MGG

Molec. gen. Genet. 147, 307-314 (1976)

© by Springer-Verlag 1976

Mapping of the polA Locus of Escherichia coli K12: Orientation of the Amino- and Carboxy-termini of the Cistron William S. Kelley and Nigel D.F. Grindley* Department of BiologicalSciences, Mellon Institute of Science, Carnegie-MellonUniversity,Pittsburgh, Pennsylvania15213, USA

Summary. Three mutations of the polA cistron, the structural gene for D N A polymerase I of E. coli, have been ordered by three factor transductional crosses. The three mutant polymerase species have altered properties which may be ascribed to defects located in different portions of the polypeptide chain. Our data indicate that the amino terminal end is encoded by the end of the polA cistron nearer to mete and that transcription and translation proceed clockwise on the E. coli circular map towards the rha locus.

Introduction In 1969 de Lucia and Cairns (1969) isolated the first mutant recognized as being defective in D N A polymerase I activity. This mutant, which they named polA1, was mapped by Gross and Gross (1969) and its location determined as being at 75 rain on the E. coli K12 map (Taylor and Trotter, 1972), linked with metE and rha by phage P1 transduction. Further enzymological analysis indicated that a non-complementing mutant designated poIA6, produced a structurally altered species of D N A polymerase I, thus confirming that the polA locus is the structural gene for D N A polymerase I (Kelley and Whitfield, 1971).. More recently a number of other pol- mutants have been isolated which are cotransducible with metE and rha and have different effects on the D N A polymerase I activity within the E. coli cell. Although not all of these mutants have been tested by complementation analysis, they are presumed to lie within the polA cistron. D N A polymerase I is a complex enzyme catalyzing several different reactions at different enzymati* Dr. Grindley's current address is: Department of Molecular Biophysics and Biochemistry,Yale University, New Haven, Connecticut 06510, U.S.A.

cally active sites within a single polypeptide chain of about 1,000 amino acid residues (Kornberg, 1969). Since the enzymatic activities of the pure protein may be analyzed selectively by independent in vitro assay procedures, it has been possible to assign several of the mutational lesions to different positions within the polypeptide, based on their effects on different enzymatic activities known to be at amino- or carboxy-terminal ends of the polymerase molecule (e.g., Lehman and Chien, 1973; Heijneker and Klenow, 1975). Since D N A polymerase I is clearly an important enzyme which functions in the replication and repair of bacterial chromosomal and plasmid D N A (for review, see Gefter, 1975), a more complete analysis of the structural defects of the many different polAalleles is necessary for total understanding of its intracellular role. Unfortunately, the isolation of mutant polymerase is frequently difficult due to altered (assayable) enzymatic activities or the instability of the protein in vitro (Kelley and Whitfield, 1971 ; Lehman and Chien, 1973). Therefore, we have turned our attention to the construction of a fine structure map of the polA cistron by standard genetic techniques. By correlating transductional mapping data with the enzymological experiments of ourselves and others we have been able to construct a preliminary map. Our experiments indicate that the amino terminus of the D N A polymerase I polypeptide is encoded by the end of the polA cistron proximal to metE.

Materials and Methods

Genetic Techniques Transductional techniques with phage P1CM (Rosner, 1972) and media have been described elsewhere(Grindleyand Kelley, 1976). Transductionswere routinelycarried out at 32° C to avoid selecting against any conditional lethal polA- mutants used in these and subsequent experiments. Phage infections were carried out at a

308

W.S. Kelley and N.D.F. Grindley: polA Intragenic Mapping

Table 1. Bacterial strains a) Original strains

JG108=E. coli K12F- thyA a lac- rha m e t E - str r 1R713=E. coIi K12 prototroph J G l l l =E. coli K12 thyA m e t E - polA6 (cys-) b KMBL1789=E. eoli K l 2 F - thyA- argA bio- p h e A - endA- polA'107 DS694=E. coli K12F- thyA lac rha- polAl str

from from from from from

J.D. Gross via J.A. Wechsler E.S. Anderson J.D. Gross H.L. Heijneker D.J. Sherratt

b) Intermediate strains

CM1068 =P1CM lysogen of DS694 CMlO69=metE + polA1 rha transductant of JG108 with phage P1CM from CM1068 CM3OO9=metE- polA6 rha + transductant of JG108 with phage P1 from J G l l 1

CM3225=P1CM lysogen of 1R713 CM3567=P1CM lysogen of KMBL1789 CM3571 = m e t E + poIA6 rha + transductant of CM3009 with phage P1CM from CM3225 CM3574=metE + polA1 rha + transductant of CM1069 with phage P1CM from CM3225 CM3576=P1CM lysogen of CM3571 CM3578=P1CM lysogen of CM3574 CM3760=metE + pol + rha- transductant of JG108 with phage P1CM from CM3225 C M 3 7 6 2 : m e t E pol + rha + transductant of JG108 with phage P1CM from CM3225 CM3764=metE + pol + rha + transductant of CM3760 with phage P1CM from CM3225 CM3842= metE + poIA'107 rha + transductant of CM3762 with phage P1CM from CM3567 c) Experimental strains

