Zinc Ion-DNA Polymer Interactions X I N JIA and LUlGl C. MARZILLI*t Department of Chemistry, Emory University, Atlanta, Georgia 30322
SYNOPSIS
The adjacent GN7-M-GN7 cross-linking and adjacent G-M-G sandwich-complex models for DNA metal ion binding were evaluated both with native DNAs differing in GC content as well as with the synthetic polymers poly[ ( dGdC)Iz, poly[ ( dAdT)I2, and poly [ (dAdC) (dGdT) 1 . The effect of Zn2+was studied in depth, and limited studies were also performed with Co2+and Mg2+.The results were compared to the extensive information available on Cu2+binding to native DNAs and poly [ (dAdT)],. At high ratios of metal/ base ( R ) ,Zn2+caused all native DNAs to denature with the same melting temperature T,, 61°C. A similar pattern was reported previously for Cu2+,but the typical T , was 35°C. The extent of renaturation on cooling DNAs denatured in the presence of Zn2+ increased with GC content, as reported previously for CU". These results, together with previously reported similarities, strongly indicate that the DNA binding characteristics of the two cations are similar. By comparison of the T , values and hyperchromicity changes monitored at 260 and 282 nm, it is clear that, during thermal denaturation in the presence of Zn2+,both AT and GC regions were denatured, even at high R. The T , vs R profile for the native DNAs was typical. The rise at low R and subsequent decrease at high R were inversely and directly related, respectively, to GC content. Except for poly [ (dAdT)12, where T, increased with R , the other synthetic polymers exhibited the increase/decrease pattern. Poly[ (dAdC) (dGdT)] gave a T , value a t high R of 54°C. In the absence of Zn", this polymer exhibited little hypochromicity on cooling of denatured polymer. However, in the presence of Zn2+,nearly complete hypochromicity was observed, although the mid15°C at R = 10. These point of the cooling curve was lower than the T , value by characteristics were similar to those with native DNAs, although viscosity and CD studies suggested that the "renatured' polymer was not identical to the unheated polymer. Furthermore, addition of Zn2+after denaturation nearly completely reversed the absorption increase. This finding contrasts with those for native DNAs, where the Zn2+must be present during denaturation in order to reverse the absorption increase nearly completely on cooling. With some caveats, poly [ (dAdC ) (dGdT) ] appears to be a good model for native DNAs since its properties, including CD and uv changes on addition of Zn2+ to premelted and melted polymer, parallel those of the native polymers. Based on these findings and the discovery that Zn2+actually inhibits renaturation of poly [ (dGdC)I,, we believe adjacent G-M-G complexes are not the primary species responsible for the spectral changes in premelted DNAs, nor are they the principal species promoting renaturation. Three interrelated hypotheses to explain these phenomena were identified for further study as follows: ( a ) a kinetic effect-the metal ion promotes renaturation of denatured regions formed during thermal denaturation with metal ion present; ( b ) an inhibiting effect-the metal ion prevents the initial formation of conformations that otherwise inhibit renaturation; and ( c ) a CN3 binding effect-the metal ion lowers T , by stabilizing the denatured state through C binding. We speculate that such CN3 binding may account for the unexpectedly poor ability of Co2' both to lower T, and to promote renaturation.
--
-
Biopolymers. Val. 31, 23-44 (1991) 0 1991 J o h n Wiley & Sons, Inc.
CCC 0 ~ 6 - 3 5 2 5 / 9 1 / 0 1 ~ 2 3 - 2 2 $ 0 4 . 0 0
* To whom correspondence should be addressed.
t Also a member of the Winship Cancer Center.
23
24
JIA AND MARZILLI
INTRODUCTION T h e special role that zinc plays in nucleic acid biochemistry prompted us a few years ago t o develop methods for obtaining solid crystalline complexes formed between aquated Zn2+(hereafter, Zn '+) and nu~leotides.'-~ Out of such studies emerged a model' for the binding of Zn2+ to DNA. T h e model was based, in part, on a study of DNA polymers with C U ~ + . ~ Specifically, we suggested one of the ZnZ+DNA binding interactions might be a n N7,N7 intrastrand cross-link between two adjacent G residues.' This GN7-M-GN7 binding mode permits retention of GC base pairing and such adducts could act as nucleation points for renaturation of DNA strands when DNA is thermally denatured in the presence of Zn2+.Before the development of this model, Daune had proposed a chemically unreasonable G-Cu-G sandwich complex to serve as one of the nucleation points for DNA r e n a t ~ r a t i o nthis ~ ; model was based on studies of native DNAs with differing GC content. Our proposal is chemically more reasonable since it was based on information that we and others had obtained since the Daune proposal. For example, Reedijk and co-workers had shown that a n N7,N7 GG intrastrand cross-link with Pt retained its base pairing under appropriate conditions in a n oligon ~ c l e o t i d e Both .~ calculations and additional nmr studies confirmed the Reedijk results.6-8 Munchausen had demonstrated that DNA platinated a t low levels (now known to constitute primarily N7,N7 GG intrastrand cross-links9"0)could be renatured more readily than untreated DNA.".'* Such anticancer-active Pt ( I1 ) compounds cannot be considered as good models for Zn2+;P t ( I 1 ) forms inert square-planar complexes, whereas Zn '+ forms labile octahedral or tetrahedral complexes. Zimmer e t al. have postulated that when it binds to G residues of DNA, Zn2+forms a n N7,PO4 chelate.I3 On the basis of studies with Cu", Daune has also proposed such a chelate as another major species in addition to the sandwich specie^.^ This chelate binding mode proposed for these labile metal species has now been found for nucleotide complexes of Pt (11) anticancer drug^'^,'^ as well as for nucleotide complexes of anticancer drugs containing Mo16 and R u . ' ~A knowledge of the binding of such inert metal centers to nucleotides, oligonucleotides, and DNA can be used judiciously to assess proposals for the complex situation of the binding of labile metal ions, such as Zn '+,to polymers. Information on labile Cu2+ complexes may be more relevant than that on P t ( I 1 ) to Zn2+ complexes. However, Cu2+ usually prefers ( a ) square-
planar geometry, ( b ) a five-coordinate geometry, or ( c ) a n octahedral geometry with significant JahnTeller distortions. Additionally, paramagnetic Cu (11) complexes are difficult t o study by the informative method of nmr spectroscopy since the contact contribution dominates T2 relaxation and contributes significantly to TI relaxation." We felt additional studies on polymers with Zn2+ were needed since the information in the literature was much less complete than that for Cu2+.Such studies are important in the design of sequences of synthetic oligonucleotides for nmr studies with Zn2+,which may prove useful in explaining Zn2+DNA binding and the effects of Zn2+ on DNA properties. Furthermore, Zn2+ is a diamagnetic species, and nmr studies with Zn2+ should be less complicated than those with Cu2+. Based on comparative effects of metal ions on DNA denaturation, the following series was established: Mg2+,Co*+, Ni2', M n 2 + ,Zn2+, Cd2', and cu2+ .19 Eichhorn has emphasized that there should be no dichotomy between the binding modes of one metal ion and those of another nearby in the series, but that these effects ( a n d hence the binding interactions) should vary to a matter of degree.lg T h e series appears to reflect increasing affinity for binding to base N's vs phosphate 0's; differences in the effects of Cu2+ and Zn2+ on DNA properties have been attributed t o the stronger binding of Cu2+to the base N's.'' More recently, in a valuable assessment of metal binding characteristics, Martin pointed out that, although the binding constants t o N7 of purine nucleosides are a n order of magnitude greater for Cu2+than for Zn2+,the relative binding toward N7 of the common nucleosides follows a similar order: adenosine inosine < guanosine.'l In this report, we further assess the similarity of binding modes of these two labile ions with studies of Zn2+binding t o native and synthetic DNAs similar to those reported previously with Cu2' .4,22*23 Indeed, many of our findings support the relationship between Zn2+and C u 2 + .Furthermore, we have also tested the models with poly[ (dAdC) ( d G d T ) ] , which can form neither intrastrand nor interstrand adjacent GG cross-links. With this polymer, as well as some native DNAs, we have further explored the contrast between the effects of Co2+and Zn2+(i.e., their separation in the above series) compared to the typically similar chemistw of these metal ions.
-
EXPERIMENTAL Materials
The following DNAs were selected for study: Micrococcus lysodeikticus ( M L D N A ) , 72% GC; Esch-
ZINC ION-DNA POLYMER INTERACTIONS
erichia coli ( E C DNA) 51% GC; Salmon sperm (SS DNA ) 42% GC; Clostridiumperfringens ( C P DNA) 30% GC; and calf thymus ( C T DNA) 44% GC. The synthetic polymers selected were poly [ (dGdC) 12, poly [ ( dAdT ) 12, poly [ ( dAdC ) ( d G d T ) ] , a n d poly[ (dIdC)J 2 . Native DNA solutions were supplied by Dr. W. D. Wilson of Georgia State University. These samples had been sonicated to average 600 base pairs in length; the solutions were dialyzed against piperazine- N,N'-bis ( 2-ethanesulfonic acid) ( P I P E S ) 10 buffer (0.01M PIPES, 10-5M EDTA, 0.10M N a N 0 3 a t p H 7.00) and passed through a 0.22-p Millipore filter.24 T h e synthetic polymers, purchased from Pharmacia, were used without sonication. ZnClz and MgC12* 6 H 2 0 (from Aldrich Chemical Co.) and CoClz 6 H 2 0 (from Fisher Scientific Co.) were reagent grade and were used without further purification.
