ( lines 4 to 6 ) .
These results suggest that transcription from the outward-directed promoter of serB1463 : : TnJO may be able to extend through the nadI locus .
The reduced transcription of nadI caused by a serB insertion is probably not caused by a simple lack of the SerB protein .
While these effects are small , the idea that transcription from serB might contribute to expression of the nadI gene is also supported by the observation that the serB1466 : : TnJOd ( Tc ) insertion partially suppresses the auxo-trophic phenotype of the nadIP mutants , while the serB1463 : : TnJO insertion provides no suppression .
We presume that a slight reduction in level of expression of superrepressor protein allows improved expression of the nadA and nadB biosynthetic genes .
Thus , type I Mu dJ insertions and serB : : TnJO insertions may reduce nadI expression by blocking transcription of nadl from the serB promoter .
Expression of transport functions by type II nadl : : Mu d ( R-T + ) insertion mutants is due to a promoter and initiator supplied by the c end of Mu .
The observation of type II Mu d insertions ( R-T + ) was surprising if the nadl gene encodes a single bifunctional protein .
How is the distal transport function expressed in these mutants ?
A possible answer is that the c end of the inserted Mu sequence might provide both a promoter and an initiation signal for expression of adjacent host sequences .
This translational start site could be within Mu d sequence or could be formed at the junction of Mu d element and host sequence .
Several lines of evidence support this hypothesis .
The type II insertion mutations were found at a very low frequency and were all Mu dJ or Mu dK insertions ( two were found among 71 nadI-pnuA insertions ) ; no TnlO or TnJOd ( Tc ) insertion mutants of the R-T + type have been found .
Transposon TnlO is known to have outward-directed promoters in its ends , but no translational start site has been detected ( 1 , 9 , 28 ) .
Transposon TnlOd ( Tc ) does not have the outward promoter ( 33 ) .
The type II Mu d insertions are all oriented with the c end of Mu toward the downstream part of the nadI locus .
However , not every Mu d insertion oriented in this way expresses the transport activity .
This finding suggests that some special requirements must be met to permit expression of the transport portion of the NadI protein ( e.g. , a Mu d element may have to be in a proper orientation , be located outside of sequences required for the transport function , and be placed in a proper reading frame ) .
The Mu dJ and Mu dK elements retain a short Mu sequence at their left end ; the Mu c gene is the only intact gene included ( 5 ) .
Transcription of the Mu c gene is toward the end of the Mu genome ( 25 ) , and no strong transcriptional terminator could be found in this region of Mu by computer analysis ( 3 ) .
Therefore , the Mu c gene promoter could provide transcripts that extend out of these Mu d elements .
Recently , the mRNA for the Mu c gene has been found to be temperature sensitive ( 16a ) .
This provides a way to test our hypothesis .
Type I and type II mutants were tested for growth on lo-M NMN at 30 , 37 , and 42 °C with nadI + strain as the control .
The type II mutants are NMN-at 42 °C , while the type I and wild-type strains transport NMN at all temperatures .
We hypothesize that type I inserts merely interrupt the ser expression slightly ( see above transcript and reduce nadI and Fig. 2 ) .
Type II insertions , we suggest , interrupt both nadl and serB transcripts , providing a downstream promoter and ( if placed in a proper reading frame ) initiator an to of distal portion of the nadl express the transport portion the gene .
provide both The hypothesis that Mu d elements a pro-tested in the his A moter and an initiator was operon .
strong rho-dependent transcriptional terminator is located inside TnJO the When element is placed hisG gene ( 8 ) .
a upstream of this the outward TnJO transcripts terminated and site , are the next gene , hisD , is not expressed ( 8 , 9 ) .
However , when a particular Mu dJ element is inserted proximal promoter to the rho-dependent terminator site , the strain remains HisD + ( 9a ) .
This finding that the Mu dJ provides suggests sequence a promoter activity and , more important , that its transcripts seem to be translated since they can avoid termination .
We believe that these Mu signals permitted isolation of the rare R-T + insertions .
ZHU AND ROTH J. BACTERIOL .