CM3791 =P1CM lysogen of CM3764 CM1069= metE + polA1 rha transductant of JG108 with phage P1CM from CM1068 CM3599= metE-polA1 rha + transductant of JG108 with phage P1CM from CM3578 CM3578 =P1CM lysogen of CM3574 is metE + polA1 rha + CM3593=metE + polA6 rha- transductant of JG108 with phage P1CM from CM3576 CM3009= metE-polA6 rha + transductant of JG108 with phage P1 from JG111 CM3576=P1CM lysogen of CM3571 is metE + polA6 rha +

CM3819 = m e t E + polA'107 rha transductant of JG108 with phage P1CM from CM3567 CM3829= metE-poIA'107 rha + transductant of JG108 with phage P1CM from CM3567 CM3863=P1CM lysogen of CM3842 is metE + polA'107 rha + Gene symbols are from Taylor and Trotter (1972) a

JG108 is a low level thymine requirer selected by a two-step trimethoprim enrichment procedure. The strain is

thyA- but there is no genetic evidence to indicate whether it is dra- or drum-

b J G t l l is the original mutagenized stock. The strain does not grow well on minimal plates unless they are supplement with cysteine multiplicity of infection of one plaque forming unit (pfu) per cell which gives a frequency of lysogenization by the phage of less than five percent. Transductants were tested for lysogeny and any lysogens were excluded from scoring of the mapping experiments (see Results, below).

Bacterial Strains

The polA - alleles were transduced from their original mutagenized background into a standard E. coli K12 W3110 derivative lacking any known plasmids. Strain construction was carried out as described in Grindley and Kelley (1976). A full list of all derivative strains with their abbreviated pedigrees is found in Table 1.

activated calf thymus DNA (Fansler and Loeb, 1974) as the template with pH 8.7 tris buffer in the standard DNA polymerase I assay system of Setlow (1974) or the kinetic assay described by Kelley and Whitfield (1971). 5 ' ~ 3' exonuclease assays were carried out using 3'-phosphoryl terminated 3H-dAT copolymer in the DEAE filter paper assay described by Setlow (1974) or the assay system of Lehman and Richardson (1964) utilizing 32p-labelled E. coli DNA (Murray, 1969). Subtilisin cleavage reactions were carried out according to the protocol of Klenow, Overgaard-Hansen and Patkar (1971). Reagents

Unless specifically noted, all reagents were from standard laboratory supply houses.

DNA polymerase I

The polA6 and wild type E. coli K12 DNA polymerase I was prepared as previously described (Kelley and Whitfield, 1971) from either the original polA6 mutant strain or a P1CM transductant of it and from an isogenic pol + E. coli K12 strain. Greatest levels of activity of the polA6 enzyme are obtained using maximally

Results The rationale of the experiments presented here is to use three factor transductional crosses to establish the position of three reference mutations within the

w.s. Kelley and N.D.F. Grindley: polA Intragenic Mapping

polA cistron. Knowing the nature of the enzymatic alteration produced by each of these mutations the orientation of the cistron may then be inferred from the location of the mutational defects with respect to the amino- and carboxy-terminal ends of the polymerase polypeptide. Three mutants were chosen for analysis, they are:

309 ity in vitro. The polA6 mutation does not complement polA1 and initial reports indicated that the enzyme was structurally similar to wild-type D N A polymerase I but was thermolabile (Kelley and Whitfield, 1971).

Localization of the polA6 Mutational Defect polA1. This is the original polA- mutant of de Lucia and Cairns (1969). Its cotransductibility with metE and rha were initially used to define the position of the polA gene. The mutation is recessive in poIA+/ poIA1 merodiploids. PolA1 is an amber mutant and the experiments of Lehman and Chien (1973) have shown that an unsuppressed polA1 E. coli strain produces small quantitities of D N A polymerase I - p r e sumably by occasional read-through of the amber codon. These cells also contain a 35,000 dalton polypeptide fragment which possesses the 5 ' ~ 3 ' exonucleolytic activity of D N A polymerase I but lacks polymerizing capacity. Evidence presented by Friedberg and Lehman (1974) indicates that this fragment is essentially identical to the 35,000 dalton fragment produced by limited proteolysis of the intact enzyme. The sequence of the amino-terminal tetrapeptide of this proteolytic fragment is identical to that of the intact enzyme (Jacobsen, Klenow and OvergaardHansen, 1974). The polAl mutation therefore appears to lie at or distal to the mid-point of the cistron, terminating the synthesis of D N A polymerase I after the amino-terminal fragment has been translated. The precise intracistronic position of the polA1 amber mutation is not demonstrated by such an experiment since the 35,000 dalton polypeptide may result from partial digestion of the amber fragment by intracellular proteases.