-
Instrumentation
All thermal denaturation and renaturation experiments were performed on a Perkin-Elmer Lambda 3B UV/VIS spectrophotometer equipped with a thermally jacketed five-cell compartment and a temperature probe, which can be placed directly in one of the five sample solutions. The temperature was controlled by a temperature bath with a programmer capable of controlling the rate of temperature change. The spectrophotometer was interfaced with an Apple I1 computer through a software program designed to record the sample temperature and its corresponding absorption as a function of temperature change.25The uv scans were also obtained from the same type of instrument. Viscosity studies were performed with a Cannon-Ubbelohde viscometer in a thermostated water bath a t 30.00"C. CD spectra were acquired with an AVIV model 62DS that is equipped with a thermally jacketed cell compartment. Methods
In a typical denaturation ( "melting") experiment, five samples were run simultaneously in sealed cuvettes with a 1-cm path length. T h e first cuvette was a blank and the temperature probe was placed in this cuvette. I t contained the MES 00 buffer solution [ 0.01M 2- (N-morpholino) ethanesulfonic acid a t p H 6.00 and passed through a 0.22-p Millipore filter]. No extra salt was added and the salt introduced by p H adjustment was approximately 5 X 10-.3M.T h e other four cuvettes contained solutions of DNA (4.2 X 10-5-1 X 10-4M,diluted from the stock solutions described above) with different
25
Zn2+concentrations in MES 00 buffer. A 0.1 M Zn2+ stock solution was prepared by weighing a n appropriate amount of ZnClz into a 50 mL volumetric flask, adding 2 small drops of 5.5M HC1, and then deionized water. The final p H = 2.63 and the Zn2+ concentration measured by atomic absorption spectroscopy was 0.1OM. The total volume of the solution in each cuvette was 2.0 mL. The temperature was increased for melting and decreased for cooling experiments at the rate of 0.5"C/min. T h e melting temperature T , of DNA was determined by plotting the first derivative, AAIAT vs T . Tmis equal to the temperature at which AAIAT is a t a maximum. Cooling experiments were performed immediately after completion of the melting experiment. Except where noted, the absorbance was monitored a t 260 nm. T h e percent hyperchromicity (Hyp % ) was calculated according t o the equation Hyp % = (A, - A , ) / A f X 100, where A f is the final absorbance after melting and A, is the initial absorbance before melting. In the experiments on the time dependence of polymer annealing on the addition of metal ion, the polymer, without metal ion added, was melted. Then the solution was cooled a t O.S"C/min to 10.0, 20.0, and 30.0"C in separate experiments. At this point, the change in absorbance a t 260 nm was monitored as a function of time, both in the absence and in the presence of added metal ion. C T DNA, poly[(dAdT)], and poly[(dAdC)( d G d T ) ] were used for viscosity studies. Each point in the viscometric titrations represents a t least three measurements with a n average deviation of less than 0.10 s. For denatured DNA, the reservoir of the viscometer containing the sample was placed in a water bath at 75°C for 15 min, and then the viscometer was returned to the original water bath at 30.00"C for the viscometric measurements. The calculation of absolute viscosity was based on the equation:
where r is the time reading in seconds and C is the DNA concentration in bases. The relative viscosity ( RV ) is defined here as the ratio ?denatured/ qnatlve. CD spectra were obtained a t 25°C. When denatured DNA was needed, the cell compartment temperature was raised to 75°C for 10 min, then decreased to 25°C a t l"C/min. T h e MES buffer for the CD studies was five times more dilute than that used for the uv melting experiments in order to improve results in the 200-230-nm region. Two CD spectral scans were recorded a t 1nm/s and averaged. T h e baseline was corrected using software supplied by the manufacturer.
26
JIA AND MARZILLI
increase in R , several phenomena could be observed from the melting curves as follows: ( a ) the melting curve became sharp; ( b ) the hyperchromicity typically decreased and ( c ) the melting temperature increased at first, but then decreased. At high R values, usually R > 10, and high GC content DNA, the solution absorption decreased as the temperature was raised well above the T,. The T, values of four native DNAs and four synthetic polymers were measured (Table I ) and plotted vs R ( Figure 2 ) . Of some interest, the T,'s of native DNAs (Figure 2a) a t high R values were very similar regardless of DNA base-pair composition and sequence. At low R values, T , increased until R 2; except for ML DNA, where the maximum was a t R 1. The values then decreased to a similar range
0.68-
A260nm
-
0.58
-
0.48 40
I
50
I
60
I
70
-
I
80
Thermal Denaturation
A typical example of melting curves as a function of Zn2+base ratio ( R ) is shown in Figure 1. With
small 2°C increase a t R = 0.3. However, the study of this polymer was limited because of aggregation, even a t R = 2. Poly [ (dAdC) (dGdT)] exhibited a T , vs R curve similar to those of the native DNAs,
Table I DNA Melting Temperatures ("C) at 260 nm with Different RznmaseValues
RZIllbase
0 0.3 0.5 0.8 1.o 1.2 1.5 1.8 2 4 6 10 12 14 15 18 20
DNA GC%
GC a 100
ML
EC 51
AC/GT~
ss
CP
A T'
72
50
42
30
0
ICd 0
82.6 84.7
67.2
60.8
58.6
58.8
49.5
31.7
26.7
Poly[(dGdC)]Z. Poly [ (dAdC)(dGdT)I. P~ly[(dAdT)]z. Poly[(dIdC)]z. This data was obtained from 282 nm.
75.2
69.7 66.2 64.5 62.0 61.1 60.7 59.8 59.6
33.3
60.8
75.4 71.7 70.0 68.5 67.4 67.0 60.0'
69.3 65.4 61.9
60.1
67.2
67.2
64.4
47.0
37.0
63.3 60.0 58.1 56.3
70.5 68.6 66.1 63.5 62.0
66.7 66.1 64.0 62.7 61.3
50.8 53.3 54.7 55.9
39.2 36.5 34.2 31.5 30.8
61.4 60.7 60.1
61.2 60.7 60.0
53.9
29.7 56.0
28.6
ZINC ION-DNA POLYMER INTERACTIONS
a
Tm?C
45
! 0
I 5
10
15
20
R(Zn/Base)
significant decreases in hyperchromicity accompanied the cooling, we performed a second melt with the solution. These data are also collected in Table 11. We assume that the relative values of the hyperchromicity in the first and the second melts are indicative of the extent of renaturation in the cooling process. Relatively closer hyperchromicity values of the second melt t o the first melt were found for DNAs with higher GC content. For any given DNA, the hyperchromicity of the first and the second melts became more similar a t higher Zn2+concentration. For ML DNA, the renaturation was almost complete a t R = 2, but for SS DNA, maximum renaturation required R = 10. Obviously, the extent of the renaturation greatly depends on two factors: GC content and Zn2+concentration. Annealing Experiments
b
Tm,"C
Examples of cooling curves a t different R values for one native DNA, C P DNA, and one synthetic polymer, poly [ (dAdC) ( d G d T ) ] , are shown in Figure 3. For C P DNA (Figure 3 a ) , a biphasic cooling profile was observed a t R values of 4 and 6: the absorption did not decrease significantly during the first 10°C decrease in temperature. Then there was a sharp decrease in absorption, which was followed by a gradual decrease. There were slight differences between the R = 4 and 6 studies: the absorption decrease of the first phase increased with Zn2+concentration. At 20°C, 60% of the initial absorption increase was reversed. At low R values ( 0 and 2, figure not shown), the cooling curves exhibit a smooth monophasic decrease in absorption, and a t 20°C the absorption had decreased by 50% of the initial increase on denaturation. For DNAs with higher GC content, somewhat different cooling profiles were observed compared to the low GC content C P DNA. For SS DNA, with a n average GC content, biphasic behavior was observed a t low R values but not high R values. For the high GC content DNA, ML DNA, biphasic cooling behavior was not observed (figure not shown). For poly [ (dGdC ) I,, poly [ (dAdT) 12, and poly[ (dAdC ) ( dGdT) 1, a t R 2 1, the cooling curves are monophasic with complete reversal of hyperchromicity (for poly [ (dGdC) l2 in the presence of Zn2+, the recovery of hyperchromicity was not complete a t longer wavelength, see below). The situation for poly [ (dAdC ) ( d G d T )] a t very low Zn2+content (or in the absence of Zn2+)is complex and is discussed below. However, for the other two polymers, the cooling profiles were monophasic, even in the absence of added Zn2+.The temperature of the mid-
-
R(Zn/Base)
Figure 2. Relationship between T, and RZnlbase.( a ) Native DNAs: -, ML DNA, * * * , EC DNA; - - - -,
--
SS DNA; and - - -, C P DNA. ( b ) Synthetic polymers: . . -. poly[ (dGdC)],; * * * ,polyj (dAdC) ( d G d T ) ]; - - - -,poly [ (dAdT) I,; - - -, poly [ (dIdC) 1,; and -, ML DNA.
