J. Mol. Biol. (1977) 116, 115-123
Studies on the Biochemical Basis of Spontaneous Mutation V.t Effect of Temperature on Mutation MAURICE
The M&o&m-Pratt The Johns Hopkins
Institute and Department of Biology University, Baltimore, Md 21218, U.X.A.
(Received 6 December 1976, and in revised form 1 June 1977) Temperature, in the range of 15°C to 4O”C, has a pronounced effect on the incorporation of 2-aminopurine deoxynucleotides into DNA by purified bacteriophage T4-induced DNA polymerase. Whereas the total rate of utilization of the 2aminopurine deoxynucleoside triphosphate increases with increasing temperature, a greater proportion is converted to the monophosphate by the editing nuclease of the enzyme. Therefore, the amount of analogue incorporated goes through n maximum and then decreases with increasing temperature. These results, obtained in vitro, have been correlated with effects of temperature on a-aminopurine induced and spontaneous mutation rates of several ~-11 markers, and they have been generalized to an hypothesis which holds that the stability of the helix immediately preceding the incoming nucleotide is an important factor in determining the accuracy of DNA replication. We suggest that there is a higher probability of making errors via base substitutions in a more stable (G + C-rich) rather than a less stable (A +- T-rich) microenvironment.
1. Introduction During the course of our studies on the synthesis of DNA in vitro by several mutant DNA polymerases, we found that the deoxynucleoside triphosphate of 2-aminopurine was a useful probe in measuring the fidelity of DNA synthesis (Schnaar et al., 1973; Bessman et al., 1974). Recently we have been developing a simplified procedure for comparing the accuracy of DNA synthesis in crude extracts of tissues. It is based on the ability of the extract to discriminate between deoxyadenosine triphosphate and 2-aminopurine deoxynucleoside triphosphate. Investigation of the factors which influence the discrimination between adenine and 2-aminopurine has revealed that temperature strongly affects the incorporation of 2-aminopurine into DNA. This paper describes some of these experiments and suggests a nexus between these observations and the non-randomness of spontaneous point mutations.
2. Materials and Methods The sources of reagents and bacterial and phage strains and the procedures described in the accompanying paper (Reha-Krantz & Bessman, 1977). t Paper IV in this series is Reha-Krantz & Bessman (1977). $ Present address: Institut de Reoherches Cliniques, 110 W. Pine Avenue, Canada. 115
ICT. J. BESSMAN
I,. J. ItEHA-KRANTZ
3. Results (a) Effect of temperature on the incorporation and twrnover of 2-aminopurine awl adenine nucleotides In a preliminary experiment (not shown) we were surprised to observe that both the rate and yield of 2-aminopurine incorporation into DPiA were decreased when was raised from 30°C to 40°C. A more detailed examination of the the temperature effect of temperature on the rate of incorporation is shown in Figure l(a). The rate of incorporation of 2-aminopurine increased as the temperature was raised from 15°C to 30°C but decreased at 37°C. This was not true for the turnover (hydrolysis) of 2-aminopurine deoxynucleoside triphosphate, that is the DNA-dependent conversion of the triphosphate to the monophosphate. Figure l(a) clearly shows that turnover continued to increase as the temperature increased, excluding the possibility that the enzyme was being inactivated at the higher temperature. The total quantity of triphosphate utilized at any temperature may be calculated as the sum of the nucleotide incorporated plus that which is turned over. These values, shown in Figure l(b),
I.0 c = z 5 P Fo L “,
“, 0.5E ‘0 +j 0,4s 0.3-
Fm. 1. Effect of temperature on the utilization of 2.aminopurine deoxynucleoside triphosphate. The reaction mixtures contained, in 0.45 ml: 66 mM-Tris (pH 8.8), 16 mM-(NH,),S04, 10 miw/3-mercaptoethanol, 6.7 mM-MgCl,, 33 PM each dCTP, dGTP and dTTP, 33 PM-[sH]%aminopurine deoxynucleosi& triphosphete (9.17 x lo7 cts/min per pmol), 76 pg serum albumin, 0.8 mMalkali-denatured salmon sperm DNA and approx. 1.2 units of polymerase (T4D, fraction III). After 30 min, samples were removed for the determination of incorporation and turnover of 2-aminopurine deoxynucleoside triphosphate. (b) Nucleotide utilized refers to the sum of the nucleotide hydrolyzed and the nucleotide incorporated. (a) Nucleotide hydrolyzed, -O-O---; nucleotide incorporated, -X-X -. (b) Nucleotide utilized, [email protected]
; nucleotide incorporated, -a-a--.
