Permease (panF) Mutants nC - Europe PMC

1 downloads 0 Views 1MB Size Report
Apr 8, 1985 - grade. * Corresponding author. a-kefoisovaleric acid. ; pan 8 ketopantoic acid aspartic acid. |panD pantoic acid. ,B-alanine. nC extracellular.
Vol. 164, No. 1

JOURNAL OF BACTERIOLOGY, Oct. 1985, p. 136-142 0021-9193/85/100136-07$02.00/0 Copyright C 1985, American Society for Microbiology

Isolation and Characterization of Escherichia coli Pantothenate Permease (panF) Mutants Department

of Biochemistry,

DAVID S. VALLARI AND CHARLES 0. ROCK* St. Jude Children's Research Hospital, and The University of Tennessee Center for the Health Sciences, Memphis, Tennessee 38101 Received 8 April 1985/Accepted 15 July 1985

Mutants of Escherichia coli K-12 defective in the pantothenate permease (panF) were isolated and characterized. The panF mutation resulted in the complete loss of pantothenate uptake and of the ibility to use extracellular vitamin for growth. The growth phenotypes of panF panD, panF panB, and panF panC double mutants showed that the cytoplasmic membrane was impermeable to external pantothenate. Anaiysis of the intracellular and extracellular metabolites from strain DV1 (panF panD) labeled with P-[3- H]alanine demonstrated that a carrier-mediated mechanism for efficient paptothenate efflux remained in the panF mutant. Genetic mapping of this nonselectable allele was facilitated by the isolation of three independent TnlO insertions close to panF. Two- and three-factor crosses located panF at minute 72 of the E. coli chromosome and established the gene order fabE panF aroE.

a-kefoisovaleric acid ; pan 8 ketopantoic acid

The vitamin pantothenate is the precursor of the predominant acyl group carriers in living systems, coenzyme A (CoA) and acyl carrier protein (1). The phosphorylation of pantothenate is the initial step in the formation of CoA (6) (Fig. 1), and results of in vivo experiments with Escherichia coli have shown that the cellular CoA coptent is controlled at this step (12, 13). The activity of the enzyme pantothenate kinase is probably under negative feedback regulation by CoA or one of its thioesters (2, 11, 15, 21), and the pantothenate that is not phosphorylated by the kinase is excreted into the medium (12, 13). Approximately 15-fold more pantothenate is exported from the cell than is used for CoA biosynthesis (12), illustrating the physiological importance of pantothenate transport. E. coli has a membranebound permease specific for pantothenate that is coupled to the membrane electrochemical gradient by a sodium ion cotransport mechanism (23). In this study we tested the role of the pantothenate permease in the regulatory system that governs the intracellular CoA concentration by isolating and characterizing mutant strains lacking pantothenate uptake ability. MATERIALS AND METHODS Chemicals and supplies. Sources for supplies were as follows: ACS scintillation solution (Amersham Corp.); 250-,um silica gel H plates (Analabs, Inc.); 0.45-,m filters (type HA; Millipore Corp.); P-[3-3H]alanine (specific activity, 40 Ci/mmol), D-[1-_4C]pantothenic acid (specific activity, 55 Ci/mol), L-[U-14C]glutamine (specific activity, 263 Ci/mol), and L-[U-14C]proline (specific activity, 273 Ci/mol) (New England Nuclear Corp.); P-alanine, D-pantothenic acid, D-pantoyl lactone, ampicillin, chloramphenicol, chlortetracycline hydrochloride, naladixic acid, streptomycin sulfate, tetracycline hydrochloride, and ethylmethanesulfonic acid (Sigrma Chemnical Co.); and DE81 filter circles (Whatman Laboratory Products, Inc.). Sodium D-pantoate was prepared from D-pantoyl lactone by autoclaving with a 1.1:1 stoichiometric ratio of NaOH (17). All other chemicals were obtained through commercial sources and were reagent grade.

aspartic acid

|panD ,B-alanine

pantoic acid

nC pantothenic acid < =

extracellular

pantothenic acid

4'-phosphopantothenic acid

---- coenzyme A FIG. 1. Synthesis and metabolism of pantothenic acid in E. coli. Defects in the synthetic pathway are panB, ketopantoate hydroxymethyl transferase: panC, pantothenate synthetase; and panD, aspartate-1-decarboxylase. Dashed line represents proposed feedback regulation of pantothenate kinase by CoA.