Initial experiments (Kelley and Whitfield, 1971) indicated that the polA6 polymerase was inactive in the polydAT-primed assays described by Jovin, Englund and Bertsch (1969). Further experimentation has revealed that this behavior represents the combined effects of several phenomena. When assayed in tris buffer the polA6 polymerase exhibits optimal polymerization activity at pH 8.7, utilizing E. coli DNA, maximally activated D N A or poly dAT. This activity is markedly reduced when the reaction is buffered with potassium phosphate at pH 7.4 (Table 2). Furthermore, the polymerization activity of the poIA6 enzyme in tris buffer at pH 8.7 may be titrated by the addition of p H 8.7 potassium phosphate buffer to the initial reaction mixture. An overall reduction of polymerase activity by twofold is observed in such an experiment when the phosphate level is increased from zero to fifty millimolar. Under identical conditions, the wild-type polymerase is actually slightly stimulated by the addition of the same amounts of phosphate.

Table 2. Specific activities of DNA polymerase I from wild-type and polA6 mutant E. coli Specific activity on respectivetemplate (units/rag enzyme)

polA'107. This mutant allele was isolated and characterized by Glickman etal. (1973). Their genetic studies indicated that it is cotransducible with metE and rha and that it fails to complement the polA1 mutant in experimental merodiploid strains. The D N A polymerase I purified from polA'107 strains is defective in 5 ' ~ 3 ' exonucleolytic activity (Heijneker et al., 1973). Cleavage of the enzyme with subtilisin releases a fragment with polymerizing activity essentially identical to that released from wild type D N A polymerase I (Heijneker and Klenow, 1975). The wild type fragment has been shown to correspond to the carboxyl-terminus of the intact enzyme (Jacobsen, Klenow and Overgaard-Hansen, 1974). The polA'107 mutation thus lies in the amino-terminal end of the molecule.

polA6. This mutant allele produces a D N A polymerase I species which is defective in polymerizing capac-

dAT random copolymer

E. coli B/r DNA

Maximally Activated Calf Thymus DNA

Wild type

pH 7.4 pH 8.7

8,066 1,658

4,336 3,458

17,404 8,310

polA6

pH 7.4 pH 8.7

4,034 17,638

778 3,802

792 5,228

Assays were carried out in triplicate at three different dilutions of enzymeusing the assay systemdescribed by Setlow (1974) buffered with either 67 mM potassium phosphate pH 7.4 or 67 mM tris pH 8.7. The assays contained primer/templateDNA as indicated-E. coli B/r DNA prepared by the method of Murray (1970), maximally activated calf thymus DNA prepared by the method of Fansler and Loeb (1974), or commercial dAT random copolymer in the absence of any added exonuclease III. For most assays linear response of incorporation to added enzymewas observed in the range of 0.05 to 0.25 micrograms of enzyme per assay. Specific activities werecalculatedon the basis of 10 nmoleunits as described by Setlow (1974)

W.S. Kelley and N.D.F. Grindley: polA Intragenic Mapping

310

#olA6

Wild Type f.~.--:~--. 48_,mM PO4 100

100

>" 80 d'~ o

"~

80

P04

E 60E

(

[

u 40 O..o,--O- --O . . . . .

0 mM PO4

4O

""O

2O

G.m----m.. 48 mM P04

L,/':

"..

3ZP-E. coli DNA During polymerization Activity: units/gg

2O

rl

' 2'0

' 4~0 ' 6'o

0~3 ' 210 ' 4'0 ' Minutes of Subtilisin Digestion

wild-type p olA6

28,000 32,400

3,257 6,140

~D

o)

00

3'-Phosphoryl dAT-Copolymer Activity: units/gg

60,

x

E

Table 3. 5 ' ~ 3 ' exonuclease assay of DNA polymerase I from wild-type and polA6 E. coli K12

