Local Alteration of Ionic Strength in a Nucleosome Core Particle and Its Effect on N7-Methyl-2′-deoxyguanosine Depurination

Abstract

Positively charged N-terminal histone tails play important roles in maintaining the nucleosome (and chromatin) structure and function. Charge alteration, including those imposed by post-translational modifications, impacts chromatin dynamics, protein binding, and the fate of DNA damage. There is evidence that N-terminal histone tails affect the local ionic environment within a nucleosome core particle (NCP), but this phenomenon is not well understood. Determining the modulation of the local ionic environment within an NCP by histone tails could help uncover the underlying mechanisms of their functions and effects. Utilizing bottom-up syntheses of NCPs containing wild-type or mutated histones and a fluorescent probe that is sensitive to the local ionic environment, we show that interaction with positively charged N-terminal tails increases the local ionic strength near nucleosomal DNA. The effect is diminished by replacing positively charged residues with neutral ones or deleting a tail in its entirety. Replacing the fluorescent probe with the major DNA methylation product, N7-methyl-2′-deoxyguanosine (MdG), revealed changes in the depurination rate constant varying inversely with local ionic strength. These data indicate that the MdG hydrolysis rates depend on and also inform on local ionic strength in an NCP. Overall, histone tail charge contributes to the complexity of the NCP structure and function by modulating the local ionic strength.

Graphical Abstract

INTRODUCTION

Nucleosome core particles (NCPs), the basic units of eukaryotic chromatin, consist of 145–147 bp of DNA wrapped ~1.6 to 1.7 times around an octameric core of histone proteins (H2A, H2B, H3, and H4). The N-terminal histone tails are highly positively charged and rich in arginine and lysine that are important for nucleosome stability.^1–4 Internucleosomal interactions between the N-terminal histone H4 tail and H2A/H2B acidic patch affect the overall structure of chromatin.^5,6 In addition, histone tails impact crucial interactions with nucleosome-binding proteins through direct contact and by modulating the accessibility of nucleosomal DNA.^7,8 Post-translational modifications (PTMs) and oncohistone mutations on histone tails that change the effective positive charge, such as lysine acetylation and Arg to Ala mutations, further complicate nucleosome dynamics, chromatin compaction, and the interactions of chromatin-associated proteins with DNA.^7,9–11 Despite the importance of electrostatic interactions, there is a gap in our knowledge concerning the effects of histone tails on the local ionic environment of the nucleic acids.

Ion counting experiments revealed that while the overall difference in the electrostatic field between free DNA and an NCP is small, deleting the histone H3 N-terminal tail results in a global increase in the negative electrostatic field around a nucleosome.¹² Computational studies predict the variation in the local ionic environment surrounding DNA within an NCP, especially with respect to sites of histone tail–DNA interactions.¹ Variations in the local ionic environment could affect DNA–DNA interactions between gyres within a nucleosome, as well as DNA–protein interactions with chromatin-associated proteins. In addition, the ionic environment can significantly affect the activation barrier of chemical reactions that involve charged species and/or transition states. Reaction rates can increase or decrease depending upon how the energies of the corresponding starting materials and the transition state of the rate-determining step respond to changes in the ionic environment.

The reactivity of damaged DNA (lesions) in NCPs is affected by the proximal N-terminal histone tails.^13–17 In addition to direct reactions with electrophilic lesions, N-terminal histone tails affect the depurination rate constants (k_Hyd) of major DNA methylation products N7-methyl-2′-deoxyguanosine (MdG) (Figure 1) and N3-methyl-2′-deoxyadenosine (MdA).^18,19 Compared to free DNA, MdG or MdA depurination decreases as much as ~6-fold within NCPs when at positions of the Widom 601 strong positioning DNA sequence that are proximal to N-terminal histone tails.²⁰ Furthermore, k_Hyd increases when positively charged tail residues are replaced by alanine or the tails are truncated.

Figure 1.

Formation and reactivity of N7-methyl-2′-deoxyguanosine (MdG). (A) Reaction scheme showing the formation of DNA–protein crosslinks (DPC) and abasic sites (AP) from MdG. (B) Reaction coordinate diagram describing MdG depurination in free DNA and an NCP.

