Studies on the biochemical basis of mutation VI

Studies on the biochemical basis of mutation VI

J. Mol. Biol. (1981) 145, 677-695 Studies on the Biochemical Basis of Mutation VI? Selection and Characterization of a New Bacteriophage T4 Mutator...

1MB Sizes 3 Downloads 76 Views

J. Mol.


(1981) 145, 677-695

Studies on the Biochemical Basis of Mutation VI? Selection and Characterization of a New Bacteriophage T4 Mutator DNA Polymerase LINDA J. REHA-KRANTZ ANDMAURICE J. BESSMAN The McCollum-Pratt Institute and The Department of Biology The Johns Hopkins University, Baltimore, MD 21218, U.S.A. (Received

19 May

1980, and in revised form

30 October


A new mutant of bacteriophage T4 has been isolated by a procedure which was designed to select for mutants with high spontaneous reversion rates. This mutant, M19, induces a defective DNA polymerase which has a degraded specificity and makes errors by inserting the incorrect nucleotide more frequently than the wildtype enzyme. In addition to M19, several other T4 polymerase amber and temperaturesensitive mutants have been located on a linear, fine-scale map. The mutants which most strongly affect mutation rates are found in two clusters at 25% and 80% of the gene. These two domains may represent the active site(s) of the polymerase and exonuclease activities.

1. Introduction The discovery of conditionally lethal point mutations in the polymerase gene which greatly alter mutation frequencies (Speyer et al., 1966; Drake et al., 1969) and the subsequent purification and characterization of the mutationally altered polymerases (Muzyczka et al., 1972) was central to demonstrating the postulated “editing” function of the resident 3’ to 5’.exonuclease activity (Goulian et al., 1968; Kornberg, 1969; Englund, 1971: Brutlag & Kornberg, 1972). A model for DNA polymerase-effected nucleotide misincorporation derived from the studies of these mutant T4 DNA polymerases (Bessman et al., 1974; Goodman et al., 1974) predicts that increased base substitution errors in mutators may be due to reduced polymerase specificity, to an impaired exonuclease “proofreading” function, or to a combination of both. Though there are several examples of T4-induced mutator and antimutator polymerases with altered exonuclease activities (Muzyczka et al., 1972: Lo & Bessman, 1976a,b; Reha-Krantz & Bessman, 1977), only one mutakor with an altered specificity of T4 DNA polymerase has been reported (Hershfield, 1973 ; Gillin & Nossal, 1976). In this paper, we describe a scheme for the selection of mutator mutants in phage T4, the isolation of such a mutant using this procedure, t Paper

V in this series is Bessman

& Keha-Krantz


677 0022~2836/81/040677-19


fc7 1981 Academic

Press Inc.




I,. .I. KEH&KI-t.-ZN’l’Z



and the characterization of this mutator as a defective DNA polymerasr having reduced specificity for the correct nucleotide. We have also localized T4 polymerase (gene 43) amber and temperature-sensitive mutants together on a linear map, and we have found that mutants which most strongly affect mutation rates are grouped in two clusters which may define the polymerase and exonuclease active site(s).

2. Materials and Methods Non-radioactive nueleotides were purchased from P-1,. Riochemicals and radioactive nucleotides from SchwarzMann. 2-[6,3H]aminopurine deoxynucleoside triphosphate was synthesized as described by Bessman et al. (1974).

The preparation of the nucleic acids used in the enzyme assays has been described (RehaKrantz & Bessman, 1977). (c) Enzymes The DNA polymerases induced by the wild-type described previously (Muzyczka et al., 197.2).

(T4D) and mutator (L56) were purified as

(d) Ewyme assuys DN‘A polymerase activity was assayed as the conversion of a radioactive deoxynucleoside triphosphate to an acid-insoluble product. Reactions (63 ml) contained 66 mw-Tris (pH X.8). 16 mr+(NH,),SO,, 10 miw-/I-mercaptoethanol, 6.7 mM-Mgcl,, 170 PM-dATP. dt”TP, and dGTP, 170 p%j3H]dTTP (2 x lo6 to 5 x 10’ cts/min per pmol), 50 pg serum albumin, 1 mM partially digested salmon sperm DNA, and approx. 0.1 to 93 unit of polymerase. Nuclease activity was assayed as the conversion of a radioactive single-stranded DNA to an acidsoluble product. The conditions were similar to the polymerase assay, except that 33 PMalkali-denatured Escherichiu coli r3H]DNA (1 x lo6 to 2 x IO6 cts/min per pmol) was used in place of the partially digested salmon sperm DNA, and no deoxynucleoside triphosphates were present. One unit of polymerase converts 10 nmol of denatured [3H ]DNA to an acidsoluble form in 30 min at 30°C (nuclease), or converts 16 nmol of a deoxgnucleoside triphosphate to an acid-insoluble product in 30 min at 30°C (polymerase). Turnover (DNA-dependent conversion of a deoxynucleoside triphosphatc to t’he corresponding deoxynucleoside monophosphate) was determined by using the standard polymerase assay with partially digested salmon sperm DNA as the template-primer. The deoxynucleosidc monophosphate formed during the incubation was detected after thin-layer chromatography on polyethyleneimine cellulose as described by Hershfield & Nossal (1972). Incorporation was determined in the same assay by removing a separate portion to measure polymerase activity by the disk assay (Muzyczka et al.. 1972). (e) Phagv and bacterial strains T4 gene 43 amber and temperature-sensitive mutants and T4 rlI mutants were generous gifts from ,J. W. Drake (Drake & Allen, 1968: Allen d al.. 1970) and lJ. D. Karam (O’Donnel & Karam, 1972). T4 gene 43/rII and rITA/rlIB double mutants were constructed by genetic