-
27
-
-
but the values of T , a t high R were lower. Also, the maximum in T , occurs at R = 1. In contrast, the T , vs R curve of poly[ (dAdT)],, which lacks GC base pairs, never decreased. Poly [ (dIdC) 1 2 had much lower T,'s than any other DNA or polymer studied here; however, the shape of the T, vs R curve for this polymer follows the typical increase / decrease behavior of the native DNAs. It is instructive t o compare the Hyp % for both native and synthetic DNAs as a function of added Zn" (Table 11). In most of the experiments where
-
28
JIA AND MARZILLI
Table I1 Hyperchromicity Data (70)of First/Second Melt of DNAs at 260 nm
R
GC a
0 0.5 1 1.5 2 4 6 10
16.4/16.2 14.0/13.2 13.3/7.2 6.216.0
ML 28.1/-
EC 28.5/-
24.9/24.7/22.4 23.1/21.0 23.3/20.7 21.9/19.1
ss
AC/GT~ 24.2/5.1
CP
25.1/-
21.8/19.1 26.1/25.9/24.1/-
20.5/19.3/19.0/18.0 18.7/18.0
27.9/5.7 24.5/19.3 22.6/18.6 20.4/18.6
AT
24.6/-
31.4/30.6
27.1/-
33.0/32.4
25.7/3.8 21.7/6.3 18.9/11.3 17.9/13.1
32.9/33.4 31.9/31.8 31.9/33.4 31.2/30.9
Poly[(dGdC)],. Poly[ (dAdC)(dGdT)1. ' Poly[(dAdT)]z. a
point of the cooling curve for most polymers and native DNAs decreased with increasing R (see Table 111). For example, for ML DNA, these midpoints were 50, 32, 23, and 21°C for R = 2, 4, 6, and 10, respectively. However, for poly [ (dAdT)12,the midpoint temperature of the cooling curve was similar t o the T,.
increase/decrease pattern of Hyp % was seen also for the high GC content DNAs. However, for C P DNA and poly [ (dAdT) 12,the Hyp % increased with increasing Zn2+ concentration. It should be noted that, although there was no hyperchromicity found a t 260 nm at R = 2 for poly [ ( dGdC)I2,the melting could be monitored at 282 nm (see Table IV) .
Denaturation Monitored at 282 nm
Absorption Spectra
In order t o assess the differences in interaction between AT and GC base pairs with Zn2+, the approach used was based on the dispersion of the hyperchromic effect with wavelength. This method was used originally t o analyze DNA base composition.26 Changes in absorbance and in hyperchromicity could be resolved into components arising from AT and GC base pairs. At 260 nm, larger absorbance changes on melting were found for AT-rich DNA than for GC-rich DNA. Since almost no absorbance change was found for poly [ (dAdT)], a t 282 nm, the melting behavior of GC base pairs can be more directly monitored at this wavelength. Hyperchromicity data for seven DNAs and polymers at 260 and 282 nm are listed in Table IV. T h e hyperchromicity changes for poly [ (dGdC) l2 were 4 times greater a t 282 than at 260 nm. I n contrast, absorption changes for poly [ ( dAdT)Izwere too small t o be seen a t 282 nm unless Zn2+was present. For DNAs with intermediate GC content, the relative hyperchromicity a t 260 and 282 nm reflected the expected trend. The Hyp 76 a t 260 nm decreased with increasing Zn2' content for DNAs with high GC content, such as poly[ (dGdC)I2, ML DNA, and poly[ (dAdC)(dGdT) ] . For CP DNA and poly [ (dAdT) 12,a small increase in Hyp % a t 260 nm was observed on addition of Zn2+ t o R = 1. T h e Hyp % decreased a t higher R values. On the other hand, at 282 nm this
There are three important points ( a , 6, c ) where metal ions could be added as shown below:
-
heat cool We examined the uv spectral properties of C T DNA, poly [ ( d G d C ) 12, poly [ ( d A d C ) ( d G d T ) 1 , a n d poly [ (dAdT ) l2 with no metal added and in separate experiments with Zn2+ added a t each of the three points. For each polymer, the spectrum was recorded a t room temperature in the absence and presence of Zn2+at R = 10, except for poly [ (dGdC) 12, where R = 1. No significant change in spectrum was observed (Figures 4a, 5a, and 6 a ) . In the absence of Zn2+,the C T DNA sample was heated to 80°C over 10-15 min and the spectrum was recorded (Figure 4b). T h e spectral shape did not change, with one band a t 257 nm and a slight red shoulder. However, there was a n approximate 27% hyperchromicity with a slightly greater increase for the unresolved shoulder. As the sample was cooled to room temperature over a 40-min period, the spectral shape and position did not change, but there was a decrease in intensity throughout, which reversed 20% of the intensity increase induced by heating (Figure 4b).
-
ZINC ION-DNA POLYMER INTERACTIONS
a
032
29
obtained when the sample was heated in the presence of Zn2'. However, on cooling, the absorption maximum had a 3-nm red shift and an 40% higher intensity relative to the initial room temperature spectrum a t 260 nm. An even higher relative intensity was observed a t longer wavelength, e.g., 50% at 272 nm (Figure 4c ) . Finally, a C T DNA sample was heated and cooled as above, and then Zn2+was added (Figure 4d). The spectral shape was similar t o that a t high temperature but the absorption reversed 60% of the intensity increase. The final Hyp % a t 260 and 272 nm were 1 2 and 14%, respectively. In experiments that parallel the absorption spectral studies, we wished to determine the effects on denaturation (monitored a t 260 nm) of adding Zn2+ at points b and c instead of a. When C?' DNA a t 5 X 10-5M (higher concentrations than in Table I ) was denatured, the Hyp % = 30. For C T DNA, the T , a t R = 10 (point a)was 64°C with Hyp % = 28%. However, for a n R = 10 solution with Zn2+added to denatured C T DNA (point b ) there was a n initial decrease in absorbance (19% of the initial absorbance increase), consistent with the uv plot (Figure 4c). On cooling, the absorbance remained constant until 43"C, whereupon a n additional 57% of the initial absorbance increase was reversed. After 30 min or 17 h, reheating led to a Hyp C b = 16 and a T, of 37°C. Likewise, the cooling curves in the presence of added Zn2+were similar to the heating curves. If the C T DNA was cooled t o 20°C and then Zn2' ( R = 10) added (point c ), the absorbance decreased (cf. Figure 4 d ) . A Hyp % = 15 and a T , = 37°C were observed when the DNA was reheated. During the reheating, the absorbance increase was complete a t 42"C, with only a slight additional increase to 70°C. A similar result was obtained when this solution was cooled and then reheated after 17 h. These results suggest that when the Zn is added a t point b or c , the hypochromicity changes do not correspond to the generation of any appreciable stretches of native DNA; otherwise, the T , values would have been much higher. The spectrum of poly [ (dGdC ) l2 has, , ,A a t 254 nm a t room temperature. The band tails to longer wavelength but no shoulder is evident. On heating to 90°C over 10-15 min, a pronounced shoulder emerged a t 275 nm with an intensity increase a t all wavelengths (Figure 5 b ) . When the sample was cooled t o room temperature, the shoulder disappeared and the intensity increase was reversed by 90%. The spectrum resembled the initial spectrum before heating (Figure 5 b ) . When a solution of poly[ (dGdC)], was heated to 90°C in the pres-
-
-
T. OC
b
036 -
-
-
0
2 6 10
!
.
,
.
.
,
20
30
I
.
40
,
50
.
.
, 60
70
TPC Figure 3. Melting and cooling curves of a native DNA and a synthetic polymer. ( a ) C P DNA (4.33 X 10-5M)in MES 00 buffer, pH 6.0. * ,R = 6 melting curve; -, R = 4 ; and - - - - , R = 6 cooling curves ( b ) Poly[ (dAdC)(dGdT)](4.11 X 10-5M) in MES 00 buffer, pH 6.0. --, R = 0 melting curve. * * * , R = 0. - - - R = 6; and - - -, R = 10 cooling curves.
.
---
- -
3
1
In the R = 10 sample, the absorption increased on heating a s above with a red shift of 5 nm obscuring the shoulder. On cooling as above, the resulting spectrum was similar to the initial spectrum of unheated C T DNA (Figure 4 a ) . When the sample was heated to 80°C in the absence of Zn2+and then Zn2+added, the absorption spectrum changed to superpose almost exactly that
-
-
'+
-
-
30
JIA AND MARZILLI
Table I11 Temperatures ("C) of the Midpoint of the Steepest Absorption Decrease on Cooling DNA Solutions DNA
GC"
0 0.5
77 60 48.5
1 2 4 6 10
ML
ss
AC/GTb
CP
AT
20
31.5
56
51
50 32 23 21
54.5 50.5 43
44 41
52 48 40
54.5 56
Poly[(dGdC)],. Poly[(dAdC)(dGdT)]. ' P~ly[(dAdT)]p. a
ence of Zn2' ( R = 1, point a), the pronounced shoulder was also observed (Figure 5 a ) . However, as the sample was cooled the shoulder did not disappear completely. The intensity increase at 275 nm was reversed by only 70% after 30 min. An R = 0 sample was heated to 90°C and then Zn2+to R = 1 was added (point b ) . The intensity at A, 254 nm decreased, but that at the 275-nm shoulder did not decrease (Figure 5c). The shoulder became slightly broader. After cooling to room temperature, the shoulder was still present and the initial intensity increase on heating was reversed by only 50%. For poly [ (dAdC) (dGdT) 1 , the phenomena observed were very similar to those found with C T DNA. In the absence of Zn2+,the sample was heated to 80"C, the spectral shape (A,, 258 nm, shoulder 272 nm) did not change, but there was 26% hyperchromicity at the spectral maximum (Figure 6b). When the sample was cooled to room temperature, the spectrum still resembled that at high temperature and the intensity decreased only 5%. In the R = 10 sample (point a),heating to 80°C
-
-
-
-
-
Table IV
-
induced 21% hyperchromicity at 258 nm and a ca. 2-nm red shift of the spectral maximum, which obscured the shoulder at 272 nm (Figure 6a). On cooling the sample, the spectrum recorded was similar to the initial spectrum with a slightly higher intensity at longer wavelengths. When the sample was heated to 80°C and then Zn2+added (point b ) ,the spectrum exhibited a red shift of 2 nm, and a shoulder at 272 nm was observed (Figure 6c). The spectrum was similar to that obtained when Zn2+was added at point a. On cooling, the resulting spectrum was similar to the initial spectrum at room temperature except for 5% hyperchromicity at 258 nm and a 10% hyperchromicity at 272 nm. The R = 0 sample was heated and cooled, and Zn2+was then added (point c ) . Following the absorption, intensity decrease with time demonstrated that the final spectrum was very close to the initial one at room temperature (Figure 6d). In comparison to the other polymers studied, the behavior was quite straightforward for poly[ ( dAdT)Iz.When the sample, either containing
-
-
-
-
Hyperchromicity Data (%) at 260/282 nm AT '
R
GC a
ML
AC/GTb
CP
0 0.5 1 1.5 1.8 2 4
14.7151.8 13.5152.7 12.5153.0 10.3151.6 5.6144.2 -126.9
28.1142.4
24.2132.1
24.6118.4
31.410
24.9144.6
21.8140.0
27.1120.0
33.0125
24.1143.2 23.1/42.1
20.5140.0 19.3138.2
25.1123.3 21.7124.4
32.916.5 31.917.0
a
Poly[(dGdC)I2;note different batch from that in Table 11.