indicate that the total utilization of triphosphate increased over the entire temperature range. The fraction of 2-aminopurine of the total utilized, which actually ended up in DNA is also depicted in Figure l(b). This fraction decreased markedly as the temperature was increased. An especially interesting comparison is seen in Figure 2. Under identical conditions, the incorporation of adenine into DNA at two different temperatures was strikingly different’ from its analogue, 2-aminopurine. Whereas the rate of incorporation of
Adenlne 38”C,) / / / / - Cl.30
FIG. 2. The incorporation of deoxyadenylic acid and P-aminopurine deoxynucleotide into DNA at 30°C. Reaction mixtures were in a final volume of 0.15 ml and contained the same concentrations of reagents as in Fig. 1. However, [‘%]dATP replaced [3H]2-aminopurine deoxynucleotide triphosphate where indicated.
the analogue decreasedat 38”C, the rate of adenine incorporation increased over that at 30°C. The selective decrease in the incorporation of 2-aminopurine at the higher temperature suggested the following experiment on the mutagenicity of 2-aminopurine in vivo. (b) Mutagenicity of 2-aminopurine at 30°C and at 43°C The effect of temperature on the mutagenicity of 2-aminopurine was tested at 30°C and 43°C by measuring the reversion frequency of a specific ~11 mutant. T4D/ UV199, which is revertible by 2-aminopurine, was grown for a single growth cycle in 0 to I.0 mg of 2-aminopurine per ml at both 30°C and 43°C. As seen in Figure 3, the number of rII.+ revertants was proportional to the concentration of 2-aminopurine at both temperatures, but there was a fourfold decrease in the proportion of revertants at the higher temperature. The result of changing the temperature in smaller increments is shown in Table 1. The number of 2-aminopurine-induced revertants did not decrease regularly with temperature but showed a break near 37°C. In this experiment we also measured
M. J. BESSMAN
L. J. REHX-ICHANT%
mutagenesis at 30°C (-O-O-) FIG. 3. 2.Aminopurine and at 43°C ([email protected]
). The procedure for 2aminopurine mutagenesis is described in Materials and Methods of the accompanying paper. The phage used was T4DlrUV199. At 1 mg/ml only 20 to 30% killing was observed concentrations. at 30°C and 43”C, with little or no killing at lower 2.aminopurine
Temperature Temperature W) 23 30 34 37 43
effect on mutation
Spontaneous revertants/108 cells 68 49 40 24 4
2-Aminopurine-induced revertants/lOs cells 3800 3200 3800 2000 420
The reversion of rUV199 was measured by following the progeny of a single phage through several replicative cycles (spontaneous) or was induced by 0.5 mg 2-aminopurine/ml. Both procedures are described in Materials and Methods in the accompanying paper. Eschetichia co& CR63 was the host bacterium. The number of spontaneous revertants was determined in one experiment and the number of 2-aminopurine-induced revertants was determined in a second experiment.
spontaneous revertants (in the absence of 2-aminopurine), with essentially the same results. Although, as expected, the total number of spontaneous revertants was much lower than those induced by 2-aminopurine, the spontaneous revertants also decreased as a function of temperature. (c) Effect of temperature on the spontaneous reversion of speci$c rII sites
Not all rI1 sites are equally sensitive to an increase in temperature in respect to their spontaneous reversion frequency. Table 2 shows the reversion frequencies of nine different sites. Two points are noteworthy. First, at a given temperature (30°C to 42°C) there was a wide distribution of spontaneous mutation rates for the different sites. Second, some of the sites were much more sensitive to temperature than others.