recombination and selection according to the method of Doermann & Double mutants were verified by back-crosses with the parental mutants wild-type recombinants. E. coli strains CR63, CR63(/\h), B, BB, and B4Osul were gifts from II. Preparation of phage stocks, media, and indicator bacteria have been Krantz & Bessman, 1977).

879 Boehner (1970). which yielded no Berger. described (Reha-

(f) Phage crosses For the mapping studies, the procedure followed was a modification of that of Krish et al. (1972). Five minutes prior to infection, E. coli CR63 growing exponentially at 30°C were diluted to 2 x 10s cells/ml in M-9 medium (Adams, 1959) containing 5 g Difco Casamino acids/l (M-9+), 2 mM-KCN. The culture was then infected with a multiplicity of 5 of each parental phage per bacterium and gently aerated. After 10 min, a portion was taken to assay unadsorbed phage, and T4-spe$ific antiserum was added in order to neutralize any unadsorbed phage in the infected culture. At 15 min post infection, the culture was diluted 5000-fold into the growth tube and an additional dilution was made to determine infected bacteria. After 100 min, chloroform was added in order to complete lysis, and the progeny were titered on E. coli CR63 at 30°C to determine total progeny, and on E. coli B at 43°C to determine wild-type recombinants. Burst sizes were calculated by dividing the total progeny by the number of infected bacteria and these ranged from 150 to 250. Recombination frequencies are expressed as a doubling of the percentage of wild-type recombinants. (g) Measurement of rii+



The reversion frequency of an rII marker was measured after several cycles of phage growth. Parallel stocks (10 ml) of E. coli BB or B40sul at 2 x 10’ cells/ml were inoculated with individual plaques of gene 43/rUV199 double mutants and aerated overnight at 30°C. Chloroform was added in order to complete lysis and to sterilize the cultures. Progeny phage were titered on CR63 at 30°C and averaged 4 x 10” to 6 x 10” phage/ml. The rII+ revertants were determined on CR63(hh) at 30°C. Four parallel cultures were grown to determine each reversion frequency and the numbers reported represent the median values of each set. (h) Chemical The reversion frequency of the aminopurine mutagenesis. Phage (r5 mg 2-aminopurine/ml (Sigma order to complete lysis. The total number of double rII+ revertants

induced mutagenesis

double ~11 mutant, rUV199/rUV183, was increased by 2were grown for a single replicative cycle in the presence of Chemical Co.) in M-9+ medium. Chloroform was added in phage yields were determined by plating on CR63, and the were scored by plating on CR63(hh).

3. Results (a) Mutator


An initial goal of these studies was to develop a simple procedure to isolate mutator mutants of T4 bacteriophage which were defective in DNA polymerase. A procedure has been developed which is based on the reversion of a double mutant, to wild-type, under the premise that strong mutators would be more likely to revert a double mutant to a detectable level. The ~11 system described by Benzer (1955) was chosen for these studies because of its high sensitivity and selectivity. As few as one: revertant in 109 phage can be detected.





The double rIIA/rIlB mutant, rUV199/rU\‘l83, was chosen as the tester strain because both of these individual ~11 mutants are known to be sensitive to mutant polymerase backgrounds (Drake et al., 1969: Reha-Krantz bz Bessman, 1977). Since the independent reversion frequency of each rII locus is about 1 x 10e6. a spontaneous reversion frequency of 1 X IO- l2 would be expected for the double mutant, which is below the level of detection in this syst’em. The sensitivit’y was expanded by growing the phage in 0.5 mg 2-aminopurine/ml. Under these conditions, the independent reversion frequencies of these loci are 3 x lO-5 to (Bessman & Reha-Krantz, 1977) leading to an expected reversion 6 x IO-’ frequency of about 1 x 1W9 for the double mutant. Approximately five revertants per 10’ phage were actually detected. The revertants were then assayed for temperature sensitivity on replica-plates (Doermann & Boehner. 1970) incubated at 30°C and 43°C”. Examination of 436 rII+ revertants yielded 19 temperaturesensitive isolates, one of which, “Mlg”, was localized in the polymerase gene, gene 43, by complementation assays which followed the same procedure as for phage crosses except that the host was E. coli B and the temperature was 43°C’. Progeny were titered on CR63 at 30°C. Additional mapping by genetic recombination confirmed the location of this mutant (see Appendix). The other 1X tst isolates, which were not located in gene 13 have not been characterized furthrr. The reversion frequency to ts+ was approximately 4 x 10Y5 which suggests a single point mutat’ion at the Ml9 site. The spontaneous mutation frequency of Ml9 was then determined by measuring the reversion frequency of rUV199 which we introduced into an Ml9 background (M19/rUV199). The results indicate that Ml9 is a strong mutator with a reversion frequency similar to L56, the most powerful mutator thus far studied (Table 1). In from a population of approximately IO” phage, one mutator summary, polymerase mutant was easily identified with this selection procedure.


