Gene, 109 (1991) 21-30 © 1991 Elsevier Science Publishers B.V. All rights reserved. 0378-1119/91/$03.50

21

GENE 06171

Null mutation in the stringent starvation protein of Escherichia coli disrupts lytic development of bacteriophage P1 (Stringent response; linear DNA transformation; P 1 vegetative growth defect; pog locus; P 1 late gene expression; stable RNA synthesis)

Mark D. Williams,.b, James A. Fuchs" and Michael C. Flickinger,.b a Department of Biochemistry and b Institute for Advanced Studies in Biological Process Technology. University of Minnesota. Saint Paul. MN 55108 (U.S.A.) Received by C.R. Hutchinson: I 1 June 1991 Revised/Accepted: 29 July/6 September 1991 Received at publishers: 17 September 1991

SUMMARY

As initial steps toward understanding the regulation and function of the stringent starvation protein (SSP) of Escherichia coli, we have isolated the ssp gene (encoding SSP), defined the operon in which ssp is found, and created insertion-deletion mutations of the ssp gene in recBC, sbc and recD strains by linear DNA transformation. During attempts to move the insertion-deletion structure to other strains by P 1 transduction, we found that P 1 was unable to form plaques on hosts lacking an intact ssp gene. The Assp nmtation, however, did not affect transduction of the 3ssp strains and mutant strains were able to support lysogenic P 1. When P 1 lytic growth was induced, an increase in P 1 DNA was detected without lysis or plaque formation. Examination of proteins synthesized in the Assp host during induction revealed the absence of PI late gene products. Also, the apparent continued synthesis of early gene products during late time points was observed in the Assp host. The results reported here suggest that the defect in P1 lyric growth brought about by the absence of SSP occurs at the point at which bacteriophage P 1 shifts from early to late gene expression. We also report the results of experiments on stable RNA synthesis following amino acid (aa) starvation induced by serine hydroxamate, and experiments on stable RNA synthesis following resupplementation of a limiting aa. These experiments show that SSP is not involved in stable RNA synthesis. Additionally, complementation studies have shown that ssp is identical to the previously described pog gene of E. coil.

INTRODUCTION

'The coordination of overall regulation of bacterial gene expression is thought to involve the interplay of pleiotropic regulatory domains' (Tomkins, 1975), known as global Correspondenceto: Dr. M.C. Flickinger, Bioprocess Institute, University of Minnesota, 240 Gortner Laboratory, 1479 Gortner Ave., Saint Paul, MN 55108 (U.S.A.) Tel. (612)624-9706; Fax (612)625-1700. Abbreviations: aa, amino acid(s); bp, base pair(s); CHEF, contourclamped homogeneous electric field; Cm, chloramphenicol; A, deletion; kb, kilobase(s) or 1000 bp; LB, Luria-Bertani (medium); MOPS, 2-(N-

regulatory systems or regulons. These include gene regulation in response to carbon, nitrogen, and phosphate source limitation, DNA damage, heat shock, and anaerobiosis, as well as aa starvation. Many of the global regulatory systems are known to involve the alteration of gene morpholino)propane-sulfonic acid); MS, magic spot (Cashel and Gallant, 1969); Nm, neomycin; nt, nucleotide(s); PAGE, polyacrylamide-gel electrophoresis; pfu, plaque-forming unit(s); pl, isoelectric point; R resistance/resistant; RNAP, RNA polymerase; SDS, sodium dodecyl sulfate; SHMT, serine hydroxamate; SSP, stringent starvation protein; sspA or peg, gene encoding SSP; sspB, second gene in the sspAB operon; Tc, tetracycline; wt, wild type; [ ], denotes plasmid-carrier state; : :, novel joint (insertion, fusion).

22 expression at the transcriptional level. This transcriptional regulation involves the interaction of alternative sigma factors or accessory proteins with RNAP in such a way as to alter the enzyme's affinity for regulatory DNA regions. The stringent response, the bacterial response to limitation of any required aa, was the first of these global regulatory systems to be discovered. It was first noticed as the ability of bacterial cells to curtail RNA accumulation in response to a limitation of a required aa (Sands and Roberts, 1952). Numerous subsequent studies have led to the understanding that the stringent response is characterized by the production of guanosine 3'-diphosphate-5'diphosphate (MS I; ppGpp), derived from guanosine 3'diphosphate-5'-triphosphate (MS II; pppGpp) whose synthesis is catalyzed by the relA gene product, on the ribosome in an idling reaction of protein elongation (Cashel and Gallant, 1968; 1969; Haseltine et al., 1972). 'This transcriptional regulation redirects energy and resources from the unproductive synthesis of additional translational components toward the synthesis of enzymes required to overcome the starvation" (Riggs et al., 1986). A common feature of the global regulatory systems, except for the stringent response, has been the identification of proteins involved in the alteration of RNAP specificity. These proteins may function as repressors, inducers, alternative sigma-factors, or DNA promoter-enhancing elements, often acting in concert with a regulatory nt (Hoopes and McClure, 1987). It has been suggested that ppGpp somehow interacts with RNAP directly to alter the affinity of the enzyme for promoters in an operon specific fashion (Reiness et al., 1975). Although proteins have been implicated in most other global regulatory systems, no protein has been shown to be involved in the interaction between