Pdvl(dAdT)Iz. Pdy\(dAdT)Iz.
ZINC ION-DNA POLYMER INTERACTIONS
or lacking Z n 2 + ,was heated to 70°C, the spectral maximum shifted from 260 to 259 nm with a ca. 28% hyperchromicity. After cooling to room temperature, the sample had a spectrum similar to the initial spectrum. When the sample in the absence of Zn2' was heated and then Zn2+added, the spectral shape and the absorption intensity did not change. On cooling t o room temperature, exactly the same spectrum as the initial one was obtained. These results agree with more limited studies reported by Shin and Eichhorn." Additional Studies
The interactions between DNA and Zn2+were also examined by using several other methods. Effect of Added Mgz+. The dependence of the shape
of the T , vs RZn/base curve on Mg2+is shown for the Zn'+-SS DNA system in Figure 7. At low Mg2+concentrations, the shape of the curve followed the behavior discussed above, i.e., the T , first increased then decreased with increasing zinc ratio. However, a t higher Mg2+concentrations, the increase a t low zinc ratios was no longer evident. It is noteworthy that a t high RZn/base values, the T , values were similar, regardless of Mg2+concentration. It should also be noted that T , increased by 22°C on addition of 2 m M Mg2+a t RZnlbase= 0, whereas a t R Z n / b a s e = 20 the increase in T , is 3°C. For ML DNA, we also investigated the influence of Mg2+on the annealing effect of Zn2+.In the absence of added Mg2+, the hyperchromicity of the second melt was only 3.1%, compared t o 18.1% for the first melt a t RZn/base = 1.25. However, for a similar solution but with added Mg2+ ( R M g / b a s e = 10; the MgC12 increased the ionic strength of the solution by 1.5 X 10-3M), the hyperchromicity of the second melt was 20.8%, compared to 22.0% for the first melt. The same experiment, but with NaCl ( 1.5 X 10-'M) substituted for MgC12such that the ionic strength of these salts was equivalent, led to only a 3.7% hyperchromicity in the second melt, a result similar to that of the experiment without Mg2+.
-
-
Effect of Added Coz+. T h e delrenaturation of ML DNA and SS DNA a t Rco/base= 0, 2 , 4, and 10 was
studied (Table V ) . From the T,,, values, it can be seen that for the high GC content DNA (ML DNA), the effect of Co2+was similar to that of Zn". The T , increased a t low RCo/base value (1-2) and then decreased a t higher R C o / b a s e . However, for the lower GC content DNA (SS DNA), the effect of Co2+re-
31
sembles the well-known ionic strength effect,28 in which T, increases a t low R and then remains relatively constant. T h e Hyp % values for the first and the second melt are also listed in Table V. There was no appreciable improvement in the renaturation for the lower GC content SS DNA, but with higher Co2+concentration, the ML DNA renaturation increased. One might expect the DNA to renature completely a t higher R values. Viscosity Studies. Solution viscosity, which is very dependent on the length of dissolved DNA, is known t o be a more sensitive method than optical spectroscopy for detecting residual denatured features such as hairpins, loops, etc., which alter DNA length and hence solution vis~osity.~' Therefore, viscosity changes were studied to assess the influence of Zn2+ on the extent of renaturation of thermally denatured C T DNA, poly [ (dAdC ) (dGdT) 1 , and poly[ (dAdT) 1 2 . In the study of C T DNA, the RV values increased with increasing Zn2+concentration, i.e., the viscosities of the cooled DNA solutions, which had been denatured in the presence of Zn2+, gradually approached that of the native DNA solutions. However, the recovery peaked a t 85% RV at R = 8 and remained constant up t o R = 11, the highest ratio examined. For poly[ (dAdT)], in the absence of Zn2+,91% RV was obtained. While 100% RV was observed when the polymer was melted and annealed in the presence of Zn2+ a t R = 5. In the studies of poly [ (dAdC ) ( d G d T )1, when the solution was heated without Zn2+,the RV value measured was only 54%. Then Zn2+ was added to this solution. After a n hour a t 30°C, the solution relative viscosity increased t o 66% a t R = 1. In similar experiments, but with R = 5 and 10,85% RV and 98% RV were observed, respectively.
CD Study. T h e CD spectra of C T DNA and the effect of the addition of Zn2+ (Figure 8) and Mg'' were studied. The strength of the broad positive CD band a t 275 nm was diminished significantly on adding Zn2+to the DNA and the A,, increased to 280 nm. T h e negative band exhibited a slight intensity decrease. On addition of Mg2+, there was a n approximately 20% decrease in the positive CD band t o R = 45. The A, shifted to lower wavelength only slightly. An intensity decrease in the negative band was observed, but the decrease was only about half of that in the Zn2+ case. The CD studies of poly [ (dAdC ) ( dGdT ) ] will be discussed below.
32
JIA AND MARZILLI
1.57
a I.I 8
\
ABS
.78
.39
0 270
245
295
WAVELENGTH (rim)
1.5
b 1.13
ABS
.75
.38
0 WAVELENGTH (nm)
Figure 4. Electronic spectra of C T DNA (1.00 X 10-4M) in MES 00 buffer, pH 6.0. ( a ) -. . R = 0, 24°C; - - - -: R = 10, 24°C; - .- - - : R = 10,80"C; and - - - - - - - : R = 10, after cooling from 80°C to 24°C. ( b ) -: R = 0, 24°C; - - -: R = 0, 80°C; and - .- * : R = 0, after cooling from 80°C to 24°C. ( c ) -: R = 0, 24°C; - - -: R = 0, 80°C; - .- - - : Zn2+ then cooling to 24°C ( d ) -: R = 0, 24°C; - -: added ( R = 10) at 80"C, and ------: R = 0,80"C; - * - * : R = 0, cooling to 24°C; and - - - - : Zn2+added ( R = 10) at 24°C.
Additional Studies with Poly [ (dAdC)( dCdT)] In contrast to the simple absorbance changes during
the cooling of solutions of the other synthetic DNAs,
the results observed with poly [ (dAdC) (dGdT) ] were more complex in the absence of added Zn2+. At R = 0, the absorbance decrease reversed 30% of the original hserchromicity by 22°C and then
-
-
ZINC ION-DNA POLYMER INTERACTIONS
33
1.69
C 1.27
ABS .84
.42
0
WAVELENGTH (nm)
1.65
d I. 2 4
ABS
.83
.41
0 WAVELENGTH
Figure 4.
-
(nml
(Continued from the previous page.)
the absorbance decreased sharply by a similar amount until 10°C, the lowest temperature examined (Figure 3b). When it was present initially in the denaturation experiment, Zn2+ was able to completely reverse the hyperchromicity change when solutions were cooled even to 20°C a t R = 1. T h e ability of Zn2+t o reverse the hyperchromicity change is evident from a comparison of the Hyp 96
for the first and second denaturation experiments in Table 11. Of some interest, when Zn2' was added a t ambient temperature t o a solution of the denatured polymer, a significant absorption decrease could be followed easily. To further explore the effect of Zn2+ on poly [ (dAdC ) ( d G d T ) 1 , we carried out additional experiments, including comparisons with Mg2+and
34
JIA AND MARZILLI .83
a *62
--
\
ABS
WAVELENGTH (nm)
.91
I\
b
\
I \\
ABS
/
.45
I -
.23
0
WAVELENGTH (nm)
Figure 5. Electronic spectra of poly[ (dGdC)], (4.31 X 10-5M) in MES 00 buffer, pH 6.0. ( a ) : R = 0, 24°C; - - -: R = 1, 24°C; - - -.: R = 1, 90°C; and ----:R = 1, after cooling from 90°C to 24°C. ( b ) -: R = 0,24"C; - - -: R = 0,91"C; and - - - : R = 0, cooling to 24°C. ( c ) -: R = 0, 24°C; - - -: R = 0,90"C; - * - -: Zn2+added ( R = 1)at 90°C; and ----:then cooling to 24°C. ~
-
Co2+.First, the effect of metal ions on annealing of poly [ (dAdC ) (dGdT)] was examined. In the absence of added metal ions, the absorption of denatured
polymer solution did not change significantly with time at 10°C, although, as noted above, the hyper-
.-
chromicity change had not been completely reversed by cooling. At 30°C, a slow decrease in absorbance with time was observed. To discuss this result and similar results, we define a parameter, AA% = (Ao - A,)/(Ao - A m ) , where Ao is the absorbance im-
ZINC ION-DNA POLYMER INTERACTIONS
35
.84
C .63
ABS
.42
.21
0
WAVELENGTH (nm)
Figure 5.