Reversion frequencies Phage

of wild-type


T4D/rUT’I99 L.%/rVV199 MlS/GVI99


and polymeraer


Wild-type Mutator Mutator


mutatoy mutants Relative

70 .5630 3x50

(1) 80 55

Reversion frequencies were measured at, 30°C in E. coli BB as dewribrd

(b) Yur$cation~

of the Ml.9

reversion frequencies

in Materials

and Methods.


polymerase was partially purified as described by Muzyczka et (1969). This procedure removes all other polymerase and activities. Briefly, an exponentially growing culture of E. coli CR63

The Ml9 induced

aZ. (1972) and by Nossal


t AbbreviaCons used: ta. temperature-sensitive: monophosphates of Saminopurine; nm, polypeptide

dAPTP and dAPMP, deoxynucleouide chain terminating amber mutants.

tri- and








Fraction I II III

Volume (ml) 36 80 64

Enzyme concentration (units/ml) 7.5 0.0 31.0

of the Ml9 polymerme

Total activity (units) 270 480 214

Protein concentration (m&4 14 O-18 <@Ol

SpeCifiP ’ activity (units/mg) 1.X 33.3 >[email protected]

Relative purification (1W 185 > 1700.0

Recovery (yO) 100 178 79

Activity was measured in the standard polymerase assay using partially digested salmon sperm DNA as the template primer. One unit of enzyme converts 10 nmol of dTTP to an acid-insoluble product in 30 min at, 30°C. Fraction I is the dialyzed, viscous cell extract produced from alumina ground cells and centrifuged to remove cell debris and alumina. Fraction II contains the peak tubes of polymerase from the DEAE-cellulose column. Fraction III is the dialyzed pool of peak tubes of polymerase activity from the phosphocellulose column. Protein was measured by the method of Lowry el al. (1951). The protein ronrentrat,ion of fraction III was below the level of detection.

(30 1) was infected at 25°C at a multiplicity of five. Infection continued for 30 minutes and then the culture was chilled on ice. About 60 g of cells were collected by centrifugation. These were ground with alumina (2 g/lg of cells), extracted with buffer (250/b (v/v) glycerol, 50 mM-Tris, pH 7.4, (b5 mi%-fl-mercaptoethanol, 10 mMMgCl,, 1 mM-EDTA), and centrifuged to remove alumina and cellular debris (fraction I, Table 2). Fraction I was loaded onto a DEAE-cellulose column in the above buffer (minus MgCI,) and eluted with a linear gradient of KCI. Peak fractions were pooled (fraction II, Table 2). Fraction II was further chromatographed on a column of phosphocellulose equilibrated with 50 mM-potassium phosphate, (pH 6.5), 1 rnM$-mercaptoethanol, 25% glycerol. The enzyme was eluted with a linear gradient of 50 mM to 500 mw-potassium phosphate (pH 6.5). The polymerase and exonuclease activities co-chromatographed and the peak fractions of enzymatic activity eluting at about 200 rnh-potassium phosphate were pooled and then dialyzed against the starting buffer containing 50% glycerol and no MgCl, (fraction III, Table 2). Fraction III was used for all subsequent studies. (c)

Heat inactivation

of the

Ml9 polymerase

The Ml9 mutant is temperature-sensitive in vivo with little growth at 43”C, and the purified Ml9 polymerasi: is also temperature-sensitive as shown in Figure 1. There is simultaneous and equal loss of both polymerase and nuclease activities at, 50°C. The non-interaction of the heat inactivation curves for a mixture of wild-type and Ml9 enzymes indicates that protein concentration or other factors present in the preparations are not responsible for the difference in heat stability of the wildtype and the Ml9 enzymes. (d) Nuclease-to-polymerase



The ratio of the exonuclease activity to polymerase activity, assayed under optimal conditions for each, has proven to be useful in understanding the mutator

I,. .I.