the regulatory nt, ppGpp, and RNAP. If ppGpp interacts directly with RNAP resulting in the alteration of transcription observed during the stringent response, this would be a unique global regulatory mechanism of transcriptional alteration. Therefore, it is important for understanding the mechanism of stringent regulation that the function of the 'stringent starvation protein' be investigated. Initially discovered during examination of protein synthesis following induction of the stringent response, SSP was reported as the protein whose synthesis was most dramatically stimulated by such response (Reeh et al., 1976). Later, SSP was isolated as an RNAP-binding protein by Ishihama and Saitoh (1979). The gene encoding SSP (Fukuda et al., 1985) and its sequence (Serizawa and Fukuda, 1987) have been described. The induction pattern and the cellular localization of SSP has led us to question whether SSP is a mediator of the interaction between ppGpp and RNAP. Fukuda et al. (!988) have reported that the SSP is dispensable for normal growth. Although the work reported here supports their conclusion that the SSP is not required for normal growth, these earlier investigators used a strain in which the ssp gene was under the control of the lactose promoter. Moreover, they did not construct an ssp null mutant. We describe the cloning of the ssp gene, definition of the operon in which the ssp gene is found, and the creation of ssp null mutants. We also present evidence that these mutants are unable to support P l vegetative growth. More specifically, we show that, when P 1 lytic growth is induced in the mutant harboring Plcmcl.ioo (Rosner, 1972), increased levels of Pl DNA are observed but no viable Pl particles are formed. Furthermore, analysis of proteins synthesized during P 1 induction clearly shows that, in ssp

TABLE I Escherichia coli K-12 strains used in this study

Strain

Partial genotype

Source

Reference

LE392 (ED8654) C600 DH5~e

F-hsdR514 supE44 supF58 F-thi-I tbr-1 ieuB6 iacYl tonA21 supE44 mcrA F - ( q~8OdlacZAM l 5 ) A(lacZ YA.argF)U i 69 recA i endA 1 hsdRl7 (r~, m~ )supE44 ~- thi-i gyrA96 relA 1 F-recB21 recC22 sbsB 15 sbcC201 leu arg his pro ara JC7623, F - sspA :: Nm R MG1655, F - A- recD::mini TnlO CAG12135, sspA::Nm R F - thr- ! [eu86 his.65 thi- 1 argH ara- ! 3 gal-3 malA ! xyl.7 intl.2 tonA2 supE447 (AR) CP78, F - sspA ::NmR F-thr-! leu86 his-65 thi-! argH ara-13 gal-3 malAi xyl-7 mtl-2 tonA2 supE447 relA (~.a) CP79, F-sspA::Nm n his-I leu-2 his-I leu-2 pog- 1

J.A. Fuchs J. Zissler J.A. Fuchs

Murray et al. (1977) Appleyard (1954) Hanahan (1983)

B. Kren This study D. Gentry This study B. Bachmann

Oishi and Cosby (1972)

JC7623 MW7 DPB267 EABI CP78 MW8203 CP79 MW9111 ES~364 ES1365

Biek and Cohn (1986) Fiil and Friesen (1968)

This study B. Bachmann

Fiil and Friesen (1968)

B. Bachmann E.C. Siegel E£. Siegel

Fill and Friesen (1968) Race (1984) Race (1984)

23 null mutants, early P 1 protein synthesis continues beyond the point at which the wt strain has begun to produce late P1 proteins and that apparently no late proteins are synthesized in the ssp null mutants. These results suggest that the defect in the P 1 life cycle created by the deletion of the ssp gene occurs at the point at which bacteriophage P 1 shifts from early gene expression to late gene expression. Our initial aim was to study stable RNA synthesis utilizing both stringent (relA + ) and relaxed (relA) ssp null mutants to determine what role, if any, SSP plays in this fundamental characteristic of the stringent response. Additionally, we wanted to investigate whether ssp is identical to a previously described gene, pog (Race, 1984).

ATGGATI-I'GTCACAGCTAACACCACGTCGTCCCTATCTGCTGCGTGCATfCTATGA GTGGTI'GCTGGATAACCAGCTCACGCCGCACCTGGTGGTGGATGTGACGCTCCC TGGCGTGCAGGTI'CCTATGGAATATGCGCGTGACGGGCAAATCGTACTCAACATIGCGCCGCGTGCTGTCC-GCAATCTGGAACTGGCG AATGATGAGGTGCGCTTI'AAC GCGCGC'ITT,GGTGGCAI-ICCGCGTCAGGI-FrCTGTGCCGCTGGCTGCCGTGCTG GCTATCTACGCCCGTGAAAATGGCGCAGGCACGATGTTIGAGCCTGAAGCTGCC TACGATGAAGATACCAGCATCA'rGAATGATGAAGAGGCATCGGCAGACAACGAA ACCGTrATGTCGGTI'ATI'GATGGCGACAAGCCAGATCACGATGATG ACACTCATC CTGACGATGAACCTCCGCAGCCACCACGCGGTGGTCGACCGGCA'I-rACGCG'I-I-G TGAAGTAATACAAAACAGG(~(~CAGGCGGCCTG'ITR'(~TCTTITIFig. 2. Sequenceof the sspBgene.The translational start and stop codons are double-underlined. Regions of dyad symmetry which can form the putative Rho-independent transcriptional terminator are underlined. The first nt shown corresponds to nt i 199 of Serizawa and Fukuda (1987). Regions of the 5.0-kb Hindlll fragment (Fig. 1) were sequenced (Sanger et al., 1977)with Sequenase(Tabor and Richardson, 1987)and [35S]ATP (Amersham Corp.). This sequence has been assigned the GenBank accession No. M69028.