(Continued from the previous page.)
mediately after cooling to the desired temperature (i.e., 10,20, or 30"C), A, is the absorbance expected if all the hyperchromicity had been recovered (i.e., the absorbance before denaturation), and A, is the absorbance at any given time. In the 10°C experiment, AA% was
SYNOPSIS
The adjacent GN7-M-GN7 cross-linking and adjacent G-M-G sandwich-complex models for DNA metal ion binding were evaluated both with native DNAs differing in GC content as well as with the synthetic polymers poly[ ( dGdC)Iz, poly[ ( dAdT)I2, and poly [ (dAdC) (dGdT) 1 . The effect of Zn2+was studied in depth, and limited studies were also performed with Co2+and Mg2+.The results were compared to the extensive information available on Cu2+binding to native DNAs and poly [ (dAdT)],. At high ratios of metal/ base ( R ) ,Zn2+caused all native DNAs to denature with the same melting temperature T,, 61°C. A similar pattern was reported previously for Cu2+,but the typical T , was 35°C. The extent of renaturation on cooling DNAs denatured in the presence of Zn2+ increased with GC content, as reported previously for CU". These results, together with previously reported similarities, strongly indicate that the DNA binding characteristics of the two cations are similar. By comparison of the T , values and hyperchromicity changes monitored at 260 and 282 nm, it is clear that, during thermal denaturation in the presence of Zn2+,both AT and GC regions were denatured, even at high R. The T , vs R profile for the native DNAs was typical. The rise at low R and subsequent decrease at high R were inversely and directly related, respectively, to GC content. Except for poly [ (dAdT)12, where T, increased with R , the other synthetic polymers exhibited the increase/decrease pattern. Poly[ (dAdC) (dGdT)] gave a T , value a t high R of 54°C. In the absence of Zn", this polymer exhibited little hypochromicity on cooling of denatured polymer. However, in the presence of Zn2+,nearly complete hypochromicity was observed, although the mid15°C at R = 10. These point of the cooling curve was lower than the T , value by characteristics were similar to those with native DNAs, although viscosity and CD studies suggested that the "renatured' polymer was not identical to the unheated polymer. Furthermore, addition of Zn2+after denaturation nearly completely reversed the absorption increase. This finding contrasts with those for native DNAs, where the Zn2+must be present during denaturation in order to reverse the absorption increase nearly completely on cooling. With some caveats, poly [ (dAdC ) (dGdT) ] appears to be a good model for native DNAs since its properties, including CD and uv changes on addition of Zn2+ to premelted and melted polymer, parallel those of the native polymers. Based on these findings and the discovery that Zn2+actually inhibits renaturation of poly [ (dGdC)I,, we believe adjacent G-M-G complexes are not the primary species responsible for the spectral changes in premelted DNAs, nor are they the principal species promoting renaturation. Three interrelated hypotheses to explain these phenomena were identified for further study as follows: ( a ) a kinetic effect-the metal ion promotes renaturation of denatured regions formed during thermal denaturation with metal ion present; ( b ) an inhibiting effect-the metal ion prevents the initial formation of conformations that otherwise inhibit renaturation; and ( c ) a CN3 binding effect-the metal ion lowers T , by stabilizing the denatured state through C binding. We speculate that such CN3 binding may account for the unexpectedly poor ability of Co2' both to lower T, and to promote renaturation.
--
-
Biopolymers. Val. 31, 23-44 (1991) 0 1991 J o h n Wiley & Sons, Inc.
CCC 0 ~ 6 - 3 5 2 5 / 9 1 / 0 1 ~ 2 3 - 2 2 $ 0 4 . 0 0
* To whom correspondence should be addressed.
t Also a member of the Winship Cancer Center.
23
24
JIA AND MARZILLI
INTRODUCTION T h e special role that zinc plays in nucleic acid biochemistry prompted us a few years ago t o develop methods for obtaining solid crystalline complexes formed between aquated Zn2+(hereafter, Zn '+) and nu~leotides.'-~ Out of such studies emerged a model' for the binding of Zn2+ to DNA. T h e model was based, in part, on a study of DNA polymers with C U ~ + . ~ Specifically, we suggested one of the ZnZ+DNA binding interactions might be a n N7,N7 intrastrand cross-link between two adjacent G residues.' This GN7-M-GN7 binding mode permits retention of GC base pairing and such adducts could act as nucleation points for renaturation of DNA strands when DNA is thermally denatured in the presence of Zn2+.Before the development of this model, Daune had proposed a chemically unreasonable G-Cu-G sandwich complex to serve as one of the nucleation points for DNA r e n a t ~ r a t i o nthis ~ ; model was based on studies of native DNAs with differing GC content. Our proposal is chemically more reasonable since it was based on information that we and others had obtained since the Daune proposal. For example, Reedijk and co-workers had shown that a n N7,N7 GG intrastrand cross-link with Pt retained its base pairing under appropriate conditions in a n oligon ~ c l e o t i d e Both .~ calculations and additional nmr studies confirmed the Reedijk results.6-8 Munchausen had demonstrated that DNA platinated a t low levels (now known to constitute primarily N7,N7 GG intrastrand cross-links9"0)could be renatured more readily than untreated DNA.".'* Such anticancer-active Pt ( I1 ) compounds cannot be considered as good models for Zn2+;P t ( I 1 ) forms inert square-planar complexes, whereas Zn '+ forms labile octahedral or tetrahedral complexes. Zimmer e t al. have postulated that when it binds to G residues of DNA, Zn2+forms a n N7,PO4 chelate.I3 On the basis of studies with Cu", Daune has also proposed such a chelate as another major species in addition to the sandwich specie^.^ This chelate binding mode proposed for these labile metal species has now been found for nucleotide complexes of Pt (11) anticancer drug^'^,'^ as well as for nucleotide complexes of anticancer drugs containing Mo16 and R u . ' ~A knowledge of the binding of such inert metal centers to nucleotides, oligonucleotides, and DNA can be used judiciously to assess proposals for the complex situation of the binding of labile metal ions, such as Zn '+,to polymers. Information on labile Cu2+ complexes may be more relevant than that on P t ( I 1 ) to Zn2+ complexes. However, Cu2+ usually prefers ( a ) square-
planar geometry, ( b ) a five-coordinate geometry, or ( c ) a n octahedral geometry with significant JahnTeller distortions. Additionally, paramagnetic Cu (11) complexes are difficult t o study by the informative method of nmr spectroscopy since the contact contribution dominates T2 relaxation and contributes significantly to TI relaxation." We felt additional studies on polymers with Zn2+ were needed since the information in the literature was much less complete than that for Cu2+.Such studies are important in the design of sequences of synthetic oligonucleotides for nmr studies with Zn2+,which may prove useful in explaining Zn2+DNA binding and the effects of Zn2+ on DNA properties. Furthermore, Zn2+ is a diamagnetic species, and nmr studies with Zn2+ should be less complicated than those with Cu2+. Based on comparative effects of metal ions on DNA denaturation, the following series was established: Mg2+,Co*+, Ni2', M n 2 + ,Zn2+, Cd2', and cu2+ .19 Eichhorn has emphasized that there should be no dichotomy between the binding modes of one metal ion and those of another nearby in the series, but that these effects ( a n d hence the binding interactions) should vary to a matter of degree.lg T h e series appears to reflect increasing affinity for binding to base N's vs phosphate 0's; differences in the effects of Cu2+ and Zn2+ on DNA properties have been attributed t o the stronger binding of Cu2+to the base N's.'' More recently, in a valuable assessment of metal binding characteristics, Martin pointed out that, although the binding constants t o N7 of purine nucleosides are a n order of magnitude greater for Cu2+than for Zn2+,the relative binding toward N7 of the common nucleosides follows a similar order: adenosine inosine < guanosine.'l In this report, we further assess the similarity of binding modes of these two labile ions with studies of Zn2+binding t o native and synthetic DNAs similar to those reported previously with Cu2' .4,22*23 Indeed, many of our findings support the relationship between Zn2+and C u 2 + .Furthermore, we have also tested the models with poly[ (dAdC) ( d G d T ) ] , which can form neither intrastrand nor interstrand adjacent GG cross-links. With this polymer, as well as some native DNAs, we have further explored the contrast between the effects of Co2+and Zn2+(i.e., their separation in the above series) compared to the typically similar chemistw of these metal ions.
-
EXPERIMENTAL Materials
The following DNAs were selected for study: Micrococcus lysodeikticus ( M L D N A ) , 72% GC; Esch-
ZINC ION-DNA POLYMER INTERACTIONS
erichia coli ( E C DNA) 51% GC; Salmon sperm (SS DNA ) 42% GC; Clostridiumperfringens ( C P DNA) 30% GC; and calf thymus ( C T DNA) 44% GC. The synthetic polymers selected were poly [ (dGdC) 12, poly [ ( dAdT ) 12, poly [ ( dAdC ) ( d G d T ) ] , a n d poly[ (dIdC)J 2 . Native DNA solutions were supplied by Dr. W. D. Wilson of Georgia State University. These samples had been sonicated to average 600 base pairs in length; the solutions were dialyzed against piperazine- N,N'-bis ( 2-ethanesulfonic acid) ( P I P E S ) 10 buffer (0.01M PIPES, 10-5M EDTA, 0.10M N a N 0 3 a t p H 7.00) and passed through a 0.22-p Millipore filter.24 T h e synthetic polymers, purchased from Pharmacia, were used without sonication. ZnClz and MgC12* 6 H 2 0 (from Aldrich Chemical Co.) and CoClz 6 H 2 0 (from Fisher Scientific Co.) were reagent grade and were used without further purification.