173 , 1991 THE BIFUNCTIONAL nadI GENE serB thr nadI ( pnuA ) r51I ( Rs T - ) Pt 509 ( RS T - ) 508 ( RST - ) pt [ 312 ( R+T - ) 1299 ( R + T - ) V 579 V ( 584 ' - [ Vt556642 ( 0ll ) ) 585 V 557 ( I ) -565 V ( 4 ( 583 60 58 V556 ( I ) V 561 ( 11 ) 555 ( I ) 560 554 ( I ) 5530110 549 ( I ) 607 serB1466 : : Tnl0d ( Tc ) 5748 ( ) V 573 -- O 574 - ~ ~ .
I 570 7V ( 597-807-817 2 v 58 9 58 V 576 V ( 563 v 596 v 5 5889885 567 606-593-591-575 603 -.600 568-587-566 598 v 558-8966-569 gf579 592 589 J 821 812 , 806 825-800-787 794 FIG. 1 .
Deletion map of the nadl-pnuA region .
Deletion mutations , listed below the heavy horizontal line designating the chromosome , were all isolated as spontaneous tetracycline-sensitive revertants of the serBl644 : : TnIOd ( Tc ) mutant .
The allele numbers are indicated above each deletion .
The deletion endpoints are indicated by vertical dotted lines .
The types of insertion and point mutations mapped are indicated as allele numbers above each deletion interval .
The letter J in a dark triangle indicates a Mu dJ insertion ; the letter K designates a Mu dK insertion .
The arrows above these insertion symbols indicate the orientation of their lac operon .
A dark triangle with a lowercase `` t '' denotes a TnJOd ( Tc ) insertion ; an uppercase `` T '' indicates a TnJO insertion .
Point mutations are designated Pt .
The phenotypes of insertion and point mutations are shown in parentheses , with R indicating repression and T indicating NMN transport ( R-T - , R-T + , or R + T - ) .
Superrepressor mutations are designated RS T-since they all show a transport defect as well as superrepression .
For Mu d insertions , type I indicates a nadI ( R + T + ) insertion and type II indicates nadI ( R-T + ) insertion .
Complementation test between the nadI ( R-T + ) and nadI ( R + T - ) mutations P-Galactosidase activity ( U ) in cells grown with indicated NA concn lo-4 M NMN Growth on StrSatirnain Relevant ggeennoottyyppee 10-6 M 357-342-321 2 x 10-4 M 5 92 + TT16005 Segregant Segregant nadB227 : : Mu dA DUP728 [ ( nadI260 ( R-T + ) ) * ( nadI299 ( R + T - ) ) ] nadB227 : : Mu dA nadI260 ( R-T + ) nadB227 : : Mu dA nadI229 ( R + T - ) + ¬ ZHU AND ROTH J. BACTERIOL .
cis effect of the nadI260 ( R-T + ) mutation on the phenotype of nadisSi P-Galactosidase activity ( U ) in cells grown with indicated NA concn lo6M 2 x14M 5 3 361 81 2 2 30 4 303 6 111 57 nadI allele ( s ) Growth on 10-4 M NMN LinL Straint e inSe a raina Copy I Copy II 1 TT15998 2 Segregant + + + + ¬ nadI260 ( R-T + ) nadI260 ( R-T + ) nadPSll ( RS T ) nadPS ( RST - ) 3 Segregant 4 TT16039 5 Segregant 6 Segregant a All contain a nadB227 : : Mu dA fusion .
nadI + ( R + T + ) nadI + ( R + T + ) nadI260 ( R-T + ) nadISJOO ( RS T - ) nadI260 ( R-T + ) nadIsSJ ( RS T - ) TABLE 4 .
Effect of the type I and type II nadl : : Mu d ( R-T + ) insertions on expression of the nadB : : lac fusion P-Galactosidase activity ( U ) in cells grown with indicated NA concn 1o-6 M 2 x 1O-4 M 3 3 13 6 257-163-247 6 281 9 307 14 Strain Relevant genotype TT16202 nadB227 : : Mu dA nadISO09 TT16208 TT16206 TT16201 TT16207 TT16205 nadB227 : : Mu dA nadPSO9 nadISS4 : : Mu dK ( type I ) nadB227 : : Mu dA nadIsSO9 nadISS3 : : Mu dK ( type II ) nadB227 : : Mu dA naddB227 : : Mu dA nadISS4 : : Mu dK ( type I ) nadB227 : : Mu dA nadISS3 : : Mu dK ( type II ) TABLE 5 .