Bacterial strains. All strains used in this study were derivatives of E. coli K-12 and are listed in Table 1. Strains defective in pantothenate uptake (panF) did not exhibit a growth phenotype in the absepce of a Pan- background (Fig. 1). panD strains were constructed by transduction with a P1 phage stock prepared from strain SJ16 and by selection for zad-220::Tnl O. To construct strain DV29 (panC zad220::TnJO), strain AT1371 (panC) was transduced with P1 phage prepared from strain SJ16, and Tetr recombinants were scored for retention of panC (4 of 83 recombinants). Selection for the loss of TnJO was performed by the method of Bochner et al. (4), as modified by Maloy and Nunn (18).

* Corresponding author. 136

CHARACTERIZATION OF panF MUTANTS

VOL. 164, 1985

137

TABLE 1. Bacterial strains used in this study

A132834 AE168 AT1371 DV1 DV4 DV5

DV6 DV8 DV9 DV11 DV12 DV14 DV21 DV24 DV26 DV29 DV32 DV33 DV34 DV36 DV39 DV41 DV42

DV44 DV46 Hfr C Hfr H KL96

kL209 (J4c) KL708

KL738 L8 (LA2-22d) NK5336

PK191 (PK19d) SJ15 SJ16 SJ100 UB1005 UQ285 W1485

Construction or Source

Genotype

Strain

aroE353 mal-352 tsx-352 A- Xr supE42 FargG his leu trp thy glpD lac- mal+ gal xyl mtl rpsL gyrA FpanC4 proA2 his4 argE3 thi-1 lacYl ga/K2 xyl-5 mtl-l tsx-29 X- supE44 metBI panD2 panFI zad-220::tnlO

met)3l panD2 panFI metBI panD2 metBI panD2 zha-6::TnlO metBI panD2 panFl zha-6::TnlO metBI panD2 zhc-9: :TnlO metBI panD2 panFI zhc-9::TnlO metBI panD2 zhc-12::TnlO metBI panD2 panFI zhc-12::TnlO argG75 lacZ4 panD2 rpoD285 zad-220::TnlO metBI panD2 zad-220::TnlO panF2 metBI panD2 zad-220::TnlO panF3 panC4 zad-220::TnlO aroE353 panD2 zad-220::TnlO metBI panB6 metBlpanD2 panF2

P1(SJ16) x UQ285 EMS mutagenesis of SJ16 EMS mutagenesis of SJ16 P1(SJ16) x AT1371 P1(SJi6) x AB2834 Tets of SJ15 Tets of DV24 Tets of DV26

metBI panD2 panF3

P1(DV29) x UB1005 Tet' of DV39 Tet5 of DV32 Tets of SJ100 P1(DV11) x DV42 J. Cronan J. Cronan J. Cronan

panC4 zad-220::TnlO metBi metBI panC4

aroE353 panD2 fabE22(Ts) panD2 aroE353 panD2 panFl Hfr Hfr Hfr

B. Bachmann (CGSCa) J. Anderson B. Bachmann (CGSC) EMSb mutagenesis of SJ16 Tets of DV1 Tets of SJ16 P1(W1485::TnlO) x DV4 P1(DV6) x DV4 P1(W1485::TnlO) x DV4 P1(DV9) x DV4 P1(W1485::TnlO) x DV4 P1(DV12) x DV4

zhc-9::TnlO

J. Cronan

Hfr

F1411/leuB6 tonA2 lacYl supE44 gal-6 X- hisGl recAl argG6 rpsL104 malAl (Xe) xyl-7 mtl-2 metBI F1401/leuB6 tonA2 lacYl supE44 gal-6 X- hisGI recAl argG6 rpsL104 malAl (Xe) xyl-7 mtl-2 metBI g/tA5 fabE22(Ts) Ict-l ara-14 lac YJ galK2 xyl-5 mtl-l rpsL20 tsx-57 tfr-5 supE44