-
Instrumentation
All thermal denaturation and renaturation experiments were performed on a Perkin-Elmer Lambda 3B UV/VIS spectrophotometer equipped with a thermally jacketed five-cell compartment and a temperature probe, which can be placed directly in one of the five sample solutions. The temperature was controlled by a temperature bath with a programmer capable of controlling the rate of temperature change. The spectrophotometer was interfaced with an Apple I1 computer through a software program designed to record the sample temperature and its corresponding absorption as a function of temperature change.25The uv scans were also obtained from the same type of instrument. Viscosity studies were performed with a Cannon-Ubbelohde viscometer in a thermostated water bath a t 30.00"C. CD spectra were acquired with an AVIV model 62DS that is equipped with a thermally jacketed cell compartment. Methods
In a typical denaturation ( "melting") experiment, five samples were run simultaneously in sealed cuvettes with a 1-cm path length. T h e first cuvette was a blank and the temperature probe was placed in this cuvette. I t contained the MES 00 buffer solution [ 0.01M 2- (N-morpholino) ethanesulfonic acid a t p H 6.00 and passed through a 0.22-p Millipore filter]. No extra salt was added and the salt introduced by p H adjustment was approximately 5 X 10-.3M.T h e other four cuvettes contained solutions of DNA (4.2 X 10-5-1 X 10-4M,diluted from the stock solutions described above) with different
25
Zn2+concentrations in MES 00 buffer. A 0.1 M Zn2+ stock solution was prepared by weighing a n appropriate amount of ZnClz into a 50 mL volumetric flask, adding 2 small drops of 5.5M HC1, and then deionized water. The final p H = 2.63 and the Zn2+ concentration measured by atomic absorption spectroscopy was 0.1OM. The total volume of the solution in each cuvette was 2.0 mL. The temperature was increased for melting and decreased for cooling experiments at the rate of 0.5"C/min. T h e melting temperature T , of DNA was determined by plotting the first derivative, AAIAT vs T . Tmis equal to the temperature at which AAIAT is a t a maximum. Cooling experiments were performed immediately after completion of the melting experiment. Except where noted, the absorbance was monitored a t 260 nm. T h e percent hyperchromicity (Hyp % ) was calculated according t o the equation Hyp % = (A, - A , ) / A f X 100, where A f is the final absorbance after melting and A, is the initial absorbance before melting. In the experiments on the time dependence of polymer annealing on the addition of metal ion, the polymer, without metal ion added, was melted. Then the solution was cooled a t O.S"C/min to 10.0, 20.0, and 30.0"C in separate experiments. At this point, the change in absorbance a t 260 nm was monitored as a function of time, both in the absence and in the presence of added metal ion. C T DNA, poly[(dAdT)], and poly[(dAdC)( d G d T ) ] were used for viscosity studies. Each point in the viscometric titrations represents a t least three measurements with a n average deviation of less than 0.10 s. For denatured DNA, the reservoir of the viscometer containing the sample was placed in a water bath at 75°C for 15 min, and then the viscometer was returned to the original water bath at 30.00"C for the viscometric measurements. The calculation of absolute viscosity was based on the equation:
where r is the time reading in seconds and C is the DNA concentration in bases. The relative viscosity ( RV ) is defined here as the ratio ?denatured/ qnatlve. CD spectra were obtained a t 25°C. When denatured DNA was needed, the cell compartment temperature was raised to 75°C for 10 min, then decreased to 25°C a t l"C/min. T h e MES buffer for the CD studies was five times more dilute than that used for the uv melting experiments in order to improve results in the 200-230-nm region. Two CD spectral scans were recorded a t 1nm/s and averaged. T h e baseline was corrected using software supplied by the manufacturer.
26
JIA AND MARZILLI
increase in R , several phenomena could be observed from the melting curves as follows: ( a ) the melting curve became sharp; ( b ) the hyperchromicity typically decreased and ( c ) the melting temperature increased at first, but then decreased. At high R values, usually R > 10, and high GC content DNA, the solution absorption decreased as the temperature was raised well above the T,. The T, values of four native DNAs and four synthetic polymers were measured (Table I ) and plotted vs R ( Figure 2 ) . Of some interest, the T,'s of native DNAs (Figure 2a) a t high R values were very similar regardless of DNA base-pair composition and sequence. At low R values, T , increased until R 2; except for ML DNA, where the maximum was a t R 1. The values then decreased to a similar range
0.68-
A260nm
-
0.58
-
0.48 40
I
50
I
60
I
70
-
I
80
Thermal Denaturation
A typical example of melting curves as a function of Zn2+base ratio ( R ) is shown in Figure 1. With
small 2°C increase a t R = 0.3. However, the study of this polymer was limited because of aggregation, even a t R = 2. Poly [ (dAdC) (dGdT)] exhibited a T , vs R curve similar to those of the native DNAs,
Table I DNA Melting Temperatures ("C) at 260 nm with Different RznmaseValues
RZIllbase
0 0.3 0.5 0.8 1.o 1.2 1.5 1.8 2 4 6 10 12 14 15 18 20
DNA GC%
GC a 100
ML
EC 51
AC/GT~
ss
CP
A T'
72
50
42
30
0
ICd 0
82.6 84.7
67.2
60.8
58.6
58.8
49.5
31.7
26.7
Poly[(dGdC)]Z. Poly [ (dAdC)(dGdT)I. P~ly[(dAdT)]z. Poly[(dIdC)]z. This data was obtained from 282 nm.
75.2
69.7 66.2 64.5 62.0 61.1 60.7 59.8 59.6
33.3
60.8
75.4 71.7 70.0 68.5 67.4 67.0 60.0'
69.3 65.4 61.9
60.1
67.2
67.2
64.4
47.0
37.0
63.3 60.0 58.1 56.3
70.5 68.6 66.1 63.5 62.0
66.7 66.1 64.0 62.7 61.3
50.8 53.3 54.7 55.9
39.2 36.5 34.2 31.5 30.8
61.4 60.7 60.1
61.2 60.7 60.0
53.9
29.7 56.0
28.6
ZINC ION-DNA POLYMER INTERACTIONS
a
Tm?C
45
! 0
I 5
10
15
20
R(Zn/Base)
significant decreases in hyperchromicity accompanied the cooling, we performed a second melt with the solution. These data are also collected in Table 11. We assume that the relative values of the hyperchromicity in the first and the second melts are indicative of the extent of renaturation in the cooling process. Relatively closer hyperchromicity values of the second melt t o the first melt were found for DNAs with higher GC content. For any given DNA, the hyperchromicity of the first and the second melts became more similar a t higher Zn2+concentration. For ML DNA, the renaturation was almost complete a t R = 2, but for SS DNA, maximum renaturation required R = 10. Obviously, the extent of the renaturation greatly depends on two factors: GC content and Zn2+concentration. Annealing Experiments
b
Tm,"C
Examples of cooling curves a t different R values for one native DNA, C P DNA, and one synthetic polymer, poly [ (dAdC) ( d G d T ) ] , are shown in Figure 3. For C P DNA (Figure 3 a ) , a biphasic cooling profile was observed a t R values of 4 and 6: the absorption did not decrease significantly during the first 10°C decrease in temperature. Then there was a sharp decrease in absorption, which was followed by a gradual decrease. There were slight differences between the R = 4 and 6 studies: the absorption decrease of the first phase increased with Zn2+concentration. At 20°C, 60% of the initial absorption increase was reversed. At low R values ( 0 and 2, figure not shown), the cooling curves exhibit a smooth monophasic decrease in absorption, and a t 20°C the absorption had decreased by 50% of the initial increase on denaturation. For DNAs with higher GC content, somewhat different cooling profiles were observed compared to the low GC content C P DNA. For SS DNA, with a n average GC content, biphasic behavior was observed a t low R values but not high R values. For the high GC content DNA, ML DNA, biphasic cooling behavior was not observed (figure not shown). For poly [ (dGdC ) I,, poly [ (dAdT) 12, and poly[ (dAdC ) ( dGdT) 1, a t R 2 1, the cooling curves are monophasic with complete reversal of hyperchromicity (for poly [ (dGdC) l2 in the presence of Zn2+, the recovery of hyperchromicity was not complete a t longer wavelength, see below). The situation for poly [ (dAdC ) ( d G d T )] a t very low Zn2+content (or in the absence of Zn2+)is complex and is discussed below. However, for the other two polymers, the cooling profiles were monophasic, even in the absence of added Zn2+.The temperature of the mid-
-
R(Zn/Base)
Figure 2. Relationship between T, and RZnlbase.( a ) Native DNAs: -, ML DNA, * * * , EC DNA; - - - -,
--
SS DNA; and - - -, C P DNA. ( b ) Synthetic polymers: . . -. poly[ (dGdC)],; * * * ,polyj (dAdC) ( d G d T ) ]; - - - -,poly [ (dAdT) I,; - - -, poly [ (dIdC) 1,; and -, ML DNA.