RESULTS AND DISCUSSION

(a) Isolation of the ssp gene

The ssp gene was isolated by taking advantage of the previously published sequence and restriction enzyme recognition site data (Fukuda et al., 1985; Serizawa and Fukuda, 1987), which showed that the ssp gene is found on a 1.75-kb SalI DNA fragment. A plasmid library was created and screened for presence of the ssp gene resulting in identification of a SalI ssp containing fragment (Fig. 1). Approximately i00 transformants from this library were screened by restriction enzyme analysis and the resulting DNA restriction fragment patterns were compared to the expected pattern for the ssp containing fragment (Fukuda et al., 1985). A 1.75-kb Sail fragment was sequenced to confirm the previously published sequence of the ssp gene and to extend the sequence of adjacent regions (Serizawa and Fukuda, 1987) (Fig. 2). Unfortunately, the 1.75-kb fragment did not contain the

~.12F1

entire coding region of the gene immediately downstream from the ssp gene whose existence was noted by Serizawa and Fukuda (1987). Since the ribosome-binding site of this second gene is within the C-terminal coding region ofthe ssp gene, we determined the structure of the ssp-containing operon. This was accomplished by isolation of a 5.0-kb HindIIl fragment from 1-12F1 of the phage i library constructed by Kohara et al. (1987). This fragment contains the ssp gene and approx. 3.3 kb of downstream region (Fig. 1). Isolation of the ssp gene from I- 12F 1 establishes the location of the ssp gene on the physical map of the E. coli chromosome at 3450 kb (Kohara et al., 1987) in the vicinity of and counter-clockwise from the mdh locus (Bachmann, 1990). Analysis of this fragment allowed determination of the entire nt sequence of the second gene (Fig. 2) and the location of a putative Rho-independent transcriptional ter-

Hindlll Sail

Hindlll Sail _

A c

G

G

c C-G

C-G

Sail

ssp

Ball I

Smal I

G-C

Sail ,

~,BssHll ~

A-U C-G

Styli

A-U

BssHll

A-U A-U A-U

-GGC-CGG-AGU-UAA-UCUGU-AUG-GAU -Gly -Arg - Ser

•

fMet- Asp-

C-G

GU-GUA-AGU-UAA-UA-UCUUUUU

"

Fig. !. Location and structure of the operon which contains the sspA gene. The sspA gene was isolated from it-12Fl of the it collection of Kohara et al. (1987). Phage it DNA was purified as described (Maniatis et aL, 1982). The sspAB operon is located on a 5.0-kb HindIII fragment which contains a 1.75-kb Sail fragment on which the entire sspA gene is found. Sequence data (Serizawa and Fukuda, 1987) have been extended beyond the end of the sspB gene. A stem-and-loop structure immediately downstream from sspB, resembling a Rho-independent terminator, has been found. The sspB gene starts 5 bp downstream from the sspA stop codon and a ribosome-binding, site sequence of GGAG is found within the sspA coding region. Enzymes used in DNA manipulations were purchased from Bethesda Research Laboratories or New England Biolabs.

24 minator (Fig. 1). Sequence analysis ofsspB suggests that it encodes a 165-aa protein with a deduced Mr of 18 263 and calculated pl of 4.1. The function of sspB is currently being investigated.

(~) Wild-type genomlc swp region HIndlll

I

Sall

Hlncn

8rnal ,'~'5 kb Sail Insert> all all

Hlncll digestion

Ball + Smsl digestion

t

I F T 4 DNA llgase Sall

~/sspA:: NmIt

"1 ApR

Nmll

Sail

Fig. 3. Construction of pSNEOSP as the source of the sspA : : NmR DNA fragment for use in linear DNA transformation. Plasmid pSV77 was digested with Ball + Sinai which removed essentiallyall of the sspA coding sequence, The NmR gene contained in a 1.6-kb Hincll fragment from pUC4K was inserted in place ofthe Ball-Smal fragmenton pSV77. Plasmid pSV77 is a derivative of pUCI9 in which the Sinai site in the multicloning region has been removed.

Sail Hindlll

I

I

S~al Ball '

I

(b) Construction of AsspA mutants As an initial step toward elucidating the function of SSP, a site-directed insertion and deletion mutation in the sspA gene was created by linear D N A transformation of a recBC, sbcB strain, JC7623, according to the method of Winans et al. (1985), Plasmid pSV77, a pUC 19 derivative (YanischPerron et al., 1985) with the SmaI site in the multicloning region deleted and containing the 1.75-kb Sail fragment (Fig. 3), was digested with Ball + SmaI to remove essentially all of the sspA coding sequence. The Nm x gene, obtained on a 1600-bp Hincll fragment from pUC4K (Taylor and Rose, 1988), was inserted in place of the sspA coding region (Fig. 3). The resulting plasmid, pSNEOSP, containing the sspA : : Nm R structure, was used to replace the genomic sspA allele in JC7623. Plasmid p S N E O S P was digested with Sail + Scal, and the 2.75-kb Sail fragment (1.75-0.6 + 1.6) was purified (Langridge et al., 1980) and used to transform JC7623 to Nm R. Purification of the 2.75-kb, ssp.Nm R fragment eliminated the possibility that a linear plasmid would recircularize within the cell resulting in plasmid-based Nm R. Southern-blot analyses were carried out on genomic

Sail I---1.75 kb

~"5kb

I

:

I

Mutant genomlc ssp region

HindlH

Sall

I

I

HindHI I

I

®t

4.1 kb

i

2

3

4

Sall HindIll

I

Nm R 2.7kb I

11 1

I

I

t 1.gkb

I

I

2

3

4

kb ,

w 4.15 2.7 1.9 1.75

e gllJ

w

Fig. 4. Maps ofthe ssp region and Southern-blot analyses.(A) The wt ssp and the dsspregion followingthe sspA::NmR linear DNA recombination, For linear DNA transformation, pSNEOSP was digested with Sail +Scal. The resulting 2.75-kb Sail fragmentwas purified (Langridge et al., 1980) and used to transform JC7623 (recBC) and DPB267 (re.cO) to NmR. Strains JC7623 and DPB267 were made competent by the method of Hanahan (1983). Transformants were selected by growth on Nm (50 #g/ml)-containing LB plates. Media components were purchased from Difco. All other compounds used were commercial products of reagent grade. (Panel B)Southern-blot analyses of the parental strain, JC7623 (lanes I and 2) and the Nma transformant, MW7 (lanes 3 and 4). Blot I was probed with the intact 1.75-kb Sail ssp fragment. Strain JC7623 contains the wt 5.0-kbHindII1 fragment (lane I ) and the 1.75-kb Sail fragment (lane 2). Strain MW? contains two HindIlI fragments (lane 3) and a single 2.7-kb Sail fragment (lane 4) as predicted for the mutagenesis. Blot II was probed with the Ball-Sinai fragment from pSV77 which was deleted in the construction of pSNEOSP. Strain JC7623 shows the two wt bands (5.0-kbHindlII band, lane I and 1.75-kb SalI band, lane 2) while strain MW7 shows no hybridizing DNA fragments to the Bali-Sinai probe (lanes 3 and 4).