Herein, we establish that the effects of histone tails on MdG depurination correlate with their effects on local ionic strength surrounding the DNA. Depurination of MdG or MdA produces an abasic site (AP) via a unimolecular nucleophilic substitution (S_N1) mechanism that produces an oxocarbenium ion (X) in the rate-determining step (Figure 1A). Since the alkylated purines are positively charged, hydrolysis results in charge dispersal en route to the transition state of the rate-determining step.²¹ Hence, the rate constant for depurination (k_Hyd) should decrease as the local ionic strength increases (Figure 1B).

EXPERIMENTAL PROCEDURES

Preparing 145-mer 601 DNA Containing Cy5 and 2-Aminopurine (2-AP) (Position 88, 247, or 72).

The 5′-Cy5 labeled 145-mer 601 DNA containing 2-AP was prepared by enzymatic ligation using chemically synthesized oligonucleotides (Figures S1C–E and S7). Oligonucleotides (15 nmol) other than the 5′-Cy5 labeled one were phosphorylated with adenosine triphosphate (ATP) (4 mM) and T4 polynucleotide kinase (PNK) (200 units) in 100 μL 1 × PNK buffer (70 mM Tris–HCl, pH 7.6, 10 mM MgCl₂, 5 mM dithiothreitol (DTT)) at 37 °C for 4 h. T4 PNK was inactivated by incubating at 65 °C for 30 min. The 5′-phosphorylated oligonucleotides were combined with the 5′-Cy5-oligonucleotide and annealed with the appropriate scaffolds (22.5 nmol). Ligation (700 μL) was performed with T4 DNA ligase (32 000 units) in 1 × T4 ligase buffer (50 mM Tris–HCl, pH 7.5, 10 mM MgCl₂, 1 mM ATP, 10 mM DTT) at 16 °C overnight. After phenol–chloroform extraction, the supernatant was mixed with 95% formamide and purified by 8% denaturing PAGE (20 × 16 × 0.1 cm). The gels were run at room temperature under limiting power (15 W) until the xylene cyanol migrated out of the gel. The product band was excised, crushed, and the DNA was eluted overnight at room temperature in 2 mL of elution buffer (0.2 M NaCl and 1 mM ethylenediamine tetraacetic acid (EDTA)). The gel particles in the solution were pelleted by centrifugation at 4 °C (10 min, 13 200 rpm). The supernatant was concentrated, and the buffer was exchanged with water (6 rounds) using a 10 K Amicon Ultra membrane at 4 °C. The 145-mer ligated DNA product (~4.5 nmol) was stored at −20 °C until further use.

Preparation and Purification of NCPs and Double-Strand DNA Containing Cy5 and 2-AP.

To prepare NCPs and double-stranded DNA containing Cy5 and 2-AP to examine the effect of NCP formation on Cy5 fluorescence, 5’-Cy5-labeled 145-mer DNA containing 2-AP (200 pmol) was hybridized with the complementary 145-mer DNA (200 pmol) in a solution (10 μL) containing 4-(2-hydroxyethyl)-1-piperazineethanesulphonic acid (HEPES) (10 mM, pH 7.5) and 100 mM NaCl via heating at 95 °C for 2 min, followed by slow cooling to room temperature. The solution was split in half, and 4 M NaCl (5 μL) was added. Histone octamer (98 μM, 1.42 μL) and H2A/H2B dimer (51 μM, 0.8 μL) were added to one sample, and 2.22 μL of octamer purification buffer (10 mM HEPES, 2 M NaCl, 1 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride (PMSF), pH 7.5) was added to the other. Both mixtures were incubated at 4 °C for 30 min before a series of dilutions at 4 °C using a buffer containing 10 mM HEPES, pH 7.5. Dilution #: volume of buffer added in μL, incubation time in min-1: 12, 60; 2: 6, 60; 3: 6, 60; 4: 10, 30; 5: 10, 30; 6: 20, 30; 7: 50, 30; 8: 100, 30. After the final dilution (total volume ~ 224 μL), any precipitate was pelleted via a brief spin (1 min) at 13 200 rpm at 4 °C. A small aliquot (2–4 μL) was mixed with 40% sucrose (3 μL) and loaded on a 7% native PAGE to check the reconstitution efficiency. The gel (10 × 8 × 0.15 cm) was run at 4 °C under the limiting power (3 W) until bromophenol blue migrated to the bottom of the gel.