60

Fig. I. Effect of phosphate on the large fragment of wild type and polA6 DNA polymerase I. DNA polymerase I was cleaved with subtilisin following the procedure described in Materials and Methods. The incubation mixture contained 75 lig of DNA polymerase I, 0.05 lzg of subtilisin, 65 gg of calf thymus DNA and 10.6 gmole of potassium phosphate, pH 6.5 in a total volume of 105 gl. At times of 0, 5, 10, 15, 20, 30, and 60 rain 5 gl samples were removed and pipetted into 0.25 ml of 10 mg/ml bovine serum albumin solution and placed in ice for assay at completion of the reaction. At the same times 10 gl samples were also removed and mixed with 10 gl volumes of 2.4 mg/ml phenylmethylsulfonyl fluoride. The latter samples were analyzed in an SDS-polyaerylamide slab gel following the procedure of Laemmli (1970) using the apparatus of Reid and Bieleski (1968). These data (not shown here) indicated that quantitative cleavage of the polymerase was completed by 10-20 rain. Comparison of the band obtained with those of known standards indicated the large fragment molecular weight as being 73,000. The diluted enzyme was assayed by the method described by Kelley and Whitfield (1971) utilizing maximally activated calf thymus DNA as a primer/template and 67raM tris buffer, pH 8.7 at 37 °. Three separate sets of assays were run with a solution of pH 8.7 potassium phosphate being added to give a final concentration of 0, 19 mM and 48 mM phosphate. Results presented here are normalized to maximal incorporation values for the respective enzymes

Nuclease assays with 3'-phosphoryl terminated polydAT or 3zp_ Labelled E. eoli DNA plus four added triphosphates as indicated in the text. Both assays measure 5 ' ~ 3 ' exonucleolytic activity preferentially

the wild-type and the polA6 enzymes were similarly active in digesting 3'-phosphoryl poly dAT or in degrading 32p labelled E. coli D N A under conditions favoring polymerization. The 3 ' ~ 5 ' exonuclease activity of DNA polymerase I is strongly inhibited by a 3'-phosphoryl terminus (Deutscher and Kornberg, 1969) and polymerization has been shown to stimulate the concomittant 5'--. 3' exonucleolytic degradation of the primer-template necessary for "nick translation" (Lehman, 1967). The results of these assays are summarized in Table 3. Thus, these experiments demonstrate that the polA6 DNA polymerase I is defective in its capacity to catalyze the polymerization of DNA chains and that this defect is localized within the carboxyl end of the molecule. The enzyme does not appear to be defective in the 5'--.3' exonucleolytic activity which is found in the amino-terminal end of the molecule.

Genetic Mapping of the poIA Locus The wild-type and polA6 polymerase were cleaved by subtilisin and polymerase activity was measured in tris buffer, pH 8.7, in the presence of added potassium phosphate (Fig. 1). The proportion of 73,000 dalton "large fragment" of the wild-type enzyme produced by cleavage increases with time of digestion and its activity is stimulated by the addition of potassium phosphate, as observed by Klenow, OvergaardHansen and Patkar (1971). The fragment produced by cleavage of the polA6 enzyme is inhibited by added phosphate under the same conditions although net polymerizing activity is not lost when assayed in the absence of phosphate. This large fragment has been shown to comprise the carboxy-terminus of the molecule by a number of criteria (e.g., Lehman and Chien, 1973; Jacobsen, Overgaard-Hansen and Klenow, 1974). Assays of the nucleolytic capacity of the intact polymerase molecule do not indicate any gross defects in the enzyme's 5 ' ~ 3 ' exonucleolytic function. Both

In an attempt to define the relative position of the three mutations within the polA cistron a series of three factor crosses was performed. In each cross metE- polA- or rha-polA- bacteria were transduced to met + and rha + respectively with a transducing lysate from another polA- mutant and the selected transductants were tested for the presence ofpol + recombinants. Since the linkage of polA to metE and to rha is not great, this experiment involves testing a large number of transductant clones to find a few recombinants. Because of the possibility of spontaneous reversion of the recipient or the presence ofpol + revertant transducing phage, great care was taken to conduct control experiments to determine the numbers of random revertants introduced by any particular cross. A) Control Crosses. Control crosses were conducted as described in Table 4 to determine the level ofpol + revertants produced by the transductional cross of any particular allele into itself. Although the recipient

311

W.S. Kelley and N.D.F. G r i n d l e y : p o l A Intragenic M a p p i n g Table 4. Control crosses to test for p o l + revertant phage Phage Stocks

Bacterial Strains

C M 3 5 7 8 - - m e t E + p o l A 1 rha + C M 3 5 7 6 - - m e t E + p o l A 6 rha + C M 3 8 6 3 - - m e t E + p o l A ' 1 0 7 rha +

C M 3 5 9 9 - - m e t E - p o l A 1 rha + C M 3 O O 9 - - m e t E - p o l A 6 rha + C M 3 8 2 9 - - m e t E - p o l A ' 1 0 7 rha +