Effect of serB insertions on expression of the type I and type II nadI : : Mu d ( R-T + ) fusions P-Galactosidase activity ( U ) in cells grown with indicated NA concn 10-6 M 2 x 10-4 M 61 59 1 1 25 30 5 5 3 3 4 4 69 57 32 29 47 46 Line Strains Relevant genotype TT15926 FT16220 TT16224 FT15924 FT16219 Ff16223 FT15922 FT16218 Ff16222 nadI562 : : Mu dJ ( type I ; R+T + ) 1 2 3 4 5 6 7 8 9 serB1466 : : TnlOd ( Tc ) nadI562 : : Mu dJ ( type I ) serB1463 : : TnlO nadI562 : : Mu dJ ( type I ) nadI561 : : Mu dJ ( type II ; R-T + ) serBJ466 : : TnlOd ( Tc ) nadI561 : : Mu W ( type II ) serB1463 : : TnlO nadI561 : : Mu dJ ( type II ) nadI563 : : Mu dJ ( R-T - ) serB1466 : : TnlOd ( Tc ) nadI563 : : Mu dJ ( R-T - ) serB1463 : : TnJO nadI563 : : Mu dJ ( R-T - ) nedi P I ( b ) .
A type I Mu d Insertion ( R-T ) typeI serB nodl .
A type I1I Mu d Insertion ( R T ) serB I type I nudl PI PiI w-il-l - .
A standard Mu d insertion ( R T serB R T nedi PI ' 1F I ) P .
Hypothesis for transcription of the nadI gene .
( a ) In a wild-type strain , nadI is expressed from both the serB promoter and the nadI promoter .
( b ) Type I insertions block serB transcripts and slightly reduce expression of nadL .
( c ) Type II insertions block both serB and nadI transcripts and express the downstream portion of nadI from its own promoter and initiator .
( d ) Other Mu d insertions ( R-T - ) in the same region as type II insertions may fail to express the transport portion of the nadI gene because they insert in an improper reading frame .
DISCUSSION Mutants of the nadI gene were initially isolated by virtue of two distinct phenotypes .
Mutations causing a defect in NMN transport were designated pnuA ; mutations showing constitutive transcription of the nadA and nadB biosynthetic genes were designated nadI by us ( nadR by Foster and co-workers [ 14 ] ) .
In this report , we present genetic evidence that the two functions are actually provided by a single bifunctional protein .
These genetic data provide functional , in-vivo confirmation of conclusions of Foster and co-work-ers , who have found a single open reading frame in the base sequence of DNA from this region ( 15a ) .
Using mutations that cause loss of one of the functions , we can show that the two functions of the nadI protein can be separated .
The fact that R-T + mutations complement R + T-mutations indicates that , at least in some mutant proteins , the two functions can act independently .
However , these two functions are not always independent .
Some Tmutations in the distal region of the gene eliminate both repression and transport , and the NMN transport function can be altered so as to interfere with the repressor function .
This finding initially suggested the possibility that the repressor and NMN transport functions are provided by a single protein .
In addition , superrepressor mutations map at the distal ( transport ) end of the nadI gene .
These point mutations cause both a dominant superrepression phenotype and a recessive NMN transport deficiency , strongly suggesting that the two distinct functions act in concert .
Furthermore , double mutant strains with an R-T + mutation and an RS Tmutation lose the superrepression phenotype ( auxotrophy ) but maintain their NMN transport defect , suggesting that the Rs mutation causes loss of transport activity because of a direct effect on the transport function of the nadl protein .
We that mutations cause a in presume nadl ( Rs T - ) change the transport portion of the protein such that the repressor portion of the molecule is active under all cellular conditions .
These data alone could be explained by a bifunctional protein or by a bifunctional complex of two nonidentical polypeptides .
A cisltrans complementation test provided support for a single bifunctional nadI protein .
A nadP mutation normally causes a superrepression phenotype ( RS ) that is dominant to a wild-type allele of nadL .
The superrepression phenotype is eliminated in a haploid strain if a nadl ( R-T + ) mutation is cis to the placed nadI ( Rs T - ) mutation .
This is expected , since repression ability is required for superrepression .