$$2-\text{AP}{\left(\text{Adj}\right)}_{\text{NCP}}=2-\text{AP}{\left(\text{Obs}\right)}_{\text{NCP}}\left(\frac{{\text{Cy}5}_{\text{Free}}}{{\text{Cy}5}_{\text{NCP}}}\right)$$ (1)

$$2-\text{AP}{\left(\text{Corr}\right)}_{\text{NCP}}=2-\text{AP}{\left(\text{Adj}\right)}_{\text{NCP}}-\left({\text{Frac}}_{\text{Free}}\times 2-{\text{AP}}_{\text{Free}}\right)$$ (2)

$$2-\text{AP}{\left(\text{Fin}\right)}_{\text{NCP}}\text{=}\frac{2-\text{AP}{\left(\text{Corr}\right)}_{\text{NCP}}}{{\text{Frac}}_{\text{NCP}}}$$ (3)

To prepare and purify NCPs and double-stranded DNA containing Cy5 and 2-AP for 2-AP fluorescence measurement, the 5′-Cy5-labeled 145-mer DNA containing 2-AP (300 pmol) was hybridized with the complementary 145-mer DNA (300 pmol) in 150 μL of HEPES (10 mM, pH 7.5) and 100 mM NaCl via heating at 95 °C for 2 min, followed by slow cooling to room temperature. The hybridized DNA was mixed with ~1.3 equiv of histone octamer and additional H2A/H2B dimer (~120 pmol) in 3 mL of HEPES (10 mM, pH 7.5) and 2 M NaCl. The sample was first dialyzed using Slide-A-Lyzer Dialysis cassette G2 (3500 Da MWCO) in a 1 L beaker containing chilled high salt buffer (600 mL, 2 M NaCl, 10 mM HEPES, pH 7.5) for 2 h at 4 °C with stirring. Subsequently, salt-free buffer (10 mM HEPES, pH 7.5) was added using a peristaltic pump (1.5 mL min⁻¹). A second pump removed the solution from the beaker at the same rate to maintain a constant volume. After carrying out the buffer exchange for 36 h, the sample was concentrated to ~60 μL by centrifugation (4000 rcf, 4 °C) using a 10 K Amicon Ultra membrane. A Mini Prep Cell apparatus (BioRad) was used for purification of NCPs and double-strand DNA.²² NCP or DNA samples (300 pmol, ~60 μL) were mixed with 20 μL of 40% (w/v) sucrose before loading on a 6.5 cm 7% native PAGE (59:1 = acrylamide/bisacrylamide) column. The gel column was run at 4 °C under limiting power (1 W) with a peristaltic pump (150 μL min⁻¹) to deliver elution buffer (10 mM Tris–HCl, 1 mM DTT, pH 7.5). A total of 80 fractions (~300 μL fraction⁻¹) were collected after NCPs or DNA (blue band) reached the 2 cm mark. To distinguish which fraction contains nucleosomes or free DNA, one out of every five fractions for the first and last 20 fractions and every other fraction for the 40 fractions in between these were checked on a 7% native PAGE (10 × 8 × 0.15 cm). The gel was run under limiting power (3 W) at 4 °C until xylene cyanol migrated to the bottom of the gel (for NCP samples) or until bromophenol blue migrated to the bottom of the gel (for free DNA samples). The fractions (5–8) containing pure nucleosomes or free DNA were pooled immediately and concentrated to ~80 μL using a 10 K Amicon Ultra membrane (4000 rcf, 4 °C). Lastly, a solution of 1 M NaCl was added to yield a final concentration of 100 mM NaCl prior to fluorescence measurement.

Fluorescence Measurement of DNA and NCPs Containing Cy5 and 2-AP.