Transductions performed

Total m e t E + transductants analyzed

Total

%

metE + pol +

pol +

4 4 1

2,466 3,506 1,296

1 1 0

0.041 0.029 0

To test for p o l A + revertants in our phage stocks or for selfing, a lysate of phage P 1 C M was prepared from each of the three strains CM3576, CM3578 and CM3863 (see Table l) at a titer of greater than 109 pfu/ml for use in these and subsequent crosses. The m e t E bacterial strains. CM3009, CM3599 and CM3819 were inoculated into broth from single clones on fresh streak plates and grown for several generations. The broth cultures were then infected with phage isolated from the transducing lysate carrying the same p o l A allele, spread on selective m e d i u m lacking methionine and incubated at 32 °. Individual m e t + transductant colonies were picked from these plates, patched onto nutrient agar masters and allowed to grow overnight. These masters were replica plated onto nutrient agar plates with and w i t h o u t 0.02% methylmethane sulfonate (MMS), incubated at 42 ° overnight and scored. P 1 C M lysogens were killed on both replicas; p o l - lines grew on the nutrient agar but not on t h e replica containing MMS. The p o l + recombinants and revertants which grew on both replicas were scored, streaked on minimal agar plates, picked and retested to verify their p o l A + phenotype The recipient culture was tested for the presence of revertants by spread plating directly onto M M S at 10 2 dilution before adding transducing phage and bY simultaneous plating at 10 4 dilution onto nutrient plates followed by replica plating onto M M S after overnight growth. These two platings indicated the frequency of p o l + revertants in the recipient culture; any cross in which the level of p o l + revertants exceeded 0.01% was discarded

Table 5. Backcrosses o f p o l A Phage Stock

CM3791 m e t E + p o I A + rha + C M 3 7 9 1 - m e t E + p o l A + rha + C M 3 7 9 1 - - m e t E + p o l A + rha + C M 3 7 9 1 - - m e t E + p o l A + rha + C M 3 7 9 1 - - m e t E + p o l A + rha + C M 3 7 9 1 - - m e t E + p o l A + rha +

recipients to p o l + Bacterial Strains

CM1069 CM3599

Transductions performed

m e t E + polA1 rha3 m e t E - p o l A 1 rha + 1 CM3593--metE + polA6 rha2 C M 3 O O 9 - - m e t E - p o l A 6 rha + 1 CM3819--metE + polA'107 rha- 2 CM3829 m e t E p o l A ' 1 0 7 rha + 1

metE +

rha +

pol +

Transductants analyzed

Transductants analyzed

Transductants

512 407 1,003

370 545 1,807 -

18 64 58 63 49 102

% pol + met +

% pol + rha +

12.5 15.5 10.2

4.9% 10.6 6.0 -

Transductions were carried out and scored as described in the text. Potential p o l + transductants were restreaked on selective medium and ten clones of each were retested. Segregation of the non-selected p o l A markers was observed in m a n y cases. Behavior of this type is consistent with data previously reported (Lennox, 1955) and presumably represents segregation of different varieties of recombinants from the muttinucleate recipient, No further segregation was observed after a second cycle of streaking and testing and the transductants presented none of the properties of defective phage lysogens. A transductants was thus scored as p o l + if one or more of the ten clones was M M S resistant

culture was rigorously checked for spontaneous pol + revertants, in the case of polA1 xpolA1 and polA6 xpolA6, pol+t. Transductants were observed at the level of 0.04% and 0.03%, respectively. These met + pol + transductants are either the result ofpol + revertants in the phage stock or of "selfing" (Demerec, 1963). Since the recipient strains have equal colony forming ability at 32 ° and the plating does not selectively enrich for pol + strains, these data indicate that revertants phage or self transductants can account for a few polA ÷ clones but that this frequency is less than 0.04% of the transductants tested. Since the same phage lysates were utilized for subsequent crosses, we may conclude that pol + transductants ap-

pearing at frequencies of greater than 0.04% represent recombinants rather than revertants. B) Back Crosses. Different data have been published expressing the linkage of polA to metE and to rha. Gross and Gross (1969) estimated the linkage as 17 to 22% for the metEpolAl linkage and 6% for the polAl rha linkage. Glickman et al. (1973) estimated 27.3% linkage for metEpolA1 and 26.5% for metEpoIA'107. Since these data are based on transduction of metEcells to met +poIA-, we carried out the reverse transduction of metE-polA2 and rha-polA2 to met + and rha + respectively to verify their results by a formal back cross. The transductions were carried out

W.S. Kelley and N.D.F. Grindley: p o l A Intragenic Mapping

312 Table 6 Phage stocks

Bacterial strains

Transductions

Transductants analyzed

pol +

5 4 6 4 7 4

3,527 4,171 3,691 5,699 4,630 3,176

2 12 6 3 18 0

0.06 0.28 0.16 0.05 0.39 0.00

4 3 4 4 4 3

5,608 2,915 4,150 1,862 2,672 3,747

22 3 0 6 0 16

0.39 0.10 0.00 0.32 0.00 0.43

% pol +

Transductants

a) Crosses o f p o l A - alleles-Met + selection CM3578--metE CM3576--metE CM3578--metE CM3863--metE CM3576 m e t E CM3863 m e t E