Tme at 50°C hn) FIG. 1. Temperature sensitivit,y of M19. Equal units incubated at, 50°C from 0 to 30 min. Samples were removed standard polymerase or nuclease assay reaction mixtures After all samples were c*ollected, the radioactive substrate incubated at 30°C for 30 min. In one experiment, equal incubated and assayed together. The curves were drawn (0) Polymerase artivit,y: (0) nuclease activity.

of purified Ml9 and T4D polymerases were at Smin int,ervals and added to chilled. 1‘C. less [‘H IdTTP or denatured E. colt’ 13H]DNA. was added and the reaction mixtures wercb units of T4D and Ml9 polymerawa were preto fit the da,ta points for polpmerasr activity.

and antimutator phenotypes of certain mutant’ polymerases. Low exonucleasr activity and a low nuclease-to-polymerase activity ratio hare been found in a number of mutator polymerases, and a high nucleasr-to-polymerase activity ratio has been found in antimutator polymerases (Muzyezka et al., 1972). A similar analysis of this mutator, however. indicates a higher nuclease-to-polymerase ratio for Ml9 compared to the wild-t’ype enzyme (Table 3). The relatively high exonuclease activity of Ml9 in contrast to the low nuclease activities reported for other mutator polymerases suggested that the mutator phenotype of Ml9 is not due to a reduced editing function.

T.ABLE 3 Sucleasp-to-polym~rase Strain ‘1’4 I) Ml9

Phenot,ype \2’il&typr Mutator

ratios Nuc,lease/polytnerase


059fW16 0.93 * 0.16

Polymerase and nuclease activities were measured under standard conditions at 30°C for 30 min using 02 or 0.1 polymerase unit of wild-type and mutant. respectively. Partially digested salmon sperm DNA was the template primer for the polymerase assays. Alkali-denatured [‘HjDNA was the substrat,e for the nuclease assays. The nucleate-to-polpmerase rat,iov are the mean values from 3 experiments.






(e) Incorporation and turnover of the base analogue, Z-aminopurine We have shown that dAPTP may be used to analyze various aspects of DNA polymerase activity, and that the amount of dAPTP incorporated into DNA in place of dATP can be related to the polymerase and exonuclease functions of variant enzymes (Bessman et al., 1974). We have used this technique in the following experiments to compare the polymerase molecule of Ml9 to a well-studied mutator, L56, and to the wild-type T4D. In order to facilitate a direct comparison of the enzymatic properties of M19, L56, and T4D, the polymerase activities were adjusted so that the incorporation ofdAPMP was approximately equal for all three enzymes and this amount is used in all subsequent experiments. As shown in Figure 2(a), although L56 incorporates dAPMP to a slightly greater extent than


Time (mid

Time (mln)



Flu. 2. Incorporation (a) and turnover (b) of dAPTP. Incorporation and turnover of dAPTP were determined in @3 ml reaction mixtures containing 66 miwTris (pH 8+3), 16 mw(NH,),SO,, 10 mw8. 33~~ each of dCTP, dTTP, and dGTP. 33 PM-2-[‘H]dAPTP merraptoethanol. 6.7 mwMgCl,, (9.2 x IO6 cts/min per pmol), 50 pg serum albumin, 1 mwpartially digested salmon sperm DNA, O? to 0.3 unit of each enzyme (fraction III): (0) T4D; (A) M19: ( x ) L56. Samples (20 ~1) were removed at lomin intervals to determine incorporation and turnover as described in Materials and Methods. Values are reported for a @3 ml reaction volume.

either Ml9 or the wild-type enzyme it removes or turns over dAPMP much less than Ml9 or T4D (Fig. 2(b)). The turnover of dAPMP is due to 3’-exonuclease activity of the enzyme. This reduced proofreading function which has been correlated with the mutator phenotype of L56 (Muzyczka et al., 1972) is reproduced here. Yet, M19, also a strong mutator, shows a surprisingly high turnover of dAPMP, higher even than the wild-type, T4D, and much higher than L56. Thus the high mutation rate in Ml9 cannot be due to an impaired exonuclease proofreading function. 23



(f) Competition




dATP a,nd dAP77’

In the previous experiments, dATP or dAPTP was used in separate incubations to compare various aspects of the mutant and wild-type enzymes. In the following experiments, dATP and dAPTP were present in bhe same incubation in order to assess the ability of the individual enzymes to discriminate between t,he correct base, adenine, and the incorrect base 2-aminopurine. For the experiments described in Figure 3, the activities of the enzymes were adjusted so that the rates of dATP

o-2 IO






Time (mm) (a)


FIG. 3. Incorporation and turnover of dATP and dAPTP by Ml9 (a) anti T4D (b). Incorporation and turnover of dATP and dAPTP were determined under the same conditions as described in the legend to Fig. 2 for dAPTP, except that parallel reaction mixtures were used with either 33 pM-1 14CjdATt’ (1.8 x 10’ cts/min per pmol) plus non-radioactive dAPTP (33 (&I): or 33 ~LM-%[~H]~APTP (92 x IO6 cts/min per pmol) plus nonradioar:t,ive dATP (33 FM). (A) dATP incorporation: (a) dATP turnover: (0) dAPTP incorporation: (0) dAPTP turnover.