The fluorescence spectra of free DNA and NCPs were collected using a 1 cm path length quartz cell (PerkinElmer) in a fluorimeter (Horiba Fluorolog-3). Excitation and emission slits were set at 5 nm. To examine the effect of NCP formation on Cy5 fluorescence, the emission spectra of Cy5 of DNA and NCP samples were recorded from 655 to 800 nm, following excitation at 647 nm. To measure 2-AP fluorescence in purified DNA and NCP samples, the emission spectra of 2-AP were recorded from 330 to 500 nm, following excitation at 310 nm. Measurements were carried out on two independently prepared samples of each DNA or NCP construct. The Cy5 fluorescence at an emission maximum of 663 nm was used to normalize the 2-AP fluorescence between NCP and free DNA samples that have different concentrations. Specifically, the observed emission spectrum of 2-AP (2-AP(Obs)_NCP) in an NCP sample was first normalized (2-AP(Adj)_NCP) by comparing its Cy5 fluorescence at 663 nm (Cy5_NCP) with that of the respective free DNA sample (Cy5_Free) (eq 1). The final 2-AP fluorescence in an NCP sample was then determined by subtracting the contribution from free DNA to the signal. The fraction of free DNA present in an NCP sample (Frac_Free) was measured using native PAGE (7%) after recording the fluorescence measurement (Figure S9). The 2-AP fluorescence attributable to the NCP in the sample (2-AP(Corr)_NCP) was determined by subtracting the product of Frac_Free and 2-AP_Free from the adjusted fluorescence (2-AP(Adj)_NCP, eq 2). Lastly, the 2-AP fluorescence contributed solely by the NCP in a sample (2-AP(Corr)_NCP) was plotted (2-AP(Fin)_NCP) after normalizing for the fraction of NCP present in an NCP sample (Frac_NCP) determined by native PAGE (eq 3). 2-AP(Fin)_NCP spectra were smoothed via adjacent averaging of 50 points using Origin 6.1.

RESULTS AND DISCUSSION

We probed the local ionic strength surrounding DNA for the first time in an NCP by utilizing a method developed by Bevilacqua, which capitalizes on the dependence of 2-aminopurine (2-AP) fluorescence on the corresponding stability (pK_a) of an adjacent protonated 2′-deoxyadenosine·2′-deoxycytidine ([AH⁺·C]) wobble base pair (Scheme 1).^23–25 The pK_a of [AH⁺·C] decreases with increasing ionic strength. Deprotonation destabilizes the wobble base pair and 2-AP stacking within the duplex, resulting in increased fluorescence of the latter. Hence, 2-AP fluorescence is an indirect measure of the local ionic strength. This method has been used as an “electrostatic meter” in free DNA by Herschlag, who proposed that it would be useful in studies on NCPs.^12,25 Although three positions within the NCP were examined, histone tail effects were only compared to one another at a given position. This reduces concerns that changes in fluorescence could be due to DNA sequence and protein structural effects.

Scheme 1.

Use of an [AH⁺·C] Base Pair and 2-Aminopurine (2-AP) Fluorescence as an Electrostatic Meter^23–25

In this investigation, the DNA sequences used were based upon the Widom 601 strong positioning sequence (Figure 2).²⁰ They were constructed containing either MdG or a dinucleotide sequence comprised of 5′-(2-AP)(A)/(C)(T) substituted for the appropriate nucleotide(s) at specific sites within the nucleosomal DNA to most closely compare the effect of ionic strength on damaged DNA reactivity (Figure 2B). NCPs containing MdG were prepared and characterized as previously described (Figures S3–S5).¹⁸ The 2′-deoxyadenosine of the dinucleotide electrostatic meter sequence that is paired with 2′-deoxycytidine was incorporated at the corresponding MdG position because this is the site at which we are interested in determining the effects of the histone tails on the local ionic environment. Cy5 was installed at the 5′-termini of strands containing 2-AP to facilitate the normalization of 2-AP fluorescence between DNA and NCP samples of different concentrations. Cy5 fluorescence at 663 nm was minimally suppressed (<5%) upon NCP formation (Figure S8). Hybridized duplex DNA and NCPs containing 5′-(2-AP)(A)/(C)(T) were purified using a Mini Prep Cell, but gel electrophoresis revealed small amounts (≤5%) of free DNA in the isolated NCPs (Figure S9). Consequently, the contribution of 2-AP fluorescence due to free DNA in NCP samples was subtracted. At each position, the trend in corrected 2-AP fluorescence was compared with MdG depurination rate constants to probe whether the chemical reactivity correlates with ionic strength. Reported fluorescence data (Figure 3A–C) are the average of two independently prepared NCP samples (Figure S10).

Figure 2.