+ p o l A 1 rha + + p o l A 6 rha + + p o l A 1 rha + + p o l A ' 1 0 7 rha + + p o l A 6 rha + + p o l A ' 1 0 7 rha +

b) Crosses o f p o l A CM3578--metE CM3576-metE CM3578--metE CM3863--metE CM3576--metE CM3863--metE

CM3OO9--metECM3599-metE CM3829--metECM3599--metECM3829--metECM3OO9--metE-

p o l A 6 rha + p o l A 1 rha + p o l A ' 1 0 7 rha + p o l A 1 rha + p o l A ' 1 0 7 rha + p o l A 6 rha +

alleles-Rha + selection

+ poIA1 rha + + p o l A 6 rha + + p o l A l rha + + p o l A ' 1 0 7 rha + + p o l A 6 rha + + p o l A ' 1 0 7 rha +

CM3593--metE metE CM3819--metE CMlO69--metE CM3819--metE CM3593 m e t E

CM1069

+ polA6 rha+ polA1 rha+ p o l A ' 1 0 7 rha + polA1 rha+ polA'107 rha+ p o l A 6 rha

using the met + pol ÷ rha + phage from CM3791 and the six possible recipient cell lines. The same transductional protocol was followed to minimize the frequency of spontaneous polA ÷ revertants among the transductants. Results of these crosses are displayed in Table 5. These data show that all of the recipient strains have the polA2 allele located near mete and rha as expected although linkage data from this type of cross are slightly different from those previously reported. Minor apparent differences in the linkage of the three polA- alleles to mete and to rha are not deemed significant for positioning the different alleles on an intragenic map, however. C) Three-factor Transductional Crosses. The three polA- alleles have been positioned with respect to each other by carrying out three factor crosses between each of the six possible combinations of metE and of rha transductants. The same precautions were taken to avoid pol + revertant bacterial recipient stocks and the same phage lysates were used. The results are displayed in Table 6. Of each pair of reciprocal crosses using the same selected marker (Table 6 a, b), only one produces pol + lines at a frequency significantly above the level of spontaneous or transduction-generated reversion (Table 4). Our results are consistent with the unambiguous gene order metE-polA'lOT-polAl-polA6-rha and show that quadruple crossovers are much less frequent than doubles despite the large distance between metE or rha and polA as illustrated in Table 6 (c). In all cases the results from rha + selection support those obtained when metE + recombinants were selected. The map order shown above is compatible with the evidence available

about the defects in the polymerase produced by these three mutant alleles. It implies that transcription and translation of the cistron proceed from that end nearer to mete toward the end nearer to rha.

Discussion

The conclusions resulting from these data and further applications of such mapping techniques should allow us to interpret more precisely the effects of other polA- mutations and hopefully eventually to understand the role of DNA polymerase in DNA replication. The conditional lethal polymerase mutations such as poIAexl (Konrad and Lehman, 1974) and polA BT4113 (Olivera and Bonhoeffer, 1974) effect the 5<-.3 ' exonucleolytic function and would be expected to be positioned near poIA'107. Numerous other polymerase mutants affecting the polymerase function of the enzyme would be expected to map near poIA6. However, because we do not presently know any detailed information about the configuration of the folded polypeptide, it is possible that single site mutations could have pleiotropic effects resulting from intramolecular conformational alterations. Unfortunately, this type of mapping is tedious because of the low linkage of mete and rha to the polA locus. We have been unable to use any of the other markers in this region for our selection at present although we hope to do so. A more closely linked external marker would facilitate the analysis of transductants by increasing the frequency of actual transduction of the polA locus. Our back cross data indicate that at present fewer than fifteen percent of the

W.S. Kelley and N . D . F . Grindley: polA Intragenic M a p p i n g

313

Table 6 (continued) c) Schematic s u m m a r y of the data

mete

polA

rha I

polA'107 "

At

mete I

CM3578

----"

I

rha P°l+/metE+

l

~ I

k__-I

I

pol+/rha +

0.06%

A1 I

metE

' I A6

A6 I

A6 I

----I

0.28%

0.10% ] A1

1

A1

C-q

I

1

I I

---

0.16%

I I

0.00%

r i

........ I

I

A'107

CM3599

,_1 . . . . . . .

. . . .

I I I i

1

I

I

I

I

0.05%

I

I

I .....

I

,

0.32%

0.39%

t I

CM3829

,r---i, 0.00% ...... I

A'107

r I I

I

I I I

I

i" .... ,

[ ]

!

[ ~. . . . .

I A'I07

0.00%

J_

1

',

~ 1

i. . . . .

I

metE

A6

rha

CM3863

1

I

J metE

CM3819

I I

I. . . . .

I

CM3009

CM1069

CM3576

I

i'

0.43%

CM3819

CM3863

1 I

I

A'107 I

A'I07 CM3863

L 2

1. . . .