incorporation were similar for Ml9 and T4D. As can be seen, the turnover of dATP was also similar and so the wild-type and mutant, enzymes respond to dATP in the same manner. However, their interaction with the analogue, dAPTP, is quite different. For example, Ml9 incorporates more of the analogue, and what is most striking, it turns over much more of the dAPTP than the wild-type enzyme. The high turnover of dAPTP coupled with its high incorporation implies that Ml9 actually inserts dAPTP into DNA at a high rate relative to wild-type.? That this is indeed the case is demonstrated in Figure 4, where the insertion (incorporation plus t This interpretation assumes no hydrolysis of the triphosphate to monophosphate prior to t,he polymerization of the incoming nucleotide. The hydrolysis of the triphosphate is absolutely dependent on DNA. but the requirement for a covalent linkage between the incoming triphosphate and the 3’. hydroxyl of the DNA prior to hydrolysis has not been demonstrated unequivocally.













Time (mm)

FIG. 4. SAminopurine insertion in the presence of dATP. Insertion is defined as the sum of incorporation plus turnover. Data for Ml9 and T4D are from Fig. 3, and the data for L56 are from RehsKrantz & Bessman (1977) and unpublished results. (0) T4D; (A) M19; ( x ) L56.

turnover) is plotted for T4D, M19, and the well-known mutator, L56. In contrast to L56, which has previously been shown to insert or utilize 2-aminopurine to the same extent as T4D, but to have a much lower turnover or removal frequency (Bessman et al., 1974; Reha-Krantz & Bessman, 1977), Ml9 has a markedly increased insertion of the base analogue. Therefore, although Ml9 and L56 share similar mutator phenotypes, the biochemical defects of the two enzymes are apparently quite different. The incorporation and turnover data for the mutants and the wild-type enzymes may be compared in Table 4. The most striking difference is that the Z(AP), which is a measure of the ability of the polymerase to discriminate between the two bases, independent of nuclease activity, is nearly threefold higher for the Ml9 polymerase. The other parameters including the removal frequencies R(AP) of dAPMP and R(A) of dAMP under synthesizing conditions, which are a measure of the magnitude of nuclease activity, and the removal probability, R(AP)/R(A), which is a measure of nuclease specificity, vary only slightly. The probability of finding a 2aminopurine in place of adenine in the final DNA product, P(AP), is higher for the mutator polymerase, and is apparently due to the high misinsertion rate. It has been shown in a number of systems in vitro that dAPTP competes with dATP for sites opposite template thymine (Cerami, 1967 : Rogan & Bessman, 1970: Clayton et al., 1979). This competition has been observed here also, but the analogue, dAPTP is more competitive with Ml9 than it is with T4D. This is demonstrated most clearly in Table 5 where the effect of adding dATP is directly compared to incubations omitting dATP. It may be seen that in respect to incorporation, turnover, and insertion, dAPTP is less affected by the presence of dATP in the mutant, M19, than in T4D. Thus Ml9 distinguishes less between dAPTP and dATP than does the wild-type enzyme.







adenine by Ml9




and uild-typr



T-l IlE.-t



f(AP) (“0)

tl(Al-‘) (“0)

/(‘A) (“,,I

K(A) (” 0 )





11 29

60 79

x9 71

23 22


polymwasw K(AP)/K(A)

2.0 34

/‘(.\I’) f” 0 1 fi


1(.4P) and I(A) are the average insertion f’wquencios of %-aminopurinr and adrnine : H(AP) and H(A) are the average removal frequencies of %aminopurine and adeninr: I’(AP) is thr average inrwrporatiou frequenry of 2-aminopurine. The insertion and removal frequencies were c&ulatetl lkom the data. in Fig. 3 as described by Besxman PI nl. (1974) awording to the following formulae: f(AP)




mo(AP)+ turn(AP)+

inc(A) + turn(A)



mc(AP + turn(AP) ’


ino mc(AP) + inc(A)

where incx = incor~oraliwc of the nwleotide into DNA and turn = ~trrwwr of the triphosphate to the monophosphate. The values for I(A) and R(A) were valrulated using the same equations by subst,ituting A for AP where appropriate.

Effect of dATP

(a) -BTP

Incorporation (1)) + ATP


TABLE 5 on dAPT1’



(nmol) T4D






2.8 1.9

0.97 1.65

Turnover (b)

+ ATP (nmol) 0.10 0.51



9.7 3.2


-ATP 1.19 193

Insertion (b)

+ ATP (nmol) 0.18 0456




The values are derived from the 20.min time points of’ Pigs 2 and 3.