UGA suppressor

Hfr metBI panB6 zad-220::TnlO relAl spoTI A- xr gyrA216 FmetBI panD2 zad-220::TnlO relAl spoTl X- A gyrA216 Fg/tAS fabE22(Ts) panD2 zad-220::TnlO metBI relAl spoTl X- Xr gyrA216 FlacZ4 rpoD285 argG75 XsupE42

B. Bachmann (CGSC) B. Bachmann (CGSC)

B. Bachmann (CGSC) J. Cronan J. Cronan S. Jackowski 12 S. Jackowski 5 B. Bachmann (CGSC) J. Cronan

CGSC, Coli Genetic Stock Center, Yale University, New Haven, Conn. b EMS, Ethylmethanesulfonate.

a

c

d

Ancestral strain. Previous designation.

Acetyl-CoA carboxylase mutants (fabE) were temperature sensitive (22) and were scored for growth at 42 versus 30°C. Media. Rich broth contained 10 g of tryptone per liter, 5 g of NaCl per liter, and 1 g of yeast extract per liter and rich medium agar was 1.5% agar. P1 top agar was rich medium containing 0.6% agar, 5 mM CaC12, and 10 mM MgC92. Lambda medium contained 10 g of tryptone per liter, 2.5 g of NaCl per liter, 1 mM MgS04, 0.2% maltose, and 0.01% yeast extract. Lambda medium agar was prepared with 1% agar, and A top agar was prepared with 0.65% agar. Luria-Bertani medium (19) contained 2.5 mM sodium pyrophosphate and tetracycline hydrochloride for the selection of TnlO insertions. Minimal medium E (24) was supplemented with thiamine (1 mg/liter), and plates contained 1.5% agar. The carbon source was glucose, glycerol, or mannitol (0.4%). Growth supplements were L-amino acids (100 mg/liter), casein hydrolysate (1 g/liter), and thymine (40 mg/liter). The supplement for aroE strains was 25 mg each of Lphenylalanine, L-tryptophan, and L-tyrosine per liter, and 10 jim shikimic acid (20). Rich broth and Bochner medium (4,

18) were supplemented with 20 ,uM P-alanine for panD panF strains and with aromatic amino acids plus shikimic acid for aroE mutants. The concentrations of antibiotics were 20 ,ug of ampicillin, 20 ,ug of naladixic acid, 100 ,ug of streptomycin sulfate, and 10 ,ug of (20 ,ug/ml in rich medium) tetracycline hydrochloride per ml. Isolation of panF strains. Strain SJ16 (panD) was grown overnight in minimal medium E containing glucose, methionine, and 1 ,uM ,B-alanine. Stationary-phase cells were exposed to 1.5% ethylmethanesulfonate for 1 h with shaking at 37°C and washed, and survivors were grown overnight in the same medium. The overnight culture was washed with medium E and suspended to a density of 5 x 107 cells per ml in fresh medium containing glucose, methionine, and 1 jiM pantothenate. Cells unable to grow on 1 jiM pantothenate were enriched with ampicillin (19) and recovered in medium containing 1 ,uM ,B-alanine. After two cycles of ampicillin selection, cells were plated on minimal medium containing 1 ,uM ,B-alanine and screened for the ability to utilize 1 jiM pantothenate. panF colonies were isolated with a 35%