D N A from the Nm x transformants and JC7623, the parental strain, to confirm that the Nm R phenotype was a result of a gene replacement event. When the blots were probed with the intact 1.75-kb SalI fragment, the results showed that the parental strain contained the wt 5.0-kb HindIII fragment and the 1.75-kb Sail fragment (Fukuda et al., 1985). In contrast, a dsspd strain would be expected to show two HindIII fragments (4.1 and 1.9 kb) and a single 2.7-kb SalI fragment (Fig. 4B). Southern analyses of the parental strain and one of the strains shown to be an sspA null mutant, MW7, are shown in Figs. 4B-I and 4B-II. The fact that MW7 shows no fragments hybridizing to the

25 0.6-kb Ball-SmaI fragment confn'ms the fact that most of the ssp region has been deleted from the chromosome during mutagenesis. P I lysate p r o d u c t i o n a n d P I t r a n s d u c t i o n s P I experiments were carried out as described by Miller (1972). P l lysate production (plaque formation) on MW7 and its parental strain, JC7323, was attempted using 104-108pfu per 107-10 s exponential phase cells. The results (Table II) show that strain MW7 does not support P l plaque formation. It is known that recBC strains are deficient in P l replication giving burst size values (pfu/plaque) 10-20-fold lower than recBC + strains (Zabrovitz et al., 1977). We saw no deficiency in plaque formation (plaque diameter) between recBC and recBC + strains, therefore we chose to examine the possibility that the combination of the recBC and Assp mutations results in the absence of P l plaques. A second sspA null mutant, EAB1, was constructed by linear D N A transformation of a recD strain, DPB267. Strains which are recD are more robust than recBC strains (Russell et al., 1989). Strain EABI was tested using Southern-blot analyses to deter(c)

TABLE I! PI plaque formation and transduction results Strain a

Plasmid b

PI p l a q u e formation on agar platesc

Transducing particle formation in liquid lysatesd

C600 JC7623 MW7 DPB267 EAB 1

none none none none none pACYCSSP pACYCSSP- 1 pACYCSSP-2 none none pACYCSSP1 pACYCSSP-2

+ + +

ND ND ND +

-

-

+ + + +

+ ND ND ND ND ND ND

EABI

EAB 1 EABI ES 1364 ES ! 365 ES 1365 ES 1365

a ssp+ orpog÷and dssporpog-I strains used in P 1 lysate formations and transductions studies (see Table I). b Plasmids used in complementation studies shown in Fig. 5. Piasmid pACYCSSP contains the entire ssp operon, pACYCSSP-1 lacks the sspA gene, and pACYCSSP-2 lacks a portion of the sspB gene. c PI lysate production was carried out using 107 pfu per 10v to i0s exponential phase cellsgro'smin LB containing 10 mM MgCI2 and 5 mM CaCI2. PI transduction experiments were carried out as described by Miller(1972) on saturated cultures using 107-10s pfu per 1-5 x l0s cells. Following a 20-rain incubation at 37°C, 0.5 vol. (100 pl) of 1.0 M Na3" citrate was added and this mixture was spread on LB plates. Plwt was obtained from the American Type Culture Collection (ATCC). d PlcmcmOOlysate formation in liquid medium was carried out in LB medium containing 10raM CaCI2/20mM MgCI2, and 10mM NaCI. Phages Plvir and PlcmcmOOwere obtained from Dr. Betsy Kren. ND, not determined.

mine that the N m R phenotype was a result o f a gene replacement event. Strain EAB 1 behaves in an identical manner as M W 7 with respect to PI plaque formation (Table II). Since both recBC and recD strains were deficient in PI growth when made AsspA, we complemented the sspA mutation to confirm that this mutation was causing the P1 replication defect. Plasmid p A C Y C S S P (Fig. 5) was constructed by insertion of the 5-kb HindIIl fragment containing the ssF4 and the sspB genes from A-12F1 into pACYCI84 (Rose, 1988). Plasmid pACYC 184, a P 15 replicon, was selected as the vector for complementation ofthe AsspA mutation in the rec strains because ColE1 replicons are unstable in recBC mutants (Bassett and Kushner, 1984). Insertion of p A C Y C S S P into EAB1 restored the ability to support P1 plaque formation (Table II). To confirm that P I plaque formation on EABI[ p A C Y C S S P ] was the result of complementation of the AsspA mutation by the plasmid-borne sspA gene and not complementation of a putative polar mutation of sspB, two derivatives of p A C Y C S S P were constructed. Plasmid p A C Y C S S P - I (Fig. 5) was constructed by removal of the sspA gene by digestion with SmaI + Bali followed by religation. Plasmid pACYCSSP-2 (Fig. 5)was constructed by removal of the majority of sspB by digestion with BssHII followed by religation. Only pACYCSSP-2 which contains an intact sspA gene allowed P 1 plaque formation on EAB 1. Plasmid ACYCSSP-1, in which the sspA gene was removed,

Hlndlll --~

/ I1oll + Sinai digestion, Religalion

1

pAUYU~P

8ssHII Sell BssHII

~'-- Hindlll

\ 8ssHII digestion, Rellgstlon

1

Fig. 5. Construction of pACYCSSP-! and pACYCSSP-2. Piasmid pACYCSSP was obtained by inserting the 5.0-kb Hindlll fragment containing the entire sspAB operon (Fig. 1). Plasmid pACYCSSP-I has the majority of the sspA coding region deleted by digestion with Ball +Sinai followedby religation. Plasmid pACYCSSP-2 has approx. one half of the sspB gene deleted by digestion with BssHll followedby religation.