(g) Localizing



in relatiori

to other T1


A genetic map has been constructed relating the position of Ml9 to six ot#her ts and 11 awL mutants of the T4 polymerase gene (see Appendix). This represent’s t’he first attempt to localize both bhe am and ts mutants encompassing most of gene 13 on the same genetic map. (An earlier study oriented some am and ts mutants at the ends of the gene 33 map (Allen et al.! 1970).) It is interesting to note that although ts mutants (lot. cit.) and am mutants are found throughout the polymerase gene. those which most strongly influence mutation rates are clustered in two regions at approximately 25o/o and 8Oo/o of the cistron (Fig. 5). (Map locations ha,ve been estimated as percentages from the marker amE being O%.) The cluster at.




Gene 43 temperature-sensltlve




Gene 43 amber mutants

PIG:. 5. Linear map of’ gene 43 temperature-sensitive and amber mutants. indicated by the symbol @ and strong antimutators by the symbol 0.

Strong mutators


about. 250/b is composed entirely of the strong mutators, tsL56, tsL98, and the new mutant, tsM19. Although these mutants share a similar mutator phenotype in ni~o. the biochemical properties of their polymerases differ. The tsL56 and tsL98 mutants are mutators due to reduced exonuclease, proofreading, activity (Bessman et al.. 1974), while the tsM19 mutant has wild-type levels of exonuclease activity, but has reduced insertion specificity (Fig. 4 and Table 4). The other cluster of mutants which strongly influences mutation rates is found at about 80% of the gene, and it is composed of both ts and am mutants. This cluster contains the strong antimutators tsL141 and tsL42 as well as the mutator tsL88. The antimutators have higher levels of exonuclease activity than the wild-type enzyme (Muzyczka et al., 1972), while tsL88 is reported to have reduced polymerase specificity (Hershfield, 1973). Some of the suppressed am mutants also profoundly influence mutation rates. It is interesting that no strong am mutators have been observed (Xlikhanian et al., 1974: L. J. Reha-Krantz & M. J. Bessman, unpublished results), but powerful antimutators, produced in a variety of suppressor bacteria, have been found. The amB22 mutation gives rise to antimutator polymerase in SICI or s/r3 hosts (Reha-Krantz & Bessman, 1977). The amE4302, amCl25, and amE mutations express strong antimutator phenotypes under sul, su2, and ~113 suplm~ssion (Appendix Fig. 7 : L. CJ.R&a-Krantz B M. J. Bessman, unpublished results). =\ correlation of biochemical properties and effects on mutation rates of certain DNA polymerase mutants with their map location has been reported by Allen et al. (1970). They suggest that the strong antimutators, tsL42 and taLl41, and thr strong mutators, tsL88 and tsL56, cluster in a region 60 %, and 750,b of the cistron on their map. The significance of the apparent clusters awaits the isolation and characterization of additional gene 43 mutants and more extensive mapping studies.




4. Discussion The selection for and characterization of a powerful T4 mutator polymerase has been described. Although this mutator polymerase selection procedure has not been extensively tested, there is reason to believe that it will be useful in selecting polymerase mutants exhibiting reduced fidelity in DNA synthesis. Similar techniques which also rely on multiple mutations have been described in other systems (Cross et aZ., 1968; Siegel & Bryson, 1963). Most of the T4 polymerase mutants previously studied have either reduced or enhanced 3’ to ti’-exonuclease proofreading activity. and models have been advanced which explain incorporation errors in DNA synthesis as a consequence of the interplay of polymerase specificity and exonuclease proofreading activity (Goodman et al., 1974: Galas & Branscomb, 197X: Clayton et al., 1979). The mutator polymerases incorporate significantly increased amounts of the analogue, dAPMP. in place of dAMP. Compared to the wild-type, the antimutator polymerases. conversely, incorporate less of the base analogue relat’ive to the wild-type enzyme (Bessman et aZ., 1974: Reha-Krantz h Bessman, 1977). In all cases. except for tsLXX (Hershfield, 1973: Gillin & Nossal, 1976), the error frequency may be correlated with the relative proofreading and polymerase activities. However. the total number of mutants available for study is relatively small. Using the selection technique described above we have isolated an interesting mutant that makes mistakes because of a reduced specificity, not proofreading capacity, and it will be of interest to examine this mutant in terms of the models’ predictions. More important, however, will be the isolation and analysis of a larger number of mutants so that the generality of the models can be tested. Another observation which would benefit from a larger collection of mutants is the clustering of mutators and antimutators in two regions of gene 33. Three of the four strong mutators are clustered at the 2576 region whereas all of the strong antimutators are found around the SOo/o region. Is this coincidence or are these two regions especially important for the proper functioning of the enzyme! One possibility is that these two domains of the polypeptide chain, although apparently far apart on the linear map. are in fact, close together in the folded threedimensional structure and make up the catalytic site of the enzyme. The bifunctional nature of this enzyme. and the availability of a large number of variants should provide an interesting subject, for studies of struct~~r~~~f~lrl(~tion relationships.