138

VALLARI AND ROCK

frequency after approximately 106-fold enrichment. This protocol was followed on three occasions to yield independent isolates. The growth phenotypes and metabolism of ,-alanine exhibited by the panF2 and panF3 isolates were identical to those described for strain DV1 (panFI). Isolation of Tn1O insertions near panF. A stock of phage XNK370 (b221 cI857 cI171::TnJO 0261; from J. Cronan) grown on strain NK5336 was used to generate a random TnWO insertion pool in strain W1485 (7). Approximately 1,500 independent insertion mutants were pooled and grown in a mixed culture. A P1 phage stock prepared from this pool was crossed with strain DV4 (panD panFI), and recombinants were selected for growth on minimal medium containing glucose, methionine, tetracycline, and 1 p.M pantothenate. Colonies were screened for both the panD and panF alleles, and three panD panF+ colonies were isolated. P1 phage were grown on each isolate (strains DV6, DV9, DV12), and the linkage of each TnWO insertion to panFI was confirmed by transductions into strain DV4. Map position of panF. P1 phage grown on strain DV12 (zhc-12::TnJO) were used to transduce HfrC, HfrH, KL96, KL209, and PK191 to Tetr. The panF allele was localized to a region of the chromosome by mating these Hfr and TnWO derivatives with strain AE168 (19), selecting recombinants for Tetr on rich medium agar and scoring for the absence of a growth requirement. Naladixic acid or streptomycin was used to counterselect against the Hfr strains. Partial diploids constructed from mating either strain KL738 or strain KL708 with DV46 were tested for expression of the panF+ growth phenotype. Strains harboring mutations with established map positions in the 70- to 74-min region of the E. coli chromosome (Table 1) were transduced with P1 phage stocks grown on strains DV6, DV9, DV12, and SJ16 (unlinked control) to determine the linkage of each TnWO insertion to the markers in this region. Cotransduction frequencies were determined by selecting TnlO and scoring the established marker, and from the reciprocal cross. Recombinants were scored directly for panF in panD derivatives of these strains (Table 1). Three-factor crosses were performed to define the orientation of panF with each of the three transposon insertions and with the fabE and aroE loci. Selections in these crosses were for either TnWO or the established marker in a panD background, and panF and the other unselected marker were scored. Growth studies. All growth studies were done at 37°C. Overnight cultures of strains DV1 (panF) and SJ16 (panF+) grown in minimal medium E containing glucose, methionine, and 10 ,uM P-alanine were washed and subcultured in medium minus ,-alanine to deplete the CoA pool (12). CoA-depleted cells were suspended to 5 x 107 cells per ml in fresh growth medium containing the indicated concentration of P-alanine or pantothenate, and cell number was tnonitored with a Klett-Summerson colorimeter with a blue filter. The colorimeter was calibrated by determining the number of CFU/ml in the range of colorimeter readings encountered. Pantothenate transport assays. Transport assays were performed as described previously (23) with 10 p.M D-[14C]pantothenate (specific activity, 55 Ci/mol), L-[U4C]proline, and L-[U-14C]glutamine and 109 cells per ml. Pantothenate kinase assays. Cells were washed twice with cold lysis buffer (50 mM Tris-hydrochloride, 10 mM NaCl, 0.5 mM MgCl2 [pH 7.4]) and broken in a French pressure cell at 18,000 lb/ih2, and the supernatant from a 30-min spin at 50,000 x g was dialyzed against the lysis buffer at 4°C. Assays contained ATP (2 mM); MgCl2 (2 mM); D-[1-

J. BACTERIOL.

"Clpantothenate (90 ,uM; specific activity, 60,000 dpmlnmol); and Tris-hydrochloride (0.1 M; pH 7.5). The reaction was initiated by the addition of 100 p,g of protein,.