NCP structure. (A) Positions at which (2-AP)(A)/(C)(T) or MdG was site-specifically introduced. Each perspective is a single gyre of DNA (PDB: 1kx5). (B) Local DNA sequences of (2-AP)(A)/(C)(T) and MdG at each nucleosomal position. (C) N-terminal tail amino acid sequences of Xenopus laevis histone H4, H2B, and H3. (D) Positively charged histone H4 amino acids in proximity to N⁸⁹. (E) Positively charged histone H2B amino acids in proximity to N²⁴⁸. Please note that DNA positions are superscripted.

Figure 3.

Measurement of the local ionic environment of an NCP compared to free DNA using 2-AP fluorescence as a function of [AH⁺·C] base pairing. (A) 2-AP⁷², [AH⁺·C]⁷³; (B) 2-AP⁸⁸, [AH⁺·C]⁸⁹; (C) 2-AP²⁴⁷, [AH⁺·C]²⁴⁸; and (D) relative 2-AP fluorescence vs relative rate constant for MdG hydrolysis at positions 89 and 248. The data presented in panels (A)–(C) were used to plot the 2-AP fluorescence of a particular NCP (FI_NCP(x), X = wild type histone, etc.) relative to that of the corresponding free DNA (Fl_{Free DNA}). The relative rate constants were determined using the average values reported in Tables 1 and 2. NCPs: (position 89, 1–4) 1, WT; 2, H4-K16,20A; 3, H4-K16,20A, R17,19A; 4, H4-Del 1–20; (position 248, 5–7) 5, WT; 6, H2B-R26,27,30A; 7, H2B-K24,25,28,31A. Please note that DNA positions are superscripted. Data plotted in panels (A)–(C) are the average of two individual experiments (Figure S10).

To establish the general effect of NCP formation on the local ionic environment and MdG depurination, position 73 (Figure 2A,B) at the dyad axis was selected as a “benchmark” due to the absence of extensive DNA–histone N-terminal tail interactions (Figure 3A).⁷ The 2-AP⁷² (please note that the position within the DNA is superscripted) fluorescence within NCP samples comprised of wild-type histones were indistinguishable from those of free DNA. This indicates that wrapping DNA around an octamer did not cause a significant change in the pK_a of [AH⁺·C] or importantly the ionic strength at the dyad axis compared to in free DNA. Deleting the histone H4 N-terminal tail (H4-Del 1–20) resulted in a less than 10% decrease in 2-AP⁷² compared to that of wild-type NCP or free DNA. In contrast, NCP samples comprised of histone H3-Del 1–37 showed ~15% reduction in 2-AP⁷² fluorescence, indicating a small reduction in local ionic strength that increased the pK_a of [AH⁺·C].

The modest but discernible effect of deleting the histone H3 N-terminal tail is consistent with molecular dynamics simulations, which indicates that this tail makes occasional contact with DNA in the dyad region.⁷ It is notable that 4 of the 10 N-terminal amino acids of histone H3, which could reach furthest from the point at which the tail protrudes from the core, are positively charged (Figure 2C). Interaction with these amino acids potentially accounts for the slight increase in ionic strength at position 73 when the H3 histone tail is present. Qualitatively, the reported rate constants for MdG⁷³ hydrolysis in free DNA and the NCP respond similarly to the fluorescence experiment. The rate constant for depurination was only weakly affected upon incorporation within an NCP (t_½ NCP/free DNA = 1.3) under similar reaction conditions (10 mM HEPES, pH 7.5, 100 mM NaCl) as the fluorescence experiments.¹⁸ Tris·HCl was substituted for HEPES in the fluorescence experiments because the latter interferes with 2-AP fluorescence.

Having established position 73 as a benchmark, two additional sites at which other studies suggested more extensive tail interactions were examined.^7,26,27 Position 89, within superhelical location 1.5, was chosen because the DNA structure at this region is altered from free B-form, and the site is a hot spot for DNA-binding molecules.^26,28 In addition, a variety of chemical studies indicate that the histone H4 N-terminal tail interacts with DNA at this position (Figure 2D).^{14,16,17,29,30} When the wobble base pair is incorporated at position 89 ([AH⁺·C]⁸⁹), 2-AP⁸⁸ fluorescence in the NCP is 40–70% greater than in the corresponding free DNA (Figure 3B), suggesting a significant effective increase in local ionic strength. In addition, 2-AP⁸⁸ fluorescence progressively decreases in NCPs comprised of histones H4-K16,20A, H4-K16,20A,R17,19A, and H4-Del 1–20. The fluorescence data indicate that deleting the positively charged amino acids that are proximal to [AH⁺·C]⁸⁹ decreases the local ionic strength, which increases the stability of the wobble base pair and decreases 2-AP fluorescence. The 2-AP⁸⁸ fluorescence is greater than that of the corresponding free DNA even when the entire N-terminal histone H4 tail is removed. This could be due to interactions with other histone tail(s), such as histone H3⁷, or reflective of other structural differences that affect 2-AP fluorescence.