If the repressor function were encoded by a gene distinct from that function , one would a nadI ( R + ) providing transport expect allele , placed in trans , to completely restore the superrepression phenotype of the nadI ( RS T - ) nadI ( R-T + ) double mutant ; this is not seen .
For a maximal superrepression phenotype , the nadI ( Rs ) mutation must be located cis to a functional repressor domain .
This is most easily explained by the existence of a single nadI protein .
Another support for the one-protein hypothesis is the existence of II nadI insertions .
II nadI insertions type Type are Mu d insertions at the particular promoter-proximal end of the nadI that are able to the region express transport function .
Other insertions of the same of type element , in the same orientation , in the same deletion interval show an R-T-phenotype .
Type II insertions would not be expected if the nadI included a unless these inserregion single protein tions could provide both a promoter and a protein initiation site to allow expression of the distal portion of the protein .
Evidence is presented that Mu d insertions can actually provide both of these signals .
The nadI ( RS T - ) mutations have two phenotypes , superrepression and NMN transport deficiency .
The RS phenotype is dominant and the T-phenotype is recessive to th wild-type allele .
This supports the idea that the transport domain of the NadI protein can be altered so as to eliminate transport and qualitatively alter repressor function .
More specifically , it predicts that the transport domain of the nadI protein may be a signal receiver as well as a component of the NMN transport system .
Occupancy of the signal receiver by NAD may determine whether the repressor assumes a conformation able to bind to the nadA and nadB operator sites .
The nadI ( Rs ) mutations could have changed the transport domain in such a way that it mimics the signal-received state and locks the NadI repressor domain into the DNA-binding conformation .
All nadI ( Rs ) mutations lack NMN transport function even though they were isolated only by their constitutive repression phenotype ; this suggests that the NadI protein can not stimulate transport when in the repressing conformation .
We suggest that the NadI protein shifts between R-T + and R + T-states in response to the intracellular NAD level .
This might be physiologically important for cells , since both the de novo synthesis of pyridines and NMN transport could be set in operation when the NAD level is low , and both could be turned off when the NAD level is high .
In the accompanying paper ( 34 ) , we present evidence that the transport function of nadI is not an integral part of the NMN transport system but rather serves to modulate ( in response to NAD ) the activity of a selfsufficient transport function provided by the pnuC gene product .
It is surprising that serB insertion mutations show polar effects on expression of nadI .
This observation leads to the hypothesis that serB transcription extends out of the serB gene into the nadI gene .
The Escherichia coli serB gene has been sequenced and is known to be transcribed clockwise ( 24 ) .
Just 48 bp distal to the stop codon of serB is a Shine-Dalgarno sequence and an open reading frame of at least 480 bp .
The possibility of a larger serB operon structure was proposed , since no strong mRNA termination site was apparent in the determined sequence ( 24 ) .
The physiological significance of this operon and its relationship to nadI is not understood .
Blazey , D. L. , and R. 0 .
Transcriptional activity of the transposable element TnlO in the Salmonella typhimurium ilvGEDA operon .
Bochner , B. R. , H.-C .
Huang , G. L. Schieven , and B. N. Ames .
Positive selection for loss of tetracycline resistance .
Brendel , V. , and E. N. Trifonov .
A computer algorithm for testing potential prokaryotic terminators .
Casadaban , M. J. , and S. N. Cohen .
Lactose genes fused to exogenous promoters in one step using a Mu-lac bacteriophage : in-vivo probe for transcriptional control sequences .
Castilho , B. A. , P. Olfson , and M. J. Casadaban .
Plasmid insertion mutagenesis and lac gene fusion with mini-Mu bacteriophage transposons .
Chan , R. K. , D. Botstein , T. Watanabe , and Y. Ogata .
Specialized transduction of tetracycline resistance by phage P22 in Salmonella typhimurium .
Properties of a high frequency transducing lysate .
Chumley , F. , R. Menzel , and J. R. Roth .
Hfr formation directed by TnlO .
Ciampi , S. M. , and J. R. Roth .
Polarity effects in the hisG gene of Salmonella require a site within the coding sequence .
Ciampi , S. M. , M. B. Schmid , and J. R. Roth .
Transposon TnlO provides a promoter for transcription of adjacent se ¬ quences .