Gene 43 Amber and Temperature-sensitive

Mutants There are many problems encountered in mapping gene -I3 mutants. These include the high reversion frequencies of some of the mutators. the alteration of recombination frequencies by some polymerase mutants (Berger et al., 1969). and






the “leakiness” or partial functioning of the mutants at both the permissive and restrictive temperatures. Most of these problems have been minimized in this study by using three and four-factor crosses to localize a third mutant to either an internal or outside position relative to a double mutant. The results of a number of crosses and the order deduced therefrom is presented in a series of four Tables. The first, Table 6, established the order of tsL88, amB22, amE304, and tsL56, and

6 gene 43 amber and temperature-sensitive mutants TABLE

Mapping Ci-OSS

lsL56 k-L56 lSL56 amB22 amB2.’ I amB22 amE kL88 tsL8S

laL88 tsL88-am,B22 tsL88-amB22 amB2StsL56

x x x x x x

x x x x x x x

amE tsL88 tsL88-amE amE tsL56 amE4304tsL56 amB22-tsL56 amE amE4304-tsL56 amB22-tsL56 tsL56 amE4304-BL56 tCyL88-a,mE4304

Recombination frequency (%) 1.4 4.7 1.4 2.6 4.8 2.4 0.4 56 4.7 0.6 3.6 2.8 0.16

Order deduced






amB22 amB22





amE amE

kL56 tnL56

The recombination frequency is a doubling of the ratio of wild-type recombinants to total progeny. The average burst size was 150 to 250 for this experiment and also for the experiments reported in Tables 7 t,o 9.

represents the core of the map. The next three Tables (Tables 7 to 9) establish the order of mutants on either side of the core and the order of mutants within the core, It was not possible to determine the order of amB22, amC125, and amE unambiguously. These mutants are separated by only @05 to 0.15 map unit. The map distances between mutants are presented in Figure 6 and represent an average of three to seven crosses. The gene covers about 10 map units from end to end. According to the mapping function of Stahl et al. (1964), this corresponds to a coding capacity for a protein of about lOO,O00 M,, which agrees well with the reported molecular weight for T4 polymerase of about 108,000 (Goulian et al., 1968 ; Nossal & Hershfield, 1971). Thus, mutants in this study encompass most of the T4 polymerase gene. The order of am and ts mutants presented in Figure 6 agrees well with the separate am maps and ts maps published by Allen et al. (1970), the am map of O’Donnel & Karam (1972), and the physical map of Huang & Lehman (1972). The only major discrepancy is the position of the strong mutator, tsL56. Allen et al. (I 970) place tsL56 near tsL88 and tsL141, while our mapping studies position tsL56 near tsL98 and the new mutator, tsM19. The positioning of gene 43 between 42 and 62 is based on the previous mapping studies of Allen et al. (1970).

M Y E e






x x x x x x x x x x

x x x x x

amE lsL42 tsG40 amB22 hLI41

tsG40 amB22 lsL141 hL88 amE

kL88 amE hL42 tsG40 amB22

tsL42-amE tsL42-amE tsL42-amE taL42-amE tsL42-amE

tsL141-amE tsL141-amE tsL141-amE kL141-amE tsL141-amE tsL88-amE tsL88-amE tsL88-amE tsL88-amE tsL88-amE 0.2 0.1 02 0.3


I.0 1.6 3.0 0.3 0.2

3.6 @OS

1.5 1.6 1.8

1.3 0.6


0.3 2.0

@I 1.4

x x x x x x x

and322 amB22 amE4335.amE amE4335-amE taL88-amB22 tsL88-amB22 amB22-tsL56

1.34 1.31 1.20

x amE x amE x amE

amB22 amC125 amE tsLl41 bL42 ksLl41 lsL42 6L141 lsL42 hL88

om6 om2 om6

x amB22-amE x amB22-amE x amE4335.amE










gene 43 amber and temperature-sensitiw

Recombination frequency (“4)

amE amC125 amB22








Order deduced






g % w E c

;o = % w E 3


(Gene 42) --


-- (Gene 62)

FIG. 6. Linear map of gene 43 temperature-sensitive and amber mutants. The recombination values (in percent) for each cross are the doubling of the ratio of the number of wild-type progeny detected by plating on E. eoli B at 43°C to the total number of progeny phage, obtained by titering on CR63 at 30°C.

I -

, -


I -

/ -


FIG. 7. The relative increase (mutators) and the relative decrease (antimutators) in reversion frequencies in host E. coli B40.9~1 was determined for 11 amber mutants and for 7 temperature sensitive mutants. The procedure is described in Materials and Methods.




The mutation frequencies of the am and ts mutants were determined in E:. co/i strain B40sul at 3W’. which is permissive for all of t,he gene 13 mutants. Tht> increase in mutation frequencies for the mutator mutants and the decrease ilr mutation frequencies for the antimutators is plotted in Figure 7. Two regions at about, 25% and S(F)& of the gene contain mutations which most strongly affect mutation frequencies. This is contribution no. 1075 from the Mc(‘ollum~Pratt Institute. This work was supported by grant no. GM18649 from the National Instkutes of Health. Special thanks go to Mark J. Krantz for a careful reading and editing of the manuscript. and to Lynn Kiflfcy. John W. Drake and Nancy G. Nossal for t,heir interest, and suggestions. Dorot,hy Regula bras responsible for the preparation of this manuscript.