and the total volume of the incubation mixture was 40 p,. After 10 min at 250C, a 35-,ul fraction was deposited on a Whatman DE81 ion-exchange filter disk that was washed with two changes of 1% acetic acid in 95% ethanol (20 ml per disk) to remove unphosphorylated pantothenate. 4'Phosphopantothenate was quantitated by counting the dried disk in 3 mnl of scintillation solution. Protein was determined with the BioRad protein assay kit and y-globulin as a standard. Analysis of 13-[3-3H]alanine metabolites. Strains DV1 (panF) and SJ16 (panF+) were depleted of CoA (12) and subcultured in medium containing 1-[3-3H]alanine. Stationary-phase cells were separated from the medium py centrifuging for 3 min at maximum speed in a Beckman microfuge. The cell pellets were washed twice with medium E and stored separately from the media samples at -20°C. ,-[33H]alanine metabolites were extracted from the cells as described previously (12). Both the cell extracts and the media samples were treated with 1 mM dithiothreitol to convert thioesters and disulfides to free sulfhydryl forms and analyzed by thin-layer chromatography on silica gel H plates developed with either ethanol-28% ammonium hydroxide (4:1 [vol/vol]) or butanol-acetic acid-water (5:2:4 [vol/vol]) to 14 cm from the origin (12). The distribution of radioactivity was quantitated by scraping 0.5-cm sections of the silica gel into scintillation vials, deactivating the silicic acid with 0.1 ml of water, and counting in 3 ml of scintillation cocktail. RESULTS Isolation of pantothenate permease mutants (panF). Transport mutants were isolated by their inability to use 1 p.M

3.0

B _~~~~.

2.0

,0

loloOM

500jM

-I;E-.) aW0Q

/f/20juM A

/

,;,R/#~~~5M j

2

6

10

2

6

10

(hours) FIG. 2. Growth phenotypes of strain DV1 (panFl panD) and strain SJ16 (panD). (A) Growth of strain SJ16 (0, A) and DV1 (0, A) on glucose minimal medium supplemented with either 1 ,uM pantothenate (A, A) or 1 ,uM P-alanine (0, 0). (B) Growth of strain DV1 in glucose minimal medium supplemented with increasing concentrations of pantothenate.

VOL. 164, 1985

CHARACTERIZATION OF panF MUTANTS

1500

pantothenate for growth after mutagenesis of strain SJ16 (panD). The panD mutation allowed counterselection of pantothenate metabolism mutants in the presence of the pantothenate precursor ,-alanine. The intracellular CoA pool was depressed to 10% the normal value by using a low (1 ,M) concentration of ,-alanine in the nonselective medium (12), thus preventing residual growth of panF mutants during ampicillin selection. The growth of strain DV1 (panFl) resembled that of the wild-type strain SJ16 in minimal medium supplemented with ,-alanine, but strain DV1 was defective in pantothenate utilization (Fig. 2). CoA-depleted cells of strain DV1 were unable to grow in minimal medium containing 1 ,uM pantothenate, whereas strain SJ16 grew with a doubling time of less than 1 h (Fig. 2A). Both strains grew with a doubling time of approximately 1 h in minimal medium containing 1 p.M P-alanine (Fig. 2A). Strain DV1 grew slowly in minimal medium containing larger supplements of commercial pantothenate (Fig. 2B) and had a doubling time of 1.1 h on 100 ,uM pantothenate. Growth on 1 p.M pantothenate was restored in

0 E

1200

Q -e 0a

900 Ca_ 0

a) 0.

0 a-

t 90

2

120

Seconds FIG. 3. Pantothenate uptake in strain DV1 (panFl panD) and strain SJ16 (panD).

merodiploids containing either the F141

or

139

F140 episome.

Biochemical characterization of strain DV1 (panFl).

TABLE 2. Three-factor analyses Cross (P1 donor x Recipient)

Selection (no.)