TABLE 2 rII
at 30°C and 43°C in E. coli BB Titer
Revertants/108 cells 30°C 43°C
uv199 uv373 UV248 UV183 uv4 SM51 uv13 UV6 UV68
60 1470 6 66 5 2 20 76 3
30°C (Phage/ml 4.6 4.5 5.4 4.8 4.2 4.7 5.4 4.1 4.2
3 2100 0.6 6 0.4 0.3 73 42 1.6 Mean
Results are from a single experiment. Reversion frequencies ,Materials and Methods of the accompanying paper.
amB22/UV199 amE4317/UV199 amBU23/UV199 Reversion ing paper.
x 10-11) 2.0 2.7 2.3 3.2 3.4 2.8 3.0 2.5
rII + Reversion frequencies of gene 43 arnberlrll 30°C and at 43°C Mutant
Revertants/lO* cells 30°C 43%
BlO(aup + 1) B40(sup + 2) B40(aup + 3) B40(aup + 1) BBO(sup+3) B4O(aup + 1)
were measured as described in Materials
0.23 365 1.2 91 324 0.9
0.01 22 0.2 37 32 0.07
and Methods of the accompany-
M. J. RERSMAN
1,. .J. H,EHA-KRAN’I’Z
4. Discussion The diminished incorporation of 2-aminopurine into DNA at, elevated temperat,urcs by purified DNA polymerase may be explained by the fact that t’he t’otal catalytic activity of the enzyme is the result of two reactions: (1) the insertion of a deoxynucleoside triphosphate at the 3’-end of a growing polydeoxynucleot’ide chain by the polymerizing function of the enzyme with the concomitant release of pyrophosphate. (2) The removal (hydrolysis) of a terminal nucleotide at’ the 3’-end by the exonuclease function of the enzyme. The most common techniques for assaying DNA synthesis involve the incorporation of a radioactive precursor into an acid-insoluble fraction. However, the incorporation of a precursor is in reality the result of reactions (1) and (2) and is only equivalent to the insertion of a triphosphate when the removal frequency, or exonuclease activity, is zero. Conditions which have a disproportionate effect on either side of the reactions may have a profound influence on the net incorporation of a particular nucleotide. On the other hand, it is impossible to know how much of a particular deoxynucleoside triphosphate was actually utilized by a given DNA polymerase merely by measuring its incorporation into DNA. This quantity, which we call the insertion (Bessman et al., 1974), is measured as the sum of the incorporation (acid-insoluble radioactivity) and the turnover (conversion of a deoxynucleoside triphosphate to a deoxynucleoside monophosphate). In a number of DNA polymerase mutants, the interplay of these two reactions has been correlated with the incorporation of 2-aminopurine into DNA, and to the spontaneous mutation rate of the organism (Bessman et al., 1974; Goodman et al., 1974; Reha-Krantz & Bessman, 1977). The observations reported in this paper may also be manifested through the mechanism of a disproportionate change in one of the activities of the enzyme. Englund (1971), Brutlag and Kornberg (1972) and Hershfield $ Nossal (1972) have all observed that an increase in temperature greatly stimulates the 3’-exonuclease activity resident in the polymerase molecule. Lo & Bessman (1976) have provided experimental evidence that this large increase in rate is most likely due to the melting out or denaturing of the ends of the DNA, thus making the 3’-terminus more susceptible to the 3’-exonuclease. For example, the was hydrolyzed by purified T4 polymerase perfectly matched poly(dA) *poly(dT) 30 times faster at 35°C than at 25°C. However, the same substrate modified to contain a mismatched 3’-terminal deoxynucleotide showed only the expected twofold increase in rate over the 10 deg. C rise in temperature. Since this substrate was already singlestranded due to the mismatch, it did not show the very high activation resulting from the increased single-strandedness of the double helix caused by melting at the ends. The data in Figure 1 are in keeping with this interpretation, and most likely reflect the destabilization at the ends of the double helix resulting in a more sensitive substrate for the single-strand specific 3’-exonuclease function of the enzyme. Although the total catalytic activity of the enzyme increased with increasing temperature, as shown by the insertion or total nucleotide utilized, the actual incorporation of 2-aminopurine into DNA went through a maximum at about 30°C and then declined. This means that a greater share of the 2-aminopurine was turned over, or hydrolyzed, at the higher temperature and this is seen in Figure l(a). Note also that the 2-aminopurine incorporated as a proportion of the total utilized dropped steadily increase with increasing temperature (Fig. l(b)). Th us, there was a disproportionate in the 3’-exonuclease, or editing function, as the temperature was increased. The incorporation of deoxyadenyhc acid, on the other hand, behaved quite differently, as