The trend in local ionic strength changes correlates qualitatively with MdG⁸⁹ hydrolysis rate constants in free DNA and NCPs containing the same histone H4 mutants (Table 1). While MdG⁸⁹ hydrolysis was reported to be 2.9-times slower within an NCP than in free DNA,¹⁸ replacing two positively charged amino acids with uncharged ones relatively close to the DNA (H4-K16,20A, Figure 3B) increases k_Hyd. Greater positive charge depletion in the histone H4 N-terminal tail (H4-K16,20A,R17,19A) further accelerates MdG⁸⁹ depurination, giving rise to k_Hyd (t_½ NCP/free DNA = 1.7) that was within experimental error of that in which the entire histone H4 tail was removed (H4-Del 1–20).

Table 1.

MdG⁸⁹ Hydrolysis in Free DNA and NCPs

substrate	kHyd (10−7 s−1)a	t1/2 (h)	rel. t1/2b
free DNAc	10.1 ± 0.5	191 ± 8
WT NCPc	3.5 ± 0.2	553 ± 27	2.9
H4-K16,20A	4.6 ± 0.2	417 ± 18	2.2
H4-K16,20A, R17,19A	5.8 ± 0.1	330 ± 7	1.7
H4 Del 1–20	5.9 ± 0.5	330 ± 30	1.7

^aValues are the ave. ± std. dev. of three independent experiments.

^bRel. t_1/2 = t_1/2 (NCP)/t_1/2 (Free DNA).

^cRef 18.

Position 248 is within the other DNA gyre that wraps around the NCP and is expected to interact with the lysine and arginine-rich histone H2B repression domain (HBR) of the H2B N-terminal tail (Figure 2E).^7,26,27 The highly conserved HBR sequence influences gene expression, chromatin assembly, and DNA damage repair.^27,31 2-AP²⁴⁷ fluorescence in the [AH⁺·C]²⁴⁸ NCP was almost twice as high as in the corresponding free DNA, suggesting elevated ionic strength within NCPs at this position (Figure 3C). Furthermore, while removing positive charges (H2B-R26,27,30A or H2B-K24,25,28,31A) resulted in 2-AP²⁴⁷ fluorescence signals that were indistinguishable from one another, the emission in these NCPs was clearly intermediate to that in the NCP comprised of wild-type histones and the corresponding free DNA.

MdG²⁴⁸ depurination was ~9-fold faster than k_Hyd for MdG⁸⁹ in free DNA (Tables 1 and 2). The MdG²⁴⁸ and MdG⁸⁹ flanking sequences are different from one another, a property that is known to affect depurination rates of N7-alkylated nucleotides.³² Despite the large difference in k_Hyd between these two positions in free DNA, incorporation within an NCP results in a comparable relative magnitude of decrease in hydrolysis (2.7-fold slower than in free DNA). In addition, just as mutation of a cluster of arginines or lysines to uncharged alanines in the HBR region resulted in an intermediate decrease in the local ionic strength, k_Hyd increased to values part way between free DNA and NCP reconstituted from wild-type histones.

Table 2.

MdG²⁴⁸ Hydrolysis in Free DNA and NCPs

substrate	kHyd (10−6 s−1)a	t1/2 (h)	rel. t1/2b
free DNA	8.4 ± 0.2	23 ± 1
WT NCP	3.1 ± 0.1	62 ± 2	2.7
H2B-R26,27,30A	5.1 ± 0.1	38 ± 1	1.7
H2B-K24,25,28,31A	6.5 ± 0.1	30 ± 1	1.3

^aValues are the ave. ± std. dev. of three independent experiments.

^bRel. t_1/2 = t_1/2 (NCP)/t_1/2 (Free DNA).