DV8 (panFl zha-6::TnlO) x DV44 (fabE)

TnlO (200)

DV11 (panF zhc-9::TnlO) x DV42 (aroE)

TnlO (236)

aroE+ (300)

classes Recombination (no.)

fabE+ panF+ (6) fabE+ panF (7) fabE panF (187) fabE panF (0) aroE+ panF+ (8)

Implied gene order

zha-6::TnlOfabE panF

zhc-9::TnlO panF aroE

aroE+ panF (68) aroE panF (58) aroE panF (102) Tetr panF (1) Tetr panF (32)

Tets panF (265) DV11 (panFl zhc-9::Tnl O) x DV44 (fabE)

TnJO (182) fabE+ (247)

DV14 (panF zhc-12::TnlO) x DV42 (aroE)

TnOO (50) aroE+ (352)

DV14 (panF zhc-12::Tnl O) x DV44 (fabE)

TnOO (237) fabE+ (232)

DV46 (panF aroE) x SJ100 (fabE)

fabE+ (190)

Tets panF (2) fabE+ panF (41) fabE+ panF (129) fabE panF (11) fabE panF (1) TetrpanF+ (33) Tet' panF (204) Tet' panF (3) Tets panF (7) aroE+ panF+ (8) aroE+ panF (18) aroE panF+ (5) aroE panF (19) TetrpanF (18) Tetr panF (24)

Tets panF (308) Tets panF (2) fabE+ panF+ (22) fabE+ panF (141) fabE panF (61) fabE panF (13) Tetr panF+ (2) Tetr panF (53) Tets panF+ (95) Tets panF (82) 22 aroE+ panF (22) 119 aroE+ panF (119) 1 aroE panF (1) 48 aroE panF (48)

zhc-9::TnlO fabE panF

panF zhc-12::TnlO aroE

fabE panF zhc-12::TnlO

fabE panF aroE

140

VALLARI AND ROCK

J. BACTERIOL.

74 rpsL

73 72

-

c,vroE zhc -12:: Tn /0 / -~pcnF

=¢ fabE

/O

zhc- 9::

Tn

70-

zho-6::

Tn /1

69

argG

71

-

FIG. 4. Location of panF, TnlO insertions, and established markers in the 69- to 74-min region of the E. coli chromosome. The relative map distances were computed from cotransduction frequencies obtained from two- and three-factor crosses using the formula of Wu (25) and the average of frequencies from reciprocal crosses.

Pantothenate uptake by an exponentially growing culture of strain DV1 was compared with that of strain SJ16. No pantothenate uptake activity by strain DV1 was detected with 10 ,uM pantothenate, whereas the pantothenate permease of strain SJ16 exhibited maximum activity at this concentration (23) (Fig. 3). We were also unable to detect uptake at higher D-[1-14C]pantothenate concentrations (up to 100 ,M). There was no difference in L-[U-14C]proline and L-[U-14C]glutamine uptake activities in the mutant and wildtype strains (data not shown), indicating the absence of a pleiotropic transport defect. Pantothenate kinase activities in extracts from strains DV1 and SJ16 were identical (0.11

nmol/min per mg of protein). Isolation of TnlO insertions near panF. TnJO insertions near panF were isolated to facilitate mapping of this nonselectable allele (16). The zad-220::TnJO insertion was removed from strain DV1 (panFl) with Bochner selection medium (4, 18), and transduction of the resulting strain (strain DV4) with P1 phage grown on strain SJ16 (zad220::TnJO) demonstrated that panF was not linked to the panBCD gene cluster. To isolate TnlO insertions near panF, TnWO was inserted randomly into strain W1485 with lambda phage NK370, and a P1 stock grown on a mixed culture of approximately 1,500 independent insertion mutants was used to transduce strain DV4. Of the recombinants that grew on minimal medium containing 1 ,uM ,-alanine and tetracycline, 2.4% (14 of 584) were able to grow on minimal medium containing pantothenate. The panD allele was retained in 3 of 14 isolates, whereas the remainder were Pan', resulting from TnWO insertions near the panD locus. The zhc-12: :TnJO element in strain DV12 had an average cotransduction frequency of 90% with panFl. The independent panF2 and panF3 isolates exhibited the same high cotransduction frequency with this insertion. The zhc-9::TnJO and zha-6::TnlO