I

A6 I

I

CM1069

CM3578

I 1. . . .

I A1

A6 I ____|

/_

I

I

1

1 k.....

,

a

1 I

I

fI

CM3576

!

I....

I A'107

A1

CM3576

I

I

A'107 I ...............

A'I07 CM3863

CM3895

I

I

I

I I

I f. . . . . .

I

,41 I ............

A1 CM3578

, 1 L_J

CM3578

I

I I

I

I I I I J

I I [

CM3829

rha [

I- ............

0.39%

I. . . . . . . . .

I A6

CM3576

CM3599

poIA6

F--I

I

CM3009

polA1

I A6

I rha

CM3593

Transductions were carried out and scored as described in the text. Potential pol + transductants were restreaked on selective m e d i u m and 50-100 clones of each were retested. Segregation p h e n o m e n a were observed as noted in the legend to Table 5. A transductant was scored as pol + if ten percent of such daughter clones were resistant to M M S . F u r t h e r restreaking and retesting of r a n d o m pol + daughters indicated no further segregation ofpol+/pol phenotype, indicating that the r e c o m b i n a n t s (or revertants) were genetically stable

mete + transductants and fewer than ten percent of the rha + transductants actually incorporate a transducing fragment covering the polA cistron. These numbers are somewhat lower than those previously reported (Gross and Gross, 1969; Glickman et al., 1973) but agree well with the value at 12.5% cotransduction predicted by the equation of Wu (1966) for markers separated by one minute on the E. coli ge-

nome. Thus, the recombinational frequencies of 0.16 to 0.45% pol+/total selected transductants actually must represent much higher frequencies of recombination between the paired homologous DNA strands covering the polA cistron during transduction. Recombination frequencies of this level may easily be rationalized with the large size of the polA cistron and the wide spacing of the three mutations used

314

in this study. However, at present we are reluctant to assign linear distances to the separations between the three alleles since our data are not extensive. Other more rapid and more precise mapping methods are being pursued. Definitive mapping of the locus with respect to outside markers could be accomplished ifpolA ÷ recombinants could be selected directly and then analyzed for metE+/metE - or rha ÷/ rha- ratios. However, all tests for pol ÷ phenotype currently employed rely on the resistance of the cells to mutagenic agents, precluding such an approach. Deletion mapping could also provide a more promising approach. At present, no deletion mutants of the polA cistron are known and the lack of such deletions may be interpreted as evidence that the polA gene is necessary for cell function. However, the arguments of Konrad and Lehman (1974) and Olivera and Bonhoeffer (1974) imply that the N-terminal portion (the 5 ' ~ 3 ' exonuclease) of the enzyme is the obligatory function. Knowing the orientation of the cistron it may now be possible to isolate partial deletions of the carboxy-terminal and of the gene to yield mutants valuable in mapping and to resolve the role of the polymerase activity of DNA polymerase I. Acknowledgments. This work was supported by U.S. Public Health Service Grant GM-19374 to WSK and by Allegheny County (Pennsylvania) Health Research and Services Foundation Grants 0-89 to WSK and R-85 to N D F G ; the latter is a postdoctoral research fellow of the Science Research Council of Great Britain. We are grateful for gifts of strains from the individuals mentioned in the text. Special thanks is due to Dr. William E. Brown for his help with the subtilisin cleavage and to Dr. David Finnegan for his helpful discussions• Mr. Norman LeDonne, Miss Jane A. Rehl and Mrs. Charlene A. Rizzo provided invaluable technical assistance in this project. •

References de Lucia, P., Cairns, J. : Isolation of an E. coli strain with a mutation affecting DNA polymerase. Nature (Lond.) 224, 1164-1166 (1969) Demerec, M. : Seller mutants of Salmonella typhimurium. Genetics 48, 1519-1531 (1963) Deutscher, M.P., Koruberg, A. : Enzymatic synthesis of deoxyribonucleic acid., XXIX. Hydrolysis of deoxyribonucleic acid from the 5' terminus by an exonuclease function of deoxyribonucleic acid polymerase. J. biol. Chem. 244, 3029-3037 (1969) Englund, P.T., Kelly, R.B., Kornberg, A. : Enzymatic synthesis of deoxyribonucleic acid., XXXI. Binding of deoxyribonucleic acid to deoxyribonucleic acid polymerase. J. biol. Chem. 244, 3045-3062 (1969) Fansler, B.S., Loeb, L.A. : Sea urchin nuclear DNA polymerase. In: Methods in enzymology, XXIX. L. Grossman, K. Moldave, • eds., pp. 53-70. New York: Academic Press 1974 Friedberg, E.C., Lehman, I.R. : Excision of thymine dimers by proteolytic and amber fragments of E. coli DNA polymerase I. Bioch. Biophys. Res. Commun. 58, 132-139 (1974) Gefter, M.L. : DNA replication. Ann. Rev. Biochem. 44, 45-78 (1975) Glickrnan, B.W.o van Sluis, C.A., Heijneker, H.L., Rfrsch, A.: A mutant of Escherichia coli K12 deficient in the 5 ' ~ 3 ' exonu-