-
27
-
-
but the values of T , a t high R were lower. Also, the maximum in T , occurs at R = 1. In contrast, the T , vs R curve of poly[ (dAdT)],, which lacks GC base pairs, never decreased. Poly [ (dIdC) 1 2 had much lower T,'s than any other DNA or polymer studied here; however, the shape of the T, vs R curve for this polymer follows the typical increase / decrease behavior of the native DNAs. It is instructive t o compare the Hyp % for both native and synthetic DNAs as a function of added Zn" (Table 11). In most of the experiments where
-
28
JIA AND MARZILLI
Table I1 Hyperchromicity Data (70)of First/Second Melt of DNAs at 260 nm
R
GC a
0 0.5 1 1.5 2 4 6 10
16.4/16.2 14.0/13.2 13.3/7.2 6.216.0
ML 28.1/-
EC 28.5/-
24.9/24.7/22.4 23.1/21.0 23.3/20.7 21.9/19.1
ss
AC/GT~ 24.2/5.1
CP
25.1/-
21.8/19.1 26.1/25.9/24.1/-
20.5/19.3/19.0/18.0 18.7/18.0
27.9/5.7 24.5/19.3 22.6/18.6 20.4/18.6
AT
24.6/-
31.4/30.6
27.1/-
33.0/32.4
25.7/3.8 21.7/6.3 18.9/11.3 17.9/13.1
32.9/33.4 31.9/31.8 31.9/33.4 31.2/30.9
Poly[(dGdC)],. Poly[ (dAdC)(dGdT)1. ' Poly[(dAdT)]z. a
point of the cooling curve for most polymers and native DNAs decreased with increasing R (see Table 111). For example, for ML DNA, these midpoints were 50, 32, 23, and 21°C for R = 2, 4, 6, and 10, respectively. However, for poly [ (dAdT)12,the midpoint temperature of the cooling curve was similar t o the T,.
increase/decrease pattern of Hyp % was seen also for the high GC content DNAs. However, for C P DNA and poly [ (dAdT) 12,the Hyp % increased with increasing Zn2+ concentration. It should be noted that, although there was no hyperchromicity found a t 260 nm at R = 2 for poly [ ( dGdC)I2,the melting could be monitored at 282 nm (see Table IV) .
Denaturation Monitored at 282 nm
Absorption Spectra
In order t o assess the differences in interaction between AT and GC base pairs with Zn2+, the approach used was based on the dispersion of the hyperchromic effect with wavelength. This method was used originally t o analyze DNA base composition.26 Changes in absorbance and in hyperchromicity could be resolved into components arising from AT and GC base pairs. At 260 nm, larger absorbance changes on melting were found for AT-rich DNA than for GC-rich DNA. Since almost no absorbance change was found for poly [ (dAdT)], a t 282 nm, the melting behavior of GC base pairs can be more directly monitored at this wavelength. Hyperchromicity data for seven DNAs and polymers at 260 and 282 nm are listed in Table IV. T h e hyperchromicity changes for poly [ (dGdC) l2 were 4 times greater a t 282 than at 260 nm. I n contrast, absorption changes for poly [ ( dAdT)Izwere too small t o be seen a t 282 nm unless Zn2+was present. For DNAs with intermediate GC content, the relative hyperchromicity a t 260 and 282 nm reflected the expected trend. The Hyp 76 a t 260 nm decreased with increasing Zn2' content for DNAs with high GC content, such as poly[ (dGdC)I2, ML DNA, and poly[ (dAdC)(dGdT) ] . For CP DNA and poly [ (dAdT) 12,a small increase in Hyp % a t 260 nm was observed on addition of Zn2+ t o R = 1. T h e Hyp % decreased a t higher R values. On the other hand, at 282 nm this
There are three important points ( a , 6, c ) where metal ions could be added as shown below:
-
heat cool We examined the uv spectral properties of C T DNA, poly [ ( d G d C ) 12, poly [ ( d A d C ) ( d G d T ) 1 , a n d poly [ (dAdT ) l2 with no metal added and in separate experiments with Zn2+ added a t each of the three points. For each polymer, the spectrum was recorded a t room temperature in the absence and presence of Zn2+at R = 10, except for poly [ (dGdC) 12, where R = 1. No significant change in spectrum was observed (Figures 4a, 5a, and 6 a ) . In the absence of Zn2+,the C T DNA sample was heated to 80°C over 10-15 min and the spectrum was recorded (Figure 4b). T h e spectral shape did not change, with one band a t 257 nm and a slight red shoulder. However, there was a n approximate 27% hyperchromicity with a slightly greater increase for the unresolved shoulder. As the sample was cooled to room temperature over a 40-min period, the spectral shape and position did not change, but there was a decrease in intensity throughout, which reversed 20% of the intensity increase induced by heating (Figure 4b).
-
ZINC ION-DNA POLYMER INTERACTIONS
a
032
29
obtained when the sample was heated in the presence of Zn2'. However, on cooling, the absorption maximum had a 3-nm red shift and an 40% higher intensity relative to the initial room temperature spectrum a t 260 nm. An even higher relative intensity was observed a t longer wavelength, e.g., 50% at 272 nm (Figure 4c ) . Finally, a C T DNA sample was heated and cooled as above, and then Zn2+was added (Figure 4d). The spectral shape was similar t o that a t high temperature but the absorption reversed 60% of the intensity increase. The final Hyp % a t 260 and 272 nm were 1 2 and 14%, respectively. In experiments that parallel the absorption spectral studies, we wished to determine the effects on denaturation (monitored a t 260 nm) of adding Zn2+ at points b and c instead of a. When C?' DNA a t 5 X 10-5M (higher concentrations than in Table I ) was denatured, the Hyp % = 30. For C T DNA, the T , a t R = 10 (point a)was 64°C with Hyp % = 28%. However, for a n R = 10 solution with Zn2+added to denatured C T DNA (point b ) there was a n initial decrease in absorbance (19% of the initial absorbance increase), consistent with the uv plot (Figure 4c). On cooling, the absorbance remained constant until 43"C, whereupon a n additional 57% of the initial absorbance increase was reversed. After 30 min or 17 h, reheating led to a Hyp C b = 16 and a T, of 37°C. Likewise, the cooling curves in the presence of added Zn2+were similar to the heating curves. If the C T DNA was cooled t o 20°C and then Zn2' ( R = 10) added (point c ), the absorbance decreased (cf. Figure 4 d ) . A Hyp % = 15 and a T , = 37°C were observed when the DNA was reheated. During the reheating, the absorbance increase was complete a t 42"C, with only a slight additional increase to 70°C. A similar result was obtained when this solution was cooled and then reheated after 17 h. These results suggest that when the Zn is added a t point b or c , the hypochromicity changes do not correspond to the generation of any appreciable stretches of native DNA; otherwise, the T , values would have been much higher. The spectrum of poly [ (dGdC ) l2 has, , ,A a t 254 nm a t room temperature. The band tails to longer wavelength but no shoulder is evident. On heating to 90°C over 10-15 min, a pronounced shoulder emerged a t 275 nm with an intensity increase a t all wavelengths (Figure 5 b ) . When the sample was cooled t o room temperature, the shoulder disappeared and the intensity increase was reversed by 90%. The spectrum resembled the initial spectrum before heating (Figure 5 b ) . When a solution of poly[ (dGdC)], was heated to 90°C in the pres-
-
-
T. OC
b
036 -
-
-
0
2 6 10
!
.
,
.
.
,
20
30
I
.
40
,
50
.
.
, 60
70
TPC Figure 3. Melting and cooling curves of a native DNA and a synthetic polymer. ( a ) C P DNA (4.33 X 10-5M)in MES 00 buffer, pH 6.0. * ,R = 6 melting curve; -, R = 4 ; and - - - - , R = 6 cooling curves ( b ) Poly[ (dAdC)(dGdT)](4.11 X 10-5M) in MES 00 buffer, pH 6.0. --, R = 0 melting curve. * * * , R = 0. - - - R = 6; and - - -, R = 10 cooling curves.
.
---
- -
3
1
In the R = 10 sample, the absorption increased on heating a s above with a red shift of 5 nm obscuring the shoulder. On cooling as above, the resulting spectrum was similar to the initial spectrum of unheated C T DNA (Figure 4 a ) . When the sample was heated to 80°C in the absence of Zn2+and then Zn2+added, the absorption spectrum changed to superpose almost exactly that
-
-
'+
-
-
30
JIA AND MARZILLI
Table I11 Temperatures ("C) of the Midpoint of the Steepest Absorption Decrease on Cooling DNA Solutions DNA
GC"
0 0.5
77 60 48.5
1 2 4 6 10
ML
ss
AC/GTb
CP
AT
20
31.5
56
51
50 32 23 21
54.5 50.5 43
44 41
52 48 40
54.5 56
Poly[(dGdC)],. Poly[(dAdC)(dGdT)]. ' P~ly[(dAdT)]p. a
ence of Zn2' ( R = 1, point a), the pronounced shoulder was also observed (Figure 5 a ) . However, as the sample was cooled the shoulder did not disappear completely. The intensity increase at 275 nm was reversed by only 70% after 30 min. An R = 0 sample was heated to 90°C and then Zn2+to R = 1 was added (point b ) . The intensity at A, 254 nm decreased, but that at the 275-nm shoulder did not decrease (Figure 5c). The shoulder became slightly broader. After cooling to room temperature, the shoulder was still present and the initial intensity increase on heating was reversed by only 50%. For poly [ (dAdC) (dGdT) 1 , the phenomena observed were very similar to those found with C T DNA. In the absence of Zn2+,the sample was heated to 80"C, the spectral shape (A,, 258 nm, shoulder 272 nm) did not change, but there was 26% hyperchromicity at the spectral maximum (Figure 6b). When the sample was cooled to room temperature, the spectrum still resembled that at high temperature and the intensity decreased only 5%. In the R = 10 sample (point a),heating to 80°C
-
-
-
-
-
Table IV
-
induced 21% hyperchromicity at 258 nm and a ca. 2-nm red shift of the spectral maximum, which obscured the shoulder at 272 nm (Figure 6a). On cooling the sample, the spectrum recorded was similar to the initial spectrum with a slightly higher intensity at longer wavelengths. When the sample was heated to 80°C and then Zn2+added (point b ) ,the spectrum exhibited a red shift of 2 nm, and a shoulder at 272 nm was observed (Figure 6c). The spectrum was similar to that obtained when Zn2+was added at point a. On cooling, the resulting spectrum was similar to the initial spectrum at room temperature except for 5% hyperchromicity at 258 nm and a 10% hyperchromicity at 272 nm. The R = 0 sample was heated and cooled, and Zn2+was then added (point c ) . Following the absorption, intensity decrease with time demonstrated that the final spectrum was very close to the initial one at room temperature (Figure 6d). In comparison to the other polymers studied, the behavior was quite straightforward for poly[ ( dAdT)Iz.When the sample, either containing
-
-
-
-
Hyperchromicity Data (%) at 260/282 nm AT '
R
GC a
ML
AC/GTb
CP
0 0.5 1 1.5 1.8 2 4
14.7151.8 13.5152.7 12.5153.0 10.3151.6 5.6144.2 -126.9
28.1142.4
24.2132.1
24.6118.4
31.410
24.9144.6
21.8140.0
27.1120.0
33.0125
24.1143.2 23.1/42.1
20.5140.0 19.3138.2
25.1123.3 21.7124.4
32.916.5 31.917.0
a
Poly[(dGdC)I2;note different batch from that in Table 11.