Overall, these experiments are consistent with enhanced local ionic strength within NCPs at DNA positions (89 and 248) that are close to positively charged N-terminal histone tails. Given that MdG depurination proceeds via a mechanism that is sensitive to changes in the local ionic environment (Figure 1), it is reasonable for the [AH⁺·C] stability and corresponding 2-AP fluorescence to correlate with k_Hyd. This was explored by carrying out kinetic experiments under the same conditions with otherwise identical NCPs as those used in fluorescence studies. Indeed, the previously reported small effect (~30% decrease) of NCP formation on MdG⁷³ hydrolysis correlates well with the comparable 2-AP⁷² fluorescence in the NCP and free DNA.¹⁸ Hydrolysis rate constants at MdG⁸⁹ and MdG²⁴⁸ respond to local changes in the ionic environment as a consequence of charge on the proximal N-terminal histone tails. A plot of the relative 2-AP fluorescence (Fl_NCP(X)/Fl_{Free DNA}) versus relative k_Hyd (k_NCP(X)/k_{Free DNA}) illustrates at a single position within the NCP that decreasing fluorescence, interpreted as a decrease in local ionic strength, correlates with increased k_Hyd (Figure 3D). When interpreting these experiments, one must be aware of the caveat that the NCP structure is heterogeneous and other factors may contribute to both measurements. However, when comparing measurements at the same nucleotide position in which the constructs differ by the structure of the tails, it is reasonable to assume that sequence and other structural factors cancel out. The fluorescence measurements at the three NCP positions are fully consistent with this premise. MdG deglycosylation in NCPs is modulated by the local ionic environment, which in turn is affected by the presence and structure of proximal N-terminal histone tails.^18,19 The correlation between the two types of experiments also suggests that either can be employed to probe the local ionic environment. The smaller quantity of material required for the experiments in which MdG hydrolysis is measured could be advantageous in the study of larger systems, such as oligonucleosomes that are more challenging to prepare.^33,34

The effect of local ionic strength on the reactivity of positively charged nucleotides (e.g., N3- and N7-alkylated purines) in nucleosomal DNA could affect product formation and should be general. For instance, decreased hydrolysis of MdG (Figure 1A) and N3-methyl-2′-deoxyadenosine (MdA) in NCPs could give rise to greater amounts of cross-links with histones (DPCs).^18,19 DNA lesions formed by reaction with other mono- and bifunctional alkylating agents are structurally analogous to MdG and MdA. Chemical lifetime variation of positively charged alkylated lesions due to positioning within an NCP could contribute to a lesion’s promutagenic effect on replication.^35–37 In addition, the increased lifetime of the initially formed adduct of a bis-alkylating agent enables the second step of the processes that consummate DNA–DNA and DNA–protein cross-linking to compete more effectively.^38–41 On the other hand, decreased local ionic strength will result in greater AP formation and the corresponding DNA–DNA and DNA–protein cross-link products.^42–46 As MdG is less generally more weakly mutagenic than AP, greater MdG depurination can also give rise to increased mutation rates.^35,36,47,48 The mutagenicity and cytotoxicity of positively charged alkylation lesions in NCPs may be further complicated by histone tails as they have been proposed to sterically block the access of repair enzymes to nucleosomal DNA.^37,49–52 Positively charged amino acids on histone tails play a crucial role in modulating repair through post-translational modifications,^53,54 as well as transiently formed chemical intermediates.¹³ However, the correlation of damaged DNA accessibility and excision by base excision repair enzymes in NCPs is imperfect, perhaps due to the contribution of other effects, including the local ionic environment.⁵⁵

Nucleosome core particles are large macromolecule ensembles, and it is possible that multiple factors can influence physical and chemical properties. Our experiments unequivocally demonstrate that N-terminal histone tails modulate the local ionic environment within an NCP. These effects are corroborated by the hydrolysis of the DNA lesion MdG, which responds to changes in ionic strength in a manner that is fully consistent with its reaction mechanism (Figure 1). Local ionic environment adds another layer of complexity to the reactivity, mutagenicity, and repair of DNA lesions in NCPs and chromatin. Additionally, histone tails and nucleosome-binding proteins target overlapping regions on nucleosomal DNA.^7,8 Histone tail-mediated effects, including those resulting from PTMs that alter charge (e.g., lysine acetylation, serine/threonine phosphorylation), on the local ionic environment may affect protein recognition and association.