26 does not allow growth of P1 (Table II). From these results we conclude that the sspA gene is required for P1 plaque formation. Experiments using phage ). and T4 show that these bacteriophages form plaques on the sspA null mutants which suggest that this defect is P l-specific (data not shown). Plate lysates of strain EAB 1[pACYCSSP] were used as a source of PI for construction of additional AsspA strains by P 1 transduction. All Nm R transductants were screened by Southern analyses to verify they were Assp. One AsspA strain, MW8203, and its wt parent, CP78, were used to determine whether the asspA mutation affected the transducibility of E. coil In experiments involving transduction of MW8203 and CP78 to his +, transduction frequencies of 2 x 10-7 and 1 x 10-7, respectively, were obtained suggesting no defect in transducibility of the asspA strain. This result indicates that the defect in P1 growth on the asspA mutants occurs at a stage in the P 1 life cycle which follows injection of PI DNA.

(d) PI lysogeny and liquid lysate formation Bacteriophage Plcmcl.too (Rosner, 1972)which carries the Cm R gene and which is heat-inducible for lytic growth as a consequence of a heat-labile C 1 repressor protein, was used to examine the ability of the sspA null mutant to support lysogenic P1. Approximately l0 s cells were mixed with 100 #1 of a 103 per ml stock solution of Plcm=t.m o, incubated for 20 min at 30 ° C, and plated on Cm-containing 10

10 to 30'C

• Iv

/~-.~

CAO~2135 ~al

I"

42"C

I I

10 e

CAG1213Spl~.) I

t m ~Bt(PI~.)-'" I i z Pt l'tter"~=t=ne ]

agar plates. PI lysogens were obtained from both strains DPB267 and EABI. Fig. 6 shows the growth of these strains and their lysogenic derivatives at 30°C (prior to induction) and at 42°C (after induction). Strain DPB267[PlemcHoo ] lyses after approx. 45 min at 42°C during which time the PI titer increases by approx. 104 (Fig. 6). In cont:ast, EABl[Plcmcl.loo ] did not lyse even after se',,eral ho~a;rs a~ 42°C (Fig. 6). Detection of transducing particles in liq~id lysates was done by attempting to transduce his +, Tc s, and Cm R. Transductions of Cm R were attempted to determine whether the defect in P I growth was a result of a packaging step defect. The presence of packaged Pl DNA (Cm R transductants from the EAB 1[ P lcm~I.~oo]lysate) and no viable P 1 particles would be evidence of such a defect. No Pl-transducing particles were detected from EAB 1 either directly from the culture medium or following chloroform lysis of the cells (Table II). The temperature shift to 42°C (Fig. 6), however, does inhibit further growth of the mutant iysogen, EABI[PlcmcI.loo]. This suggests that PI, in the AsspA mutant strain, is carrying out some of its normal functions during lytic growth.

(e) Examination of P1 DNA replication in an sspA mutant To determine whether P1 DNA replication was occuring in the AsspA mutant strain, pulsed-field gel electrophoresis was used to determine the DNA species present at various times following induction of lytic growth in EAB 1[ P lcmc t. 1oo]' Agarose gels (0.8 %) containing plugs of 100 pl were run at 3-6 V/cm for 24 h with pulse times of 30 s. P1 DNA does accumulate in the mutant strain following induction (Fig. 7), suggesting that the dsspA mutation affects P I development following P I DNA replication.

1

E_

tOe

~

10 ?

~-"

< OA

1os

0,01

_ L

I

~

I

•

I

,

I

,

I

-3,o-2.s-2,o-l,s-l,o-o.s

,

.

n

o.0 o.s

•

,

•

~

.

,

•

,

.

10 s

1,o 1,5 2,o 2,s 3.o

Time (hours) Fig. 6. Induction of Plcmc,.too at 42°C. Growth curves of DPB267, EAB 1, and their lysogonic derivatives are shown. Cultures were grown at 30°C until an absorbance (600 nm) of 0.25 was obtained and then shifted to 42°C. PI titers were determined for the lysogens at 15-rain intervals during growth at 42°C. Absorbances are shown as blackened symbols; the PI titers from DPB267 (Plcm~.loo) are shown as open symbols. The PI titers (not shown) from EAB 1 (P lcm~t.,oo) were determined to be zero.

(f) Examination of proteins synthesized during induction Since DNA replica'ion, the primary function of early gene expression in P 1, was shown to occur in AsspA strains, late gene expression was examined by protein labeling experiments (Fig. 8). These results indicate that some proteins which are synthesized late (40- and 50-min time points) in the sspA+ strain, DPB267[Plcm, Lloo], are absent from the AsspA strain, EABl[Plcmclaoo]. One of these proteins appears to be the 44-kDa major head protein (Walker and Walker, 1981) (Fig. 8). Proteins expressed early in both strains continue to be synthesized during late time points in the AsspA strain. These results suggest that SSP may be required for both the termination of early gene expression and the initiation of late gene expression. (g) Comparison of sspA and pog Due to similarities between the sspA gene and a gene designated pog for 'growth of P 1' by Race (1984), we corn-

27 1

2

3

4

5

6

0

IO

20

30

40

50

rain

wt/mt wt/mt wt/mt wt/mt wt/mt wt/mt

kb

- 200

-100

Fig. 7. Pulsed-field electrophoresis agarose gel of EABI (Plcmc,.,oo) DNA aiderlytic cycleinduction at 42°C. CHEF separation ofE. coliand Pl chromosomalDNAs was carried out as describedby Chu et al. (1986). Cells harvested before and during induction of Pl iytic growth were chilled, centrifuged, and resuspended in 0.5 × TBE (1 × TBE: 89 mM Tris. borate/89 mM boric acid/2 mM EDTA)containing0.8 % agarose at a concentration of 109per ml and lysedby treatment at 30°C for 48 h with agitation in 10 mM Tris- HCI pH 8.0/100mM EDTA/I ~ Sarkosyl/ 100 #g per ml proteinase K. The electrophoresis apparatus used is identical to that described by Chu et al. (1986). Gels were run at 10°C for 24 h in 0.5 × TBE (in the presence ofethidium bromide) at 3 to 6 V/cm with pulse times of 30 s. Samples applied were purified Pl DNA, lane 1; EABI(no Plcmc,.~oo)-0min, lane 2; EABl(PlcmcHOo)-0min, lane 3; EABl(PlcmcLloo)-30min, lane 4; EABl(Plcmct.loo)-60min, lane 5; EAB I(PlcmcI.Ioo)'90rain, lane 6.

pared the characteristics of these two genes. The sspA gene has been mapped by transduction studies to 69.5 min (Serizawa and Fukuda, 1987) and pog to 69.8 min (Race, 1984) on the E. coli chromosome. Race (1984) showed that the pog-1 strain, ES 1365, can support lysogenic P 1, that P 1 D N A is synthesized during induction of the pog-1 lysogen, and that P 1 plaque formation is blocked; all of which are characteristics seen in the AsspA mutant. Furthermore, we have complemented thepog-1 mutation with p A C Y C S SP-2 (which contains the intact sspA gene) and found that this restored the ability of ES1365 (Table If) to support P1 plaque formation to wt levels (data not shown). These results suggest that pog and sspA are identical.

Fig.& SDS-PAGE of total cellular proteins labeled with [35S]. methionine during induction of Pl lytic growth. Cultures were grown at 30°C in MOPS medium(Neidhardt et al., 1974)supplementedwith 0.4~ giucose/0.2 mM adenine/0.2mM guanine/0.2mM cytosine/0.2mM uracil/0.1 mM thiamine/and 50 pg per ml of 19 aa (no methionine)until an absorbance (600 nm) of approx. 0.25 was reached and then shined to 42°C. Samples (1.0 ml)were removed and labeled for 5 min with 15 pCi of [sSS]methionine. A~er the 5-min labeling period, 167pi of 0.2 M methioninewas added. The designations (wt) and (mr) within each time point represent DPB267(Plcm~j.=oo) and EABl(Plcm~t.loo), respectively.The unlabeled lane at the fight contains Pl proteins,of which only the 44-kDa major protein of the viral coat can be seen (arrow). (h) Examination of the effect of thesis

sspA on stable RNA syn-

We have also examined the effect of the AsspA mutation on stable R N A synthesis to determine if SSP is involved in the curtailment of R N A synthesis upon limitation of any required aa or the resumption of stable RNA synthesis following resupplementation of a limiting aa. We constructed AsspA derivatives of CP78 and CP79 (MW8203 and MW9111, respectively) for the purpose of these experiments. MW8203 and MW9111 were determined to be AsspA by Southern-blot analyses (Fig. 4). To provoke aa limitation S H M T (Shand et al., 1989) was used at 0.6 mM. The results shown in Fig. 9A (the stringent pair, CP78 and MW8203) and Fig. 9B (the relaxed pair, CP79 and MW9111) indicate that the AsspA strains respond to aa limitation in very similar fashions as their sspA + parents. To investigate the possibility that SSP was involved in stable R N A synthesis during resumption of growth following aa limitation, cultures of CP78 and MW8203 were grown in medium containing a limiting amount of arginine. Following a period of approx. 12 h in stationary

28 90000 6000

CP7B-SHMT

/y

5000

80000

//

.w.2o3

70000

MW8

60000

4000

E 50000

E et

O

o 3000

40000 30000

2000

20000 1000

1O00O 0 O • v •

6000 5000

E £1. O

CP79 CP79-SHMT MW9111 MW9111-SHMT

/

v

0

30

60

90

120

150

Time (rain) Fig. 10. The effect ofsspA on recovery from aa limitation. Experiments were carried out in the same medium as described in Fig. 9 legend except that serine was included at 50 #g/ml and arginine was added to ouly 4 pg/ml final concentration. Cultures were limited for arginine for approx. 12 h after which 0.75 #Ci of a:PO 4 was added to 3.0-ml cultures, after 5 min, arginine was added (40 pg/ml). Samples of 75 pl were removed to 2-5 ml 10% cold trichloroacetic acid at 15-min intervals.

i

4000 3000 2000 1000 0

I

0

10

]

I

20

30

40

Time (rain) Fig. 9. The effect of sspA on stable RNA synthesis. Experiments were carried out as described by Ross ct al. (1990). Cultures were grown at 30°C in MOPS medium (Neidhardt et al., 1974)supplemented with 0.4% glucose/0.2 mM adenine/0.2 mM guanine/0.2 mM cytosine/0.2 mM uracil/0.1 mM thiamine/and 50 #g per ml of 19 aa (no serine). Serine starvation was induced by the addition of SHMT to 0.6 mM. After 5 rain 0.25 #Ci of 3"PO4 was added to 0.5-ml aliquots which had been removed to prewarmed culture tubes. Samples of 150 #1 were removed to 2-5 ml 10% cold trichloroacetic acid at 10, 20, and 30 min following SHMT addition,

phase, arginine was added and stable RNA synthesis was measured. The AsspA strain, MW8203, responded to aa resupplementation following aa starvation similarly to the sspA + parent, CP78 (Fig. 10).

(i) Conclusions (1) The sspA gene was isolated on ~.-12F1 from the Kohara ~.-library on a 5.0-kb Hindlll and located at 3450 kb (Kohara et al., 1987) on the physical map of the

E. colichromosome in the vicinity of and counter-clockwise from the mdh locus (Bachmann, 1990). Sequencing indicated that the sspA containing operon includes a second gene, sspB, consisting of 495 bp and a putative Rho-independent terminator structure immediately downstream. (2) Construction ofsspA null mutants was carried out to examine the role of S SP during aa limitation. Using SHMT, we have found that SSP is not involved in the regulation of

stable RNA synthesis during the aa downshift and that SSP is not required for stable RNA synthesis following resupplcmentation of arginine to an arginine-starved culture. (3) Construction of sspA null mutants has led to the unanticipated finding that these strains are defective in their ability to support P 1 vegetative growth. Comp[,ementation studies have shown that sspA andpog-1 (Race, 1984) represent the same gen¢. This conclusion is supported by the fact that the characteristics ofsspA and its effect on PI growth are identical to those reported forpog- 1 by Race (1984). The effect of the sspA mutation on phages /t and T4 was examined and found not to affect replication of these phages. More extensive work by Race (1984) has shown that bacteriophages T7, Mu, and P2 as well as T4 and show normal plaquing efficiencivs on peg-1 (sspA) mutants and that the absence of bacteriophage growth is specific for P 1 and the related phages P7 and D6. (4) P1 replication in zisspA mutants indicate that this mutation does not affect lysogenic growth and does not inhibit P 1 DNA replication. However, we detected the absence of some P 1 late gene products in the sspA mutants. The evidence that early P 1 proteins are synthesized during late time points and that some P1 late proteins are absent in the zisspA mutant, suggests that the inability of P1 to replicate in the AsspA mutant is due to the inability of P 1 to synthesize late gene products. It appears that SSP is involved in the transition from P1 early to P I late gene expression. (5) Work continues in this laboratory to determine whether the effect of SSP on P 1 late gene expression occurs at the level of transcription or translation. Two facts suggest

29 that S S P is a transcription factor which P1 utilizes during vegetative growth. First, the initial report on S S P purification (Ishihama a n d Saitoh, 1979) suggested that S S P was an R N A P - b i n d i n g protein; and second, m a n y bacteriophages, including P 1, use their host's cellular a p p a r a t u s via R N A P (Geiduschek a n d Kassavetis, 1988; Myers and Landy, 1973; Yarmolinsky and Sternberg, 1988). Understanding the role of S S P during P I vegetative growth m a y clarify the function o f S S P within E. coll.

ACKNOWLEDGEMENTS We thank Dr. D a n Gentry for providing strain D P B 2 6 7 , and Dr. Betsy Kren for providing strain JC7623 a n d Plvir in addition to m a n y useful suggestions. We wish to thank Dr. Eli C. Siegel for providing strains ES 1364 and ES1365 and for encouraging us to c o m p a r e AsspA and p o g - I strains. W e especially thank Dr. B a r b a r a J. B a c h m a n n for reading the manuscript and m a k i n g several very helpful suggestions. This work was supported in part by grant C D R 8 6 0 4 5 3 9 from the N a t i o n a l Science Foundation.

REFERENCES Appleyard, R.K.: Segregation of new lysogenic types during growth of a doubly lysogenic strain derived from Escherichia coli K-12. Genetics 39 (1954) 440-452. Bachmann, B.J.: The linkage map of Escherichia coli K-12, Edition 8. Microbiol. Rev. 54 (1990) 130-197. Bassett, C.L. and Kushner, S.R.: Exonucleases I, Ill, and V are required for stability ofColEl-related plasmids in Escherichia coll. J. Bacterioi. 157 (1984) 661-664. Biek, D.P. and Cohn, S.N.: Identification and characterization of recD, a gene affecting plasmid maintenance and recombination. J. Bacteriol. 167 (1986) 594-603. Cashel, M. and Gallant, J.A.: Control of RNA synthesis in Escherichia coil, I. Amino acid dependence of the synthesis of the substrate of RNA. J. Mol. Biol. 34 (1968) 317-330. Cashel, M. and Gallant, J.A.: Two compounds implicated in the function of the RC gene ofEscherichia coli. Nature 221 (1969) 838-841. Chu, G., Vollrath, D. and Davis, R.W.: Separation of large DNA molecules by contour-clamped homogeneous electric fields. Science 234 (1986) 1582-1585. Dykstra, C.C., Prasher, D. and Kushner, S.R.: Physical and biochemical analysis of the cloned recB and recC genes of Escherichia coil K-12. J. Bacteriol. 157 (1984) 21-27. Fill, N. and Friesen, J.D.: Isolation oPrelaxed' mutants of Escherichia coll. J. Bacteriol. 95 (1968) 729-731. Fukuda, R., Nishimura, A. and Serizawa, H.: Genetic mapping of the Escherichia coil gene for the stringent starvation protein and its dispensibility for normal cell growth. Mol. Gen. Genet. 211 (1988) 511-519. Fukuda, R., Yano, R., Fukui, T., Hase, T., Ishihama, A. and Matsubara, H.: Cloning of the Escherichia coil gene for the stringent starvation protein. Mol. Gen. Genet. 201 (1985) 151-157.