insertions of strains DV9 and DV6 cotransduced with panFI at a rate of 66 and 2%, respectively. Map position of TnlO insertions and panF. P1 phage grown on strain DV12 (zhc-12::TnJO) were used to transduce several Hfr strains to Tetr (Table 1). The mating of Hfr KL209/zhc-12::TnJO with strain AE168 and selection for TnWO resulted in 93% linkage with rpsL, 54% with argG, 20% with thyA, and 34% with mtl. Selection for Arg+ yielded 100% Tetr recombinants. From the cross PK191/zhc12::TnJO x AE168, selection for TnWO yielded 67% linkage with argG and 50% linkage with thyA when streptomycin was used to counterselect against PK191/zhc-12: :TnJO. Phage P1-mediated transductions determined the linkage between the TnlO insertions of strains DV6, DV9, and DV12 and the chromosomal markers, argG,fabE, aroE, and rpsL. The zha-6::TnJO insertion (2% linkage with panF) cotransduced 17% with argG and 15% withfabE, but did not cotransduce with aroE or rpsL (>100 recombinants were scored). The zhc-9::TnJO insertion (66% linkage with panF) averaged a cotransduction frequency of 95% withfabE, 16% with aroE, and 1% or less with argG and rpsL. The zhc12::TnJO insertion (90% linkage with panF) was 79% cotransducible with fabE, averaged 33% with aroE and 2% with rpsL, and did not cotransduce with argG (200 recombinants were scored). These data suggest the placement of the panF locus between fabE and aroE. Transductions with panD derivatives of argG, fabE, and aroE strains indicated that panF was closest to fabE, demonstrating 75% cotransduction. Some two-factor crosses yielded different results, depending on the direction of the cross or the donor allele. The zhc-12::TnJO insertion and aroE cotransduced 43% when selection was for Tetr, but only 13% in the reciprocal cross. When aroE+ was selected, cotransduction withfabE was 54% when the donor wasfabE+, but was 13% when the donor was fabE, which is consistent with the poor transduction of fabE noted by other investigators (22). Similarly, selection for fabE+ yielded 42% cotransduction with aroE when the donor was aroE+ and 13% when the donor was aroE. A cotransduction frequency of 30% for aroE with fabE has been reported previously (22). Three-factor crosses unambiguously determined the orientation of the panF gene with each of the TnWO insertions and thefabE and aroE loci. The gene order zha-6: :TnJOfabE panF was established by transductions with phage P1 grown on strain DV8 (Table 2). Transductions with strain DV11 as the donor established the orders zhc-9: :TnJO panF aroE and zhc-9::TnJO fabE panF (Table 2). The orders panF zhc12: :TnJO aroE and fabE panF zhc-12: :TnJO were determined with phage P1 grown on strain DV14 (Table 2). The data

TABLE 3. Synthesis of CoA and pantothenate from

P-[3-3H]alanine in strains DV1 (panFl panD) and SJ16 (panD) Extracellular Intracellular CoA f3-Alanine pantothenate ,1o-Alanine Intracellsla: (pmoW101 (mll ina: ces)i-:cells)

supplement

(1±M)

DV1

SJ16

DV1

SJ16

1 8 16

7.6 49 83

10 91 85

3.5 84 286

1.0 35

236

Both intracellular CoA and extracellular pantothenate were quantitated by thin-layer chromatography as described in the text from stationary-phase cultures of strains DV1 and SJ16 grown on glucose minimal medium and the indicated concentration of ,-[3-3H]alanine. a