REFERENVES Adams, M. H. (1959). In Bacteriophages, p. 445. Interscienrr Publishers, New York. Alikhanian. S. I., Piruzian, N. 8. & Kobetz, N. S. (1974). Mol. &)I. Nevwt. 130, 327.. 331. Allen, E. F.? Albrecht, I. di Drake. J. W. (1970). Genetics, 65, 187-200. Benzer, S. (1955). Proc. Nat. Acad. Sci., I:.S.A. 41, 344-354. Berger, H. A., Warren, A. J. & Fry, K. E. (1969). J. 1.irol. 3, 17I-- 175. Bessman, M. J. & Reha-Krantz, L. J. (1977). J. Mol. Riol. 116, 115-123. Bessman, M. J., Muzyczka, N., Goodman, M. F. & Schnaar. R. I,. (1974). J. Mol. Biol. 88. 409-42 1. Brutlag, D. & Kornberg, A. (1972). J. Riol. (‘helrc. 247. 241 248. Cerami, A. (1967). Ph.D. thesis, Rockefeller Unircarsit,y. Xew York. Clayton, L. K., Goodman, M. F., Branscomb. E. W. & (ialas. D. ,J. (1979). J. Niol. (‘htnr. 254, 1902-1912. Doermann, A. H. & Boehner, L. (1970). (/~//vtjcs, 66, 117-428. Drake, J. W. & Allen, E. F. (1968). (‘old Spring Hnrhor ~Syvnp. Qctant. Wiol. 33. 339~-344. Drake, J. W., Allen, E. F., For&erg. S. A.. Preparata. R. 11. & (ireening. E. 0. (1969). Satrue (London), 221, 1128-l 131. Englund, I’. T. (1971). J. Riol. Chum. 246, 6684-5687. Galas, D. 8: Branscomb, E. W. (1978). J. Mol. Riol. 124, 653487. Gillin, F. D. & Nossal, N. G. (1976). J. Hiol. Chem. 251, 5225-6232. Goodman, M. F., (iore’, W. C’., Muzyczka, N. & Bessman, M. .I. (1974). J. :IIol. Riol. 88. ti3435. Goulian, M. Z., Lucas. ,J. $ Kornberg, .I. (1968). J. Hiol. f’h.rnL. 243, 627 -63X. (:ross, J. D., Karamata. D. & Hempstead: P. 1:. (1968). (‘t&i Sprit?.g Hnrttor Syrn/~. Qw//f. Riol. 33, 307-312. Hershfield, M. S. (1973). J. Rio/. (‘hpm. 248, 1417-1523. Hershfield, M. S. & Nossal, X. (:. (1972). J. Rio/. Chem. 247, 3393-3104. Huang, W. H. Pr Lehman, I. R. (1972). J. Kiol. (‘hum. 247. 7663-7667. Kornberg, A. (1969). Scirnce; 163, 1410-1118. Krish, H. M., Hamlett, N. V. & Bergrr, H. (1972). C;rwtics, 72, 1X7 203. Lo, K. 8 Bessman; M. .l. (1976a). .I. Rio/. (‘hem. 251. 2375-2479. Lo. K. Br Bessman, M. J. (1976b). J. Hiol. (‘hem. 251: 248+2486. Lowry, 0. H., Rosebrough. S. J., Farr. A. L. & Randall. R. ,J. (1951). .I. /liol. (‘hum. 193. 215-269. Muzyczka. N.. Poland. R. b;- Bessman, M. .J. (1972). .J. /lid. t’hrm. 247, 71 l(i~-7122. Nossaf, ri. U. (1969). J. Hiol. C’hpm. 244, 218-220. Sossal: pi. G. CuHershfield. M. S. (1971). .J. Jjiol. t’h,un. 246, +541-C-5426. O’Donnel: P. V. & Karam. J. D. (1972). ,I. I:irot. 9, 99cC-998.



Reha-Krantz, L. J. & Bessman, M. J. (1977). J. Mol. Biol. 116, 99-113. Rogan, E. G. & Bessman, M. J. (1970). J. Racteriol. 103, 622-633. Siegel, E. C. & Bryson, V. (1963). In Antimicrobial Agentsand Chemotherapy (Sylvester, J. C., ed.), Proc. of the Third I~nterscience Conference on Antimicrobial Age& and Chemotherapy, Washington, D.C., pp. 629-634, Braun-Brumfield, Inc., Ann Arbor, Michigan. Speyer, J. F., Karam, J. D. & Lenny, A. B. (1966). Cold Spring Harbor Symp. Quant. Riol. 31, 693497. Stahl. F. W., Edgar, R. S. & Steinberg, J. (1964). Genetics, 50, 539-552.