Geiduschek, E.P. and Kassavetis, G.A.: Changes in RNA polymerase, Vo!. I. In: Calendar, R. (Ed.), The Bacteriophages. Plenum, New York, 1988, pp. 93-115. Hanahan, D.: Studies on transformation ofEscherichia coli with plasmids. J. Mol. Biol. 166 (1983) 557-580. Haseltine, W.A., Block, R., Gilbert, W. and Weber, K.: MSI and MSII made on ribosome in idling step of protein synthesis. Nature 238 (1972) 381-384. Hoopes, BC. and McClure, W.R.: Strategies in regulation of transcription initiation. In: Neidhardt, F.C., Ingraham, J.L., Low, K.B., Magasanik, B., Schaechter, M. and Umbarger, H.E. (Eds.), Escherichia coil and Salmonella typhimurium: Cellular and Molecular Biology, Voi. 2. American Society for Microbiology, Washington, DC, 1987.. pp. 1231-1240. Ishihama, A. and Saitoh, T.: Subunits of RNA polymerase in function and structure, IX. Regulation of RNA polymerase activity by stringent s~arvation protein (SSP). J. Mol. Biol. 129 (1979) 517-530. Kohara, Y., Akiyama, K. and Isono, K.: The physical map of the whole E. coil chromosome: application of a new strategy for rapid analysis of a large genomic library. Cell 50 (1987) 495-508. Langridge, J., Langridge, P. and Bergquist, P.L.: Extraction of nucleic acids from agarose gels. Analyt. Biochem. 103 (1980) 265-271. Maniatis, T., Fritsch, E.F. and Sambrook, J.: Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982. Meyers, D.E. and Landy, A.: The role of host RNA polymerase in PI phage development. Virology 51 (1973) 521-524. Miller, J.H.: Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1972. Murray, N.E., Brammer, WJ. and Murray, K.: Lambdoid phages that simplify the recovery of in vitro recombinants. Mol. Gen. Genet. 150 (1977) 57-61. Neidhardt, F.C., Bloch, P.L. and Smith, D.F.: Culture medium for enterobacteria. J. Bacteriol. 119 (1974) 736-747. Oishi, M. and Cosloy, S.D.: The genetic and biochemical basis of the transformability of Escherichia coli K-12. Biophys. Res. Commun. 49 (1972) 1568-1572. Race, H.M.: An Escherichia coli Mutation that Blocks the Growth of Bacteriophage PI. Ph.D. Thesis, Tufts University, 1984. Reeh, S., Pedersen, S. and Friesen, J.D.: Biosynthetic regulation of individual proteins in relA + and relA strains ofEscherichia coli during amino acid starvation. Mol. Gen. Genet. 149 (1976) 279-289. Reiness, G., Yang, H.-L, Zuhay, G. and Cashel, M.: Effects of guanosine tetraphosphate on cell-free synthesis of Escherichia coil ribosomal RNA and other gene products. Proc. Natl. Acad. Sci. USA 72 (1975) 2881-2885. Riggs, D.L., Mueller, R.D., Kwan, H.-S. and Artz, S.W.: Promoter domain mediates guanosine tetraphosphate activation of the histidine operon. Proc. Natl. Acad. Sci. USA. 83 (1986) 9333-9337. Rose, R.E.: The nucleotide sequence of pACYC184. Nucleic Acids Res. 16 (1988) 355. Rosner, J.L.: Formation, induction and curing of bacteriophage Pl lysogens. Virology 49 (1972) 679-689. Ross, W., Thompson, J.F., Newlands, J.T. and Gourse, R.L.: E. coil Fis protein activates ribosomal RNA transcription in vitro and in vivo. EMBO J. 9 (1990) 3733-3742. Russell, C.B., Thaler, D.S. and Dahlquist, F.W.: Chromosomal transformation of Escherichia coil recD strains with linearized plasmids. J. Bacterioi. 171 (1989) 2609-2613. Sands, M.K. and Roberts, R.B.: The effects of a tryptophan-histidine deficiency in a mutant of Escherichia coll. J. Bacteriol. 63 (1952) 505-51 I. Sanger, F., Nicklcn, S. and Coulson, A.R.: DNA sequencing with chain-

30 terminating inhibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 5463-5467. Serizawa, H. and Fukuda, R.: Structure of the gene for the stringent starvation protein of Escherichia coll. Nucleic Acids Res. 15 (1987) ! 153-1163. Shand, R.F., Blum, P.H., Mueller, R.D., Riggs, D.L. and Artz, S.W.: Correlation between histidine operon expression and guanosine 5diphosphate levels and amino acid downshif~in stringent and relaxed strains of Salmonella typhimurium. J. Bacterioi. 171 (1989) 737-743. Tabor, S. and Richardson, C.C.: DNA sequence analysis with a modified bacteriophage T7 DNA polymerase. Proc. Natl. Acad. Sci. USA 84 (1987) 4767-4771. Taylor, L.A. and Rose, R.E.: A correction in the nucleotide sequence of the Tn903 kanamycin resistance determinant in pUC4K. Nucleic Acids Res. 16 (1988) 358. Tomkins, G.M.: The metabolic code. Science 189 (1975) 760-763.

Walker, J.T. and Walker, D.H.: Structural proteins of coliphage P 1. In: DuBow M.S. (Ed.), Progress in Clinical and Biological Research, Vol. 64. Liss, New York, 1981, pp. 69-77. Winans, S.C., EUedge, S.J., Kreuger, J.H. and Walker, G.C.: Site-directed insertion and deletion mutagenesis with cloned fragments in Escheri. chia coil J. Bacteriol. 161 (1985) 1219-1221. Yanisch-Perron, C., Vieira, J. and Messing, J.: Improved MI3 phage cloning vectors and host strains: nucleotide sequences of the Ml3mpl8 and pUCI9 vectors. Gene 33 (1985) 103-119. Yarmolinsky, M.B. and Sternberg, N.: Bacteriophage Pl. In: Calendar, R. (Ed.), The Bacteriophages, Vol. 1. Plenum, New York, 1988, pp. 291-438. Zabrovitz, S., Segev, N. and Cohen, G.: Growth of bacteriophage Pl in recombination-deficient hosts of Escherichia coli. Virology 80 (1977) 233-248.