CHARACTERIZATION OF panF MUTANTS

VOL. 164, 1985

from these five three-factor crosses with the TnOO insertions were consistent with the clockwise gene order fabE panF aroE. We confirmed this arrangement by crossing strain DV46 with strain SJ100 (Table 2). The positions of panF and the three TnOO insertions on the current E. coli genetic map (3) are summarized in Fig. 4. P-Alanine metabolism by strain DV1. P-Alanine-derived metabolites were uniformly labeled in CoA-depleted cultures, and cell pellets and media samples from stationaryphase cultures were analyzed by thin-layer chromatography (12, 13). CoA was the predominant intracellular metabolite (Table 3). Strains DV1 and SJ16 had a similar CoA content when labeled with 16 ,uM or higher concentrations of 3alanine (Table 3). However, the strains differed in the amount of P-alanine supplement required to saturate the CoA pool. This difference was greatest with an 8 ,uM P-alanine supplement which yielded the maximum CoA concentration in strain SJ16 (12) but only 60% of the maximum concentration in strain DV1 (Table 3). The intracellular label that was incorporated into the 4'-phosphopantetheine prosthetic group of acyl carrier protein comprised 10 to 15% of the total and fluctuated with the CoA concentration (data not shown). Trace amounts of intracellular 4'-phosphopantetheine and pantothenate were found in both strains (data not shown). [3-3H]pantothenate was the predominant labeled metabolite found in the media of stationary-phase cultures of both strains DV1 and SJ16 (Table 3), illustrating that the panF defect did not block pantothenate efflux. The media of strain DV1 contained as much as a 3.5-fold higher concentration of pantothenate compared with media of strain SJ16, even when intracellular CoA was depressed (Table 3). Less 4'phosphopantetheine was present in the media of strain DV1 than the media of strain SJ16, and P-[3-3H]alanine was not present in either case (data not shown). When samples were removed and analyzed throughout the exponential growth phase and into the early stationary growth phase, pantothenate was again found in the media of strain DV1 in higher concentrations than in the media of strain SJ16 (data not shown), suggesting that the panF mutation increased the net rate of pantothenate efflux by specifically blocking pantothenate uptake during the later stages of exponential growth. Growth phenotypes of panF panB, panF panC, and panF panD double mutants. The zhc-12::TnJO insertion was used to place panF into panB, panC, and panD (Fig. 1) derivatives of strain UB1005 and strain AT1371 (panC) to test for the presence of a low-affinity pantothenate transport system. Transduction of strain DV5 (panD) with P1 phage grown on strain DV14 (panFl zhc-12::TnJO) resulted in approximately 80% cotransduction between zhc-12::TnJO and panF whether the selection medium contained ,-alanine (1 FM) or pantothenate (100 ,uM) (Table 4). A similar linkage between panF and the TnOO insertion was obtained by crossing strain DV33 (panB) with the same phage lysate when the selection medium contained pantoate (100 ,uM); however, no panF panB colonies were obtained when the medium contained 100 puM pantothenate (Table 4). Likewise, panF panC colonies were not obtained from transductions of strains DV41 and AT1371 (Table 4) and by selection on 100 ,uM pantothenate. Growth of panD panF strains on 100 ,uM pantothenate was attributed to their ability to efficiently use ,B-alanine produced by the breakdown of pantothenate. The susceptibility of pantothenate to hydrolysis is well known (10), and panD strains are capable of growing to stationary phase on as little as 0.5 FM P-alanine (9), whereas this concentration

141

TABLE 4. Growth phenotypes of panF panB, panF panC, and panF panD double mutants Selection supplement Transduction Recipienta

DV5 (panD) DV33 (panB) DV41 (panC) AT1371 (panC)

(concn [uM])"

frequency

P-Alanine (1) Pantothenate (100) Pantoate (100) Pantothenate (100) Pantothenate (100) Pantothenate (100)

0.80 (40 of 50) 0.82 (82 of 100) 0.79 (57 of 72)