shown in Figure 2. In this case, the rate of incorporation was increased at 38°C relative to 30°C whereas 2-aminopurine was again lower at the higher temperature. This disparity may be due to the difference in the relative stability of the adenine * t,hymine base-pair in comparison to the 2-aminopurine . thymine base-pair. Cerami (1969) and Reich (personal communication) have indicated that double-stranded polydeoxynucleotides containing 2-aminopurine as partial replacement for adenine are much more unstable and denaturable than the corresponding unmodified polymer. At equivalent temperatures, the 2-aminopurine~thymine base-pair would be expected to have more single-stranded character than the adenineethymine pair, thus rendering it more susceptible to the editing function of the polymerase. This line of reasoning was tested in viva by measuring the reversion frequency (mutation rate) induced by 2-aminopurine at 43°C and at 30°C. Figure 2 demonstrates clearly t’he reduction in the 2-aminopurine-induced reversion frequency at the elevated temperature, in keeping with the notion that the editing function is more efficient at t,he higher temperature. A less interesting explanation of these results would be that one of the several steps in the metabolic pathway leading from free 2-aminopurine outside of the cell to the deoxynucleoside triphosphate inside the cell (Rogan t Bessman, 1970) was sensitive to small increases in temperature, thereby reducing the effective pool size of the triphosphate. We have had technical difficulties in attempts to measure the small pool sizes of the 2-aminopurine derivatives, so that this possibility has not been excluded directly. However, we do not believe t’hat this explanation is likely, since we have observed the same effect of temperature on spontaneous mutations not involving 2-aminopurine or any other added base analogue. In Tables 1, 2 and 3, large decreases in spontaneous mutation rates are evident even in the absence of 2-aminopurine. How may the effect of temperature be generalized to explain the results with 2-aminopurine on the one hand, and spontaneous mut,ations on the other? We believe t,hat both cases are manifestations of the same phenomenon. That is, increases in temperature have a destabilizing effect at the ends (growing points) of the DNA helix, imparting more single-stranded character to the structure. This effect would be exaggerated for any base-pair already having a weak bond, such as the 2-aminopurine.thymine pair. It would also be expected to play a role in any point mutation involving insertion of the incorrect base leading to a transition or transversion. The probability of incorporating an incorrect base would then be determined partly by environment. We submit that one the stability of the double helix in its immediate important basis for the wide variation at particular sites is the nucleotide sequence in t’he immediate environment. This possibility has already been considered by Benzcr (1961) in his elegant studies on the topography of the rII cistron, and was discussed in relation to the occurrence of hot spots of mutabihty along the gene. Some of our own data showing site-specific variation in mutation frequencies are presented in Table 2, where it can be seen that the mutation frequency can vary over 2OOO-fold. One might be tempted to think that the regions of high mutability would be intrinsically less stable A+T-rich regions of the DNA, allowing for more flexibility in accommodating mismatched bases. On the contrary, we propose that the high mutability may be associated with a G +C-rich microenvironment, which has a stabilizing effect on an inserted incorrect base. Specifically, we believe that, the immediately preceding base-pairs or nearest-neighbors would have the most effect in determining the double-strandedness at the growing point and t,hus be intimately
M. J. BEAHMAN
L. J. RRHA-KRANTZ