Connor , C. , N. Zhu , and J. R. Roth .
Cookson , B. T. , B. M. Olivera , and J. R. Roth .
Genetic characterization and regulation of the nadB locus of Salmonella typhimurium .
Davis , R. W. , D. Botstein , and J. R. Roth .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 12 .
Demerec , M. , E. Adelberg , A. J. Clark , and P. E. Hartman .
A proposal for a uniform nomenclature in bacterial genetics .
Elliott , T. , and J. R. Roth .
Characterization of TnlOdCam : a transposition-defective TnlO specifying chloramphenicol resistance .
Foster , J. W. , E. A. Holley-Guthrie , and F. Warren .
Regulation of NAD metabolism in Salmonella typhimurium genetic analysis and cloning of the nadR repressor locus .
Foster , J. W. , D. M. Kinney , and A. G. Moat .
Pyridine nucleotide cycle of Salmonella typhimurium : isolation and characterization of pncA , pncB , and pncC mutations and utilization of exogenous nicotinamide adenine dinucleotide .
Foster , J. W. , Y. K. Park , T. Fenger , and M. P. Spector .
Regulation of NAD metabolism in Salmonella typhimurium : molecular sequence analysis of the bifunctional nadR regulator and the nadA-pnuC operon .
Goosen , N. , and P. van de Putte .
Role of Ner protein in bacteriophage Mu transposition .
Higgins , D. Personal communication .
Holley , E. A. , M. P. Spector , and J. W. Foster .
Regulation of NAD biosynthesis in Salmonella typhimurium : expression of nad-lac gene fusions and identification of a nad regulatory locus .
Hughes , K. T. , and J. R. Roth .
Conditional transpositiondefective derivative of Mu dl ( Amp Lac ) .
Hughes , K. T. , and J. R. Roth .
Transitory cis complementation : a method for providing transposition function to defective transposons .
Kinney , D. M. , J. W. Foster , and A. G. Moat .
Pyridine nucleotide cycle of Salmonella typhimurium : in-vitro demonstration of nicotinamide mononucleotide deamidase and characterization of pnuA mutants defective in nicotinamide mononucleotide transport .
Maloy , S. R. , and W. D. Nunn .
Selection of loss of tetracycline resistance by Escherichia coli K-12 chromosome .
Maloy , S. R. , and J. R. Roth .
Regulation of proline utilization in Salmonella typhimurium : characterization of put : : Mu d ( Ap , lac ) operon fusions .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 24 .
Neuwald , A. F. , and G. V. Stauffer .
DNA sequence and characterization of the Escherichia coli serB gene .
Priess , H. , B. Brauer , C. Schmidt , and D. Kamp .
Sequence of the left end of Mu , p. 277-296 .
In N. Symonds , A. Toussant , P. Van de Putte , and M. Howe ( ed .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 25a .
Roberts , G. Unpublished data .
Schmid , M. B. , and J. R. Roth .
Genetic methods for analysis and manipulation of inversion mutations in bacteria .
A method for detection of phage mutants with altered transducing ability .
Simons , R. W. , B. C. Hoopes , W. R. McClure , and N. Kleckner .
Three promoters near the termini of IS10 : pIN , pOUT , and pIII .
Spector , M. P. , J. M. Hill , E. A. Holley , and J. W. Foster .
Genetic characterization of pyridine nucleotide uptake mutants of Salmonella typhimurium .
Vogel , H. J. , and D. M. Bonner .
Acetyornithinase of Escherichia coli : partial purification and some properties .
Way , J. C. , M. A. Davis , D. Morisato , D. E. Roberts , and N. Kleckner .
New TnlO derivatives for transposon mutagenesis and for construction of lacZ operon fusions by transposition .
Zhu , N. , B. M. Olivera , and J. R. Roth .
Identification of a repressor gene involved in the regulation of NAD de novo biosynthesis in Salmonella typhimurium .
Zhu , N. , B. M. Ofivera , and J. R. Roth .
Genetic characterization of the pnuC gene , which encodes a component of the nicotinamide mononucleotide transport system in Salmonella typhimurium .
Zhu , N. , B. M. Olivera , and J. R. Roth .
Activity of the nicotinamide mononucleotide transport system is regulated in Salmonella typhimurium .