W.S. Kelley and N.D.F. Grindley: polA Intragenic Mapping cleolytic activity of DNA polymerase I, I. General Characterization. Molec. gen. Genet. 124, 6 9 4 2 (1973) Grindley, N.D.F., Kelley, W.S. : Effects of different alleles of the E. coli K12 polA gene on the replication of non-transferring plasmids. Molec. gen. Genet. 143, 311-318 (1976) Gross, J.D., Gross, M.M. : Genetic analysis of an E. coli strain with a mutation affecting DNA,polymerase. Nature (Lond.) 224, 1166-1168 (1969) Heijneker, H.L., Ellens, D.J., Tjeerde, R.H., Glickman, B.W., van Dorp, B., Pouwels, P.H. : A mutant of Escherichia coli K12 deficient in the 5'--+3' exonucleolytic activity of DNA polymerase I, II. Purification and properties of the mutant enzyme. Molec. gen. Genet. 124, 83-96 (1973) Heijnker, H.L., Klenow, H. : Involvement of Escherichia coli DNA polymerase I-associated 5 ' ~ 3 ' exonuclease in excision-repair of UV-damaged DNA. In: in Molecular Mechanisms for the Repair of DNA, R.B. Setlow, P.C. Hanawalt, eds., pp. 219-223. NewYork: Plenum Publishing Company 1975 Jacobsen, H., Klenow, H., Overgaard-Hansen, K. : The N-terminal amino-acid sequences of DNA polymerase I from Escherichia coli and of the large and the small fragments obtained by a limited proteolysis. Europ. J. Biochem. 45, 6234527 (1974) Kelley, W.S., Whitfield, H.J. : Purification of an altered DNA polymerase from an E. coli strain with a pol mutation. Nature (Lond.) 230, 33-36 (1971) Klenow, H., Overgaard-Hansen, K., Patkar, S.A.: Proteolytic cleavage of native D N A polymerase into two different catalytic fragments. Influence of assay conditions on the change of exonuclease activity and polymerase activity accompanying cleavage. Europ. J. Biochem. 22, 371-381 (1971) Konrad, E.B., Lehman, I.R.: A conditional lethal mutant of Escherichia coli K 12 defective in the 5 ' ~ 3' exoniaclease associated with DNA polymerase I. Proc. nat. Acad. Sci. (Wash.) 71, 2048-2051 (1974) Kornberg, A. : Active center of DNA polymerase. Science 163, 1410-1418 (1969) Laemmli, U.K. : Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.) 227, 6804585 (1970) Lehman, !.R. : Deoxyribonucleases: Their relationship to deoxyribonucleic acid synthesis. Ann. Rev. Bioch. 36, 6544568 (1967) Lehman, I.R., Chien, J.R. : Persistence of deoxyribonucleic acid polymerase I and its 5'--*3' exonuclease activity in polA mutants of Escherichia coli K12. J. biol. Chem. 248, 7717-7723 (1973) Lehman, I.R., Richardson, C.C.: The deoxyribonucleases of Escherichia coli, IV. An exonuclease activity present in purified preparations of deoxyribonucleic acid polymerase. J. biol. Chem. 239, 233-241 (1964) Lennox, E.S. : Transduction of linked genetic characters of the host by bacteriophage P1. Virology 1, 190-206 (1955) Murray, K. : Nucleotide maps of digests of DNA. Biochem. J. 118, 831-841 (1970) Olivera, B.M., Bonhoeffer, F. : Replication of Escherichia coli requires DNA polymerase I. Nature (Lond.) 250, 513-514 (1974) Reid, M.S., Bielski, R.L. : A simple apparatus for vertical flat sheet gel electrophoresis. Analyt. Biochem. 22, 374-381 (1968) Rosner, J.L. : Formation, induction and curing of bacteriophage P1 lysogens. Virology 48, 6794589 (1972) Setlow, P. : DNA polymerase I from Escherichia coli. In: Methods in enzymology XXIX. L. Grossman, K. Moldave, eds., pp. N 12. New York: Academic Press 1974 Taylor, A.L., Trotter, C.D. : Linkage map of Escherichia coli strain K12. Bact. Rev. 36, 504-524 (1972) Wu, T.T. :~ A model for three-point analysis of random general transduction. Genetics 54, 405410 (1966)

Communicated by B.A. Bridges •Received and accepted March 3, 1976