Pdvl(dAdT)Iz. Pdy\(dAdT)Iz.
ZINC ION-DNA POLYMER INTERACTIONS
or lacking Z n 2 + ,was heated to 70°C, the spectral maximum shifted from 260 to 259 nm with a ca. 28% hyperchromicity. After cooling to room temperature, the sample had a spectrum similar to the initial spectrum. When the sample in the absence of Zn2' was heated and then Zn2+added, the spectral shape and the absorption intensity did not change. On cooling t o room temperature, exactly the same spectrum as the initial one was obtained. These results agree with more limited studies reported by Shin and Eichhorn." Additional Studies
The interactions between DNA and Zn2+were also examined by using several other methods. Effect of Added Mgz+. The dependence of the shape
of the T , vs RZn/base curve on Mg2+is shown for the Zn'+-SS DNA system in Figure 7. At low Mg2+concentrations, the shape of the curve followed the behavior discussed above, i.e., the T , first increased then decreased with increasing zinc ratio. However, a t higher Mg2+concentrations, the increase a t low zinc ratios was no longer evident. It is noteworthy that a t high RZn/base values, the T , values were similar, regardless of Mg2+concentration. It should also be noted that T , increased by 22°C on addition of 2 m M Mg2+a t RZnlbase= 0, whereas a t R Z n / b a s e = 20 the increase in T , is 3°C. For ML DNA, we also investigated the influence of Mg2+on the annealing effect of Zn2+.In the absence of added Mg2+, the hyperchromicity of the second melt was only 3.1%, compared t o 18.1% for the first melt a t RZn/base = 1.25. However, for a similar solution but with added Mg2+ ( R M g / b a s e = 10; the MgC12 increased the ionic strength of the solution by 1.5 X 10-3M), the hyperchromicity of the second melt was 20.8%, compared to 22.0% for the first melt. The same experiment, but with NaCl ( 1.5 X 10-'M) substituted for MgC12such that the ionic strength of these salts was equivalent, led to only a 3.7% hyperchromicity in the second melt, a result similar to that of the experiment without Mg2+.
-
-
Effect of Added Coz+. T h e delrenaturation of ML DNA and SS DNA a t Rco/base= 0, 2 , 4, and 10 was
studied (Table V ) . From the T,,, values, it can be seen that for the high GC content DNA (ML DNA), the effect of Co2+was similar to that of Zn". The T , increased a t low RCo/base value (1-2) and then decreased a t higher R C o / b a s e . However, for the lower GC content DNA (SS DNA), the effect of Co2+re-
31
sembles the well-known ionic strength effect,28 in which T, increases a t low R and then remains relatively constant. T h e Hyp % values for the first and the second melt are also listed in Table V. There was no appreciable improvement in the renaturation for the lower GC content SS DNA, but with higher Co2+concentration, the ML DNA renaturation increased. One might expect the DNA to renature completely a t higher R values. Viscosity Studies. Solution viscosity, which is very dependent on the length of dissolved DNA, is known t o be a more sensitive method than optical spectroscopy for detecting residual denatured features such as hairpins, loops, etc., which alter DNA length and hence solution vis~osity.~' Therefore, viscosity changes were studied to assess the influence of Zn2+ on the extent of renaturation of thermally denatured C T DNA, poly [ (dAdC ) (dGdT) 1 , and poly[ (dAdT) 1 2 . In the study of C T DNA, the RV values increased with increasing Zn2+concentration, i.e., the viscosities of the cooled DNA solutions, which had been denatured in the presence of Zn2+, gradually approached that of the native DNA solutions. However, the recovery peaked a t 85% RV at R = 8 and remained constant up t o R = 11, the highest ratio examined. For poly[ (dAdT)], in the absence of Zn2+,91% RV was obtained. While 100% RV was observed when the polymer was melted and annealed in the presence of Zn2+ a t R = 5. In the studies of poly [ (dAdC ) ( d G d T )1, when the solution was heated without Zn2+,the RV value measured was only 54%. Then Zn2+ was added to this solution. After a n hour a t 30°C, the solution relative viscosity increased t o 66% a t R = 1. In similar experiments, but with R = 5 and 10,85% RV and 98% RV were observed, respectively.
CD Study. T h e CD spectra of C T DNA and the effect of the addition of Zn2+ (Figure 8) and Mg'' were studied. The strength of the broad positive CD band a t 275 nm was diminished significantly on adding Zn2+to the DNA and the A,, increased to 280 nm. T h e negative band exhibited a slight intensity decrease. On addition of Mg2+, there was a n approximately 20% decrease in the positive CD band t o R = 45. The A, shifted to lower wavelength only slightly. An intensity decrease in the negative band was observed, but the decrease was only about half of that in the Zn2+ case. The CD studies of poly [ (dAdC ) ( dGdT ) ] will be discussed below.
32
JIA AND MARZILLI
1.57
a I.I 8
\
ABS
.78
.39
0 270
245
295
WAVELENGTH (rim)
1.5
b 1.13
ABS
.75
.38
0 WAVELENGTH (nm)
Figure 4. Electronic spectra of C T DNA (1.00 X 10-4M) in MES 00 buffer, pH 6.0. ( a ) -. . R = 0, 24°C; - - - -: R = 10, 24°C; - .- - - : R = 10,80"C; and - - - - - - - : R = 10, after cooling from 80°C to 24°C. ( b ) -: R = 0, 24°C; - - -: R = 0, 80°C; and - .- * : R = 0, after cooling from 80°C to 24°C. ( c ) -: R = 0, 24°C; - - -: R = 0, 80°C; - .- - - : Zn2+ then cooling to 24°C ( d ) -: R = 0, 24°C; - -: added ( R = 10) at 80"C, and ------: R = 0,80"C; - * - * : R = 0, cooling to 24°C; and - - - - : Zn2+added ( R = 10) at 24°C.
Additional Studies with Poly [ (dAdC)( dCdT)] In contrast to the simple absorbance changes during
the cooling of solutions of the other synthetic DNAs,
the results observed with poly [ (dAdC) (dGdT) ] were more complex in the absence of added Zn2+. At R = 0, the absorbance decrease reversed 30% of the original hserchromicity by 22°C and then
-
-
ZINC ION-DNA POLYMER INTERACTIONS
33
1.69
C 1.27
ABS .84
.42
0
WAVELENGTH (nm)
1.65
d I. 2 4
ABS
.83
.41
0 WAVELENGTH
Figure 4.
-
(nml
(Continued from the previous page.)
the absorbance decreased sharply by a similar amount until 10°C, the lowest temperature examined (Figure 3b). When it was present initially in the denaturation experiment, Zn2+ was able to completely reverse the hyperchromicity change when solutions were cooled even to 20°C a t R = 1. T h e ability of Zn2+t o reverse the hyperchromicity change is evident from a comparison of the Hyp 96
for the first and second denaturation experiments in Table 11. Of some interest, when Zn2' was added a t ambient temperature t o a solution of the denatured polymer, a significant absorption decrease could be followed easily. To further explore the effect of Zn2+ on poly [ (dAdC ) ( d G d T ) 1 , we carried out additional experiments, including comparisons with Mg2+and
34
JIA AND MARZILLI .83
a *62
--
\
ABS
WAVELENGTH (nm)
.91
I\
b
\
I \\
ABS
/
.45
I -
.23
0
WAVELENGTH (nm)
Figure 5. Electronic spectra of poly[ (dGdC)], (4.31 X 10-5M) in MES 00 buffer, pH 6.0. ( a ) : R = 0, 24°C; - - -: R = 1, 24°C; - - -.: R = 1, 90°C; and ----:R = 1, after cooling from 90°C to 24°C. ( b ) -: R = 0,24"C; - - -: R = 0,91"C; and - - - : R = 0, cooling to 24°C. ( c ) -: R = 0, 24°C; - - -: R = 0,90"C; - * - -: Zn2+added ( R = 1)at 90°C; and ----:then cooling to 24°C. ~
-
Co2+.First, the effect of metal ions on annealing of poly [ (dAdC ) (dGdT)] was examined. In the absence of added metal ions, the absorption of denatured
polymer solution did not change significantly with time at 10°C, although, as noted above, the hyper-
.-
chromicity change had not been completely reversed by cooling. At 30°C, a slow decrease in absorbance with time was observed. To discuss this result and similar results, we define a parameter, AA% = (Ao - A,)/(Ao - A m ) , where Ao is the absorbance im-
ZINC ION-DNA POLYMER INTERACTIONS
35
.84
C .63
ABS
.42
.21
0
WAVELENGTH (nm)
Figure 5.
(Continued from the previous page.)
mediately after cooling to the desired temperature (i.e., 10,20, or 30"C), A, is the absorbance expected if all the hyperchromicity had been recovered (i.e., the absorbance before denaturation), and A, is the absorbance at any given time. In the 10°C experiment, AA% was