involved in determining the efficiency of the 3’-exonuclease during DNA synthesis. Data consistent with this hypothesis have been reported by Koch (197 1), who showed that changing an A.T to a G.C pair increased the mutation frequency at a neighboring site tenfold. In the light of this discussion, it is of interest to note some consistencies in the data given in Table 2. The reversion frequency of UV373, which is already very high, is relatively unaffected by temperature. We suggest that UV373 is G + C-rich, which stabilizes the helix allowing more frequent incorporation (less frequent removal) of a mismatched base and also confers more resistance to the effect of temperature. Drake et al. (1969) have reported that the reversion of UV373 is relatively unaffected when tested in an antimutator background. This could be attributed to the 3’exonuclease activity responsible for the antimutator phenotype (Muzyczka et al., 1972; Lo & Bessman, 1976) having a diminished activity on the relatively stable G + C-rich region. On the other hand, UV199 and IJV183, which show the largest effect of temperature and which we hypothesize may represent an A+T-rich environment, are very highly affected in the antimutator background (Drake et al., 1969), probably due to the enhanced activity of the editing function in an already less stable region. Similar correlations may be noted with other mutants listed. It is interesting that UV6 and UV58, two frameshift mutant’s, do not show a relatively significant effect of temperature. Two other points should be noted. First, reconstruction experiments with revertant and mutant phage reveal that there is no selective advantage for either at either temperature, ruling out a selection artifact as a possible explanation for our observations. Second, we have noted a small but consistent decrease in the phage titer when cultures are grown at 43°C instead of at 30°C (Table 2). This may be due to a slightly diminished rate of DNA synthesis at the elevated temperature as a result of the destabilization of the double strands at the growing point. Lo & Bessman (1976) have reported that the temperature sensitivity of an antimutator phage, L141, which leads to very low burst sizes is due to the exaggeration of the already high 3’-exonuclease function of the mutant polymerase resulting in a reduced rate of DNA synthesis. On the other hand, increased temperature may affect other proteins known to be involved in DNA replication, thereby reducing the rate of DNA synthesis. This could reduce burst sizes and also increase the opportunity for editing. Our experiments in vitro on the effect of temperature on the incorporation of 2aminopurine into DNA and on the frequency of base analogue induced and spontaneous mutation rates have led us to hypothesize that the frequency of a base substitution is strongly determined by the base sequence immediately preceding the mistake. For simplicity, we have considered G *C and A *T pairs as playing the dominant role in determining stability of the helix. However, other specific interactions such as base-stacking must be considered and these may be important in determining the probability that a specific base will or will not be edited. These ideas are subject to experimental test in vitro and in vivo and we are presently engaged in both types of experiments. This is contribution no. 923 from the McCollum-Pratt by a grant (no. GM18649) from the National Institutes
Institute. This work of Health.
REFERENCES Benzer, S. (1961). Proc. Nat. dcad.Sci., U.S.A. 47, 403-415. Bessman, M. J., Muzyczka, N., Goodman, M. F. & Srhnaar, R. L. (1974). J. Mol. Biol. 88, 409-421. Brutlag, D. & Komberg, A. (1972). J. Biol. Chem. 247, 241-248. Cerami, A. (1969). Ph.D. Thesis, Rockefeller University, New Pork. Drake, J. W. & Allen, E. F. (1968). Cold Spring Harbor Xymp. Quant. Biol. 33, 339-344. Drake, J. W., Allen, E. F., Forsberg, A. A., Preparat,a, K. M. & Greening, E. 0. (1969). IYature (London), 221, 1128-1131. Englurld, P. T. (1971). J. BioZ. Chem. 246, 5684-568i. E’reesc, E. B. & Freese, E. (1967). I’roc. Nat. Acad. Sci., U.S.A. 57, 650.-657. Goodman, M. F., Gore, W. C., Muzyczka, N. & Bessman, M. J. (1974). J. Mol. BioZ. 88, 423-435. Hershfield, M. S. & Nossal, N. G. (1972). J. BioZ. Chem. 247, 3393-3404. Koch, R. E. (1971). Proc. Nat. Acad. Sci., U.S.A. 68, 773-776. Lo, K. & Bessman, M. J. (1976). J. BioZ. Chem. 251, 2475--2479. Muzyczka, N., Poland, R. & Bessman, M. J. (1972). J. BioZ. Chem. 247, 7116-7122. Reha-Krantz, L. & Bessman, M. J. (1977). J. Mol. BioZ. 116, 99-113. Hogan, E. G. & Bessman, M. J. (1970). J. Bacterial. 103, 622-633. Schnaar, R. L., Muzyczka, N. & Bessman, M. J. (1973). Genetics, 73, 137-140. Spfa>,er, J. P., Karam, J. D. & Lenny, A. B. (1966). Cold Spring Harbor Symp. Quant. Biol. 31, 693 -697. ~V:R.tsoI~, ,J. D. & Crick, P. H. C. (1953). Cold Spring Harbor Symp. Quant. BioZ. 18, 123131.