DNA coronas resist nuclease degradation

Abstract

The interaction of cell-free DNA with biological particles has been linked to autoimmune diseases such as systemic lupus erythematosus, but mechanistic details are lacking. Our recent work has shown that DNA adsorbed on the surface of synthetic particles, forming a DNA “corona,” leads to an enhanced immunostimulatory response in macrophages, providing a model system to understand how DNA-particle interactions may lead to autoimmune diseases. This current study provides a detailed examination of DNA (500–600 base pairs and ∼10,000 base pairs) interacting with synthetic particles (40 nm to 10 μm) and planar surfaces. Of specific interest is how DNA adsorbed on the surface of particles is resistant to degradation by DNase 1, a common nuclease. DNA-particle complexes are characterized by a colorimetric DNA concentration assay (PicoGreen), spectroscopy (NanoDrop), dynamic light scattering (DLS), confocal fluorescence microscopy, and transmission electron microscopy. These studies show that the protective effect of the particle is size dependent, with smaller (40 and 200 nm) particles providing less protection. Correlated with this lack of protection is significantly increased particle aggregation, suggesting that a DNA corona formed on the larger particles is protective, whereas particle aggregation, which dominates the smaller particles, is not protective. The formation of a single-stranded DNA corona leads to the opposite protective effect, with smaller (200 nm) particles leading to near-complete protection of DNA from nuclease degradation. Overall, this study provides an important biophysical basis for the interaction of DNA with particles with the goal of guiding future in vitro and in vivo studies of cell-free DNA and particles in autoimmune disease.

Significance

Cell-free DNA is an important marker for systemic lupus erythematosus. DNA outside of the nucleus or mitochondria is normally rapidly degraded by nucleases as DNA outside of these organelles can stimulate the innate immune system. We find that DNA adsorbed on the surface of particles forming a DNA “corona” is resistant to nuclease activity. Particles of all diameters (40 nm to 10 μm) provide some protective effect for all types of DNA (500–600 bp, ∼10,000 bp, double-stranded DNA, single-stranded DNA), but the level of protection depends on a complex combination of DNA length and persistence length, particle diameter, and especially DNA-induced particle aggregation. This biophysical study of DNA-particle interactions provides insight into DNA-particle interactions relevant to autoimmune disease.

Introduction

DNA is a biological polymer normally localized in the nucleus and mitochondria of cells (1). Cell-free DNA is generated when cells undergo cell death, such as apoptosis and necrosis, or in response to foreign pathogens during an inflammatory response (2,3,4). The length of cell-free DNA depends on its method of generation. When cells undergo apoptosis, short fragments of DNA between 100 and 250 base pairs (bp) are generated. These fragments correspond to the length of nucleosomal DNA (2,3,4). Because necrosis is a faster process than apoptosis, cell-free DNA generated by necrosis is not degraded by nucleases before release. This results in longer DNA fragments >1 kbp (2,3,4). Cell-free DNA is an important biomarker for cancers and autoimmune diseases such as systemic lupus erythematosus (SLE or lupus) (2,5,6).

It has previously been reported that >90% of cell-free DNA may be associated with extracellular vesicles (7). Extracellular vesicles are biological particles generated by cells that range in size from 40 nm to 5 μm (8). Extracellular vesicles have been strongly implicated in autoimmune diseases such as SLE (9,10,11). For example, studies have demonstrated that, in patients with SLE, microparticles, a class of extracellular vesicles, have anti-DNA antibodies bound to their surface (11). This observation suggests that cell-free DNA can be both inside and on the surface of extracellular vesicles. Our own previous work showed that DNA electrostatically adsorbs on the surface of synthetic particles, forming a DNA “corona” (12), which leads to an enhanced immunostimulatory response from macrophages (13). Outside of the nucleus or mitochondria, DNA is subject to rapid degradation by nucleases including DNase 1, DNase II, and DNase 1L3 (14,15). This is a protective mechanism of the cell as DNA outside of these organelles, including cell-free DNA, can stimulate the innate immune system through pattern recognition receptors such as TLR9 and cGAS-STING (16). Our work showed that DNA on the surface of particles is protected from degradation by DNase 1, an effect that was correlated with an enhanced immunostimulatory response from macrophages (13).

Our goal in this current study is to better understand how DNA on the surface of particles is protected from enzymatic degradation. By using a well-controlled model system of polystyrene particles (40 nm to 10 μm) and planar surfaces, we can more directly probe the biophysics of the system. In comparison, biological particles obtained from blood or tissue culture fluid are highly heterogeneous in terms of size and composition of DNA, RNA, and proteins, making mechanistic studies difficult (6). The particle diameters used in this study include the range of biological particle diameters (40 nm to 5 μm) (8,10,17) and also allow us to examine variation in radius of curvature. We use two different lengths of DNA obtained from Escherichia coli (sheared 500–600 bp and unsheared ∼10,000 bp). The lengths of DNA were chosen to mimic the small fragments of cell-free DNA that serve as autoimmune disease biomarkers (∼100–200 bp) and the larger DNA fragments that are released during necrosis (>1000 bp) (3,6,10). The use of double-stranded (ds) DNA and single-stranded (ss) DNA provides a comparison of nuclease protection as a function of DNA persistence lengths (50 and 0.7–6 nm, respectively) (18,19,20).

We first measured the amount of dsDNA adsorbed on the surface of particles (40 nm to 10 μm) as a function of the amount of DNA used to form the corona. DNA-particle complexes were characterized by a colorimetric DNA concentration assay (PicoGreen), spectroscopy (NanoDrop), dynamic light scattering (DLS), confocal fluorescence microscopy, and transmission electron microscopy (TEM). Of specific interest is how the concentration of the DNA corona changes for complexes formed from stoichiometric concentrations of DNA (∼1 molecule of DNA per particle) as compared to excess (∼27 molecules of DNA per particle) concentrations of DNA. We then measure the amount of DNA remaining in the corona after treatment with DNase 1 for both particles and a planar surface. In addition to these studies with dsDNA, we also examined ssDNA using S1 nuclease instead of DNase 1 due to its specificity for ssDNA (21). Overall, this biophysical study of DNA-particle interactions provides insight relevant to DNA-particle interactions relevant to autoimmunity.

Materials and methods

Particle characterization

Cationic, amine-modified, polystyrene particles (40 nm, 200 nm, 1 μm, 2 μm, and 10 μm; #37352, #37356, #37362, #37366, and #37359, respectively; Thermo Fisher, Waltham, MA) were used for all experiments. Before use, particles were vortexed briefly and sonicated with a cup-horn sonicator (50 s on/10 s off; 30% amplitude; Q500 Sonicator, Q Sonica Sonicators, Newtown, CT). The hydrodynamic diameter (d_h by intensity) and polydispersity index (PdI) were measured using DLS (Malvern Zetasizer, Nano-Z, Malvern Instruments, Worcestershire, UK). Measurements were carried out in triplicate (25 pM in water). Each measurement consisted of 30 runs. TEM (Tecnai G2 TWIN, FEI, Hillsboro, OR) was also used to measure particle diameter. Particles (500 pM in water) were drop-cast onto a carbon-coated grid (#FCF200-Cu, Electron Microscopy Sciences, Hatfield, PA) and blotted with filter paper after 5 min. Particles were imaged at 160 kV.

DNA corona formation and characterization

Our DNA corona formation method has been described previously (12,13). In brief, DNA coronas were formed by incubating particles (500 pM) with E. coli DNA (0.25 mg/mL, #J14380, Fisher Scientific, Hampton, NH), in Dulbecco’s PBS (pH 7.3–7.5, #28374, Thermo Fisher). The length of DNA used to form a corona was either 500–600 or ∼10,000 bp, as confirmed in our previous publication and noted in the text (13). The DNA and particles were mixed by pipetting up and down and then incubated at room temperature for 45 min. 1 μL of hydrochloric acid (1 M) was added to the 1 and 2 μm particles when incubating with DNA to protonate the amine functional groups. Unbound DNA was removed by centrifugation (18,000 rcf) and resuspension in PBS (×3). Removal of unbound DNA using this washing process was confirmed previously (12). The resulting particles with a DNA corona were characterized with DLS in the same way as the bare particles, as described above. The concentration of DNA in the corona was measured using the PicoGreen DNA quantification assay (#P7589, Thermo Fisher) according to the manufacturer’s instructions. Fluorescence intensity (Ex, 480 nm; Em, 520 nm) was measured with a plate reader (SpectraMax iD3, Molecular Devices, San Jose, CA). The background signal from the bare particles was subtracted from the overall signal.

The concentration of DNA used to form the corona (stoichiometric to excess) was determined using a 500 pM particle concentration (assuming 6.02 × 10²³ polystyrene particles per liter is 1 M) of 200 nm particles and converting mass of DNA to moles assuming a single DNA molecule is 10,000 bp with a molecular weight of 651.98 g/mol for each base pair (22). The stoichiometric amount of DNA is 1 mole of DNA per mole of 200 nm particles (3 μg of DNA per 1 mL volume) for 200 nm particles. For the other particles (1 μm to 10 μm), a stoichiometric concentration is matched by surface area of the particles. Specifically, the number of particles that have the same surface area as 500 pM of 200 nm particles were incubated with 3 μg of DNA, resulting in molar ratios of 25 moles of DNA/mole of 1 μm particles; 100 moles of DNA/mole of 2 μm particles; and 2500 moles of DNA/mole of 10 μm particles. We defined excess DNA as 27 moles of DNA/mole of 200 nm particles (81 μg per 1 mL volume) and 675 moles of DNA/mole of 1 μm particles; 2700 moles of DNA/mole of 2 μm particles; and 67,500 moles of DNA/mole of 10 μm particles. This concentration matches that used to form the DNA corona in our previous studies (13). For a planar surface, stoichiometric and excess concentrations of DNA were not calculated since there is no particle number to normalize against (3 and 81 μg of DNA were used per well in a 96-well plate).

Sheared DNA (DNA_S) was prepared by sonicating DNA with a cup-horn sonicator (50 s on/10 s off, six cycles, 30% amplitude; Q500 Sonicator, Q Sonica Sonicators, Newtown, CT). The length of DNA_S is 500–600 bp, confirmed using agarose gel electrophoresis (1% gel, 80 V, 45 min) in our previous publication (13).

For experiments using sheared DNA (DNA_S; 500–600 bp), we used the same concentrations of DNA as for experiments with the longer (∼10,000 bp) DNA and describe these coronas as low (3 μg of DNA_S) or high (27 μg of DNA_S) DNA concentrations corresponding to ∼20 moles of DNA_S/mole of 200 nm particles and ∼540 moles of DNA/mole of 200 nm particle, respectively. This results in molar ratios of 0.4 moles of DNA_S/mole 40 nm particles; 500 moles of DNA_S/mole of 1 μm particles; and 2000 moles of DNA_S/mole 2 μm particles when using a low amount of DNA. When using a high amount of DNA, these molar ratios become 10.8 moles of DNA_S/mole of 40 nm particle; 13,500 moles of DNA_S/mole of 1 μm particle; and 54,000 moles of DNA_S/mole 2 μm particle. A true stoichiometric corona formed from DNA_S (150 ng of DNA_S) results in a corona concentration below the detection limit of PicoGreen (100 ng). The absolute mass of DNA_S used was kept consistent with the mass of longer DNA to keep surface area coverage of particles consistent.

For a planar surface, stoichiometric and excess concentrations of DNA were not calculated since there is no particle number to normalize against (3- and 81-μg amounts of DNA were used per well in a 96-well plate).

DNA degradation by DNase 1

DNA-particle complexes were prepared in PBS containing magnesium and calcium (#14040133, Thermo Fisher), which are necessary for DNase 1 activity. DNase 1 (#4716728001, Millipore Sigma, Burlington, MA) was added to solutions at a concentration of 1U/μg DNA and the mixture was incubated at 37°C for 30 min. DNase 1 was heat inactivated by incubating the sample at 75°C for 10 min before DNA concentration was quantified using the PicoGreen Assay, as described above. In comparison, ssDNA is generated at 95°C. For DNA on planar surfaces, DNA was incubated with black poly-L-lysine coated plates (#356515, Corning Life Sciences, Corning, NY) for 30 min in PBS containing calcium and magnesium. DNA was then treated with DNase 1 as described above.

Preparation, quantification, and degradation of ssDNA

ssDNA was prepared by heating dsDNA (unsheared) to 95°C for 10 min in Eppendorf tubes vented with a surgical needle tip at the top of the tube. DNA was rapidly cooled in ice and stored in −20°C until ready to use. The concentration of ssDNA was quantified using the Quant-iT OliGreen ssDNA Assay Kit (#11492, Thermo Fisher) according to manufacturer’s instructions and measured with a plate reader as described above for dsDNA. ssDNA particles for degradation assays were prepared in S1 nuclease reaction buffer (#EN0321, Thermo Fisher). S1 nuclease (#EN0321, Thermo Fisher) was added to solutions of ssDNA and ssDNA particles at a concentration of 5 U/μg DNA and the mixture was incubated at 37°C for 10 min. S1 nuclease was heat inactivated by heating to 75°C for 10 min before ssDNA concentration was quantified with the OliGreen assay.

Statistical analysis

Data are presented as mean ± standard deviation (SD) or mean ± standard error of the mean (mean ± SE) as noted in the text. The statistical significance of the data was measured using multiple t-tests, one-way ANOVA, or two-way ANOVA with post hoc tests detailed in each figure caption. All data were analyzed using commercially available software (GraphPad Prism 10, San Diego, CA).

Results and discussion

Particle characterization

Bare particles were characterized using TEM (Fig. 1) and DLS (Table 1). Most of the particles (200 nm to 2 μm) have hydrodynamic diameters near the expected values and relatively low polydispersity values. The 40 nm particles remain aggregated (d_h = 238.2 ± 0.1 nm) even after vortexing and sonication. The 10 μm particles are most likely dimers (27,498.3 ± 16,597.5 nm).

Figure 1.

TEM images of bare particles, in the absence of DNA. The image of the 2 μm particles was taken from our previous study (13).

Table 1.

DLS Including Hydrodynamic Diameter and Polydispersity Index of Particles in the Absence of DNA

Particle	dh (nm)	PdI
40 nm	238.2 ± 0.1	0.17 ± 0.003
200 nm	198.8 ± 3.4	0.004 ± 0.05
1 μm	1716.3 ± 179.6	0.18 ± 0.1
2 μm	1982.7 ± 93.8	0.15 ± 0.1
10 μm	27,498.3 ± 16,597.5	0.61 ± 0.4

SD was calculated from n = 3 distinct samples. dh, hydrodynamic diameter; PdI, polydispersity index.

DNA adsorbs on the surface of particles as a function of initial DNA concentration

We first quantified the amount of DNA adsorbed on the surface of the particles as a function of particle diameter (200 nm and 2 μm) and DNA incubation concentration (1–81 μg) (Fig. 2 A). For both particle diameters, we found that, when incubated with ≥27 μg of DNA, there is no significant increase in the amount of DNA adsorbed on the particles. For subsequent experiments, we consider a stoichiometric amount of DNA used to form the corona (3 μg, 1 mole of DNA/mole of 200 nm particles) compared to an excess (81 μg = 27 moles of DNA/mole of 200 nm particles) as described in section materials and methods.

Figure 2.

DNA adsorbs on the surface of particles. (A) Concentration (log) of DNA adsorbed on 200 nm and 2 μm particles as a function of concentration of DNA used to form the DNA-particle complexes. Particles are normalized to 500 pM to allow for comparison between different diameters, as described in materials and methods. Significance was determined using a two-way ANOVA with Tukey’s multiple comparisons post hoc. ^∗∗∗∗p < 0.0001; nonsignificant comparisons are not included. n = 3 distinct samples. Error bars reflect ±1 SE. (B) TEM images of 200 nm particles with a DNA corona formed using stoichiometric (3 μg) and excess (81 μg) amounts of DNA. The excess amount of DNA was chosen to match the concentration used in our previous study (13). (C) TEM images of 2 μm particles with a DNA corona formed using stoichiometric (3 μg) and excess (81 μg) amounts of DNA.

DNA leads to aggregation of 200 nm particles

A notable feature of the 200 nm particles is the aggregation that results from the presence of DNA (Fig. 2 B; Table 2). Stoichiometric concentrations of DNA lead to an increased diameter (Δd_h = 2.26 μm) compared to bare particles, visible as aggregates in TEM images, and massive aggregation in response to excess DNA (Δd_h = 45 μm). In comparison, the 2 μm particles show a slight decrease (Δd_h = −244 nm) in diameter with stoichiometric DNA and a slight increase (Δd_h = 1 μm) with excess DNA, suggesting the DNA corona helps to stabilize, rather than aggregate, these larger particles. We estimate that the DNA is longer than particle diameters for the 200 nm particles (∼10,000 bp, 0.3 nm/bp, length of ∼3000 nm) (23), leading to aggregation. For the larger 2 μm particles, with particle diameters on the same order of magnitude as the length of DNA, this effect is not observed.

Table 2.

DLS of 200 nm and 2 μm Particles with DNA Coronas

Particle	DNA Amount	dh (nm)	Δ dh (nm)	PdI	Δ PdI
200 nm	stoichiometric	2800.3 ± 389.9	2620	0.39 ± 0.05	0.39
200 nm	excess	45,620 ± 2714.2	45,422	0.58 ± 0.4	0.58
2 μm	stoichiometric	1738.0 ± 114.5	−244	0.28 ± 0.2	0.13
2 μm	excess	3020.3 ± 308	1038	0.20 ± 0.1	0.05

The DNA corona was formed with stoichiometric (3 μg) or excess (81 μg) amounts of DNA. Change (Δ) relative to bare particles is included for d_h and PdI.

DNA adsorbs onto the surface of particles as a function of particle diameter

After establishing general trends with 200 nm and 2 μm particles (Fig. 2; Table 2), we examined four different particle diameters (200 nm, 1 μm, 2 μm, and 10 μm) comparing stoichiometric and excess concentrations of DNA used to form the corona (Fig. 3). The 40 nm particles were excluded as the massive amount of aggregation that resulted from incubation with DNA made the samples impossible to pipette. We found significantly more DNA adsorption on particles when incubated with excess DNA than with stoichiometric amount of DNA (Fig. 3 A).

Figure 3.

Amount of DNA adsorbed on particles using stoichiometric (3 μg) or excess (81 μg) amounts of DNA to form the corona. The excess amount of DNA (81 μg) was chosen to match the concentration used in our previous study (13). (A) Concentration (log) of DNA adsorbed on the surface of particles. The particle concentration was normalized to 500 pM to allow for comparison between different particle diameters. Significance was determined using multiple unpaired t-tests. ^∗p < 0.05, n = 3 distinct samples. Error bars reflect ±1 SE. Similar DNA concentrations were obtained using a direct spectroscopic measurement (Table S1). (B) TEM images of particles with a DNA corona formed using an excess concentration of DNA. The images for the 200 nm and 2 μm particles are repeated from Fig. 2B for comparison. Confocal fluorescence microscopy was also used to image the DNA and 2 μm particles showing the relative concentration of DNA on the particles and particle aggregation (Figs. S1 and S2).

DNA-particle diameter is dependent on the amount of DNA in the corona

As described above for 200 nm and 2 μm particles (Fig. 2 B; Table 2), the amount of DNA used to form the corona altered the DNA-particle diameter and aggregation of the 1 and 10 μm particles (Fig. 3 B; Table 3). We next assessed the diameters of all DNA-particle complexes with coronas formed using either a stoichiometric or excess amount of DNA. For 1 and 2 μm particles, particles with both stoichiometric and excess DNA coronas are similar in diameter to bare particles and fairly monodisperse. For the 10 μm particles, coronas formed using excess DNA, and aggregates have a larger size than when coronas are formed using a stoichiometric amount of DNA (71,028 and 39,208 nm, respectively). In addition, the amount of DNA associated with the 10 μm particles (excess DNA) is ∼5× greater than the amount of DNA expected based on scaling by surface area.

Table 3.

DLS of Bare Particles and Particles with a DNA Corona Formed Using a Stoichiometric (3 μg) or Excess (81 μg) Amount of DNA

Particle	DNA Amount	dh (nm)	Δ dh (nm)	PdI	Δ PdI
200 nm	stoichiometric	2800.3 ± 389.9	2602	0.39 ± 0.05	0.39
200 nm	excess	45,620 ± 2714.2	45,422	0.58 ± 0.4	0.58
1 μm	stoichiometric	1879.7 ± 32.1	163	0.14 ± 0.09	−0.04
1 μm	excess	1323.7 ± 26.0	−393	0.15 ± 0.06	−0.03
2 μm	stoichiometric	1738.0 ± 114.5	−244	0.28 ± 0.2	0.13
2 μm	excess	3020.3 ± 308	1038	0.20 ± 0.1	0.05
10 μm	stoichiometric	66,706 ± 8031.6	39,208	0.43 ± 0.31	−0.18
10 μm	excess	98,526 ± 73,204.0	71,028	0.73 ± 0.28	0.12

Change (Δ) relative to bare particles is included for d_h and PdI. Values for 200 nm and 2 μm particles with DNA coronas are repeated from Table 2.

Particles protect DNA from degradation by DNase I

Our previous work showed that DNA adsorbed on the surface of particles (200 nm and 2 μm) is protected from degradation by DNase 1 (13). In comparison, free DNA in solution is completely degraded by DNase 1 (Fig. S3). Expanding the scope of this work to a wider range of particle diameters (200 nm, 1 μm, 2 μm, and 10 μm) and varying concentrations of DNA coronas (stoichiometric and excess) shows a more complex response to DNase 1 degradation (Fig. 4). For the 200 nm particles, a corona formed from a stoichiometric concentration of DNA shows nearly complete protection from DNase 1-mediated degradation (27% ± 10% degraded). A value of 1 μg/mL DNA would show complete (100%) protection from degradation. In comparison, a corona formed from an excess concentration of DNA shows minimal (74% ± 0.6% degraded) protection from DNase 1. For 1 and 2 μm particles, coronas formed from excess DNA were slightly more protected than coronas formed from stoichiometric concentrations of DNA. For 10 μm particles and planar surfaces, there were no significant differences in degradation between stoichiometric and excess amounts of DNA. In addition, planar surfaces show minimal (86% ± 0.1% degradation for both excess and stoichiometric) protection of DNA from degradation by DNase 1.

Figure 4.

Concentration of DNA present on particles and on a planar surface after treatment of DNA-particle complexes with DNase 1 (1 U/μg; 30 min, 37°C). The concentration of DNA present is normalized to the treatment of the same mass of DNA in the absence of particles. A value of 1 μg/mL DNA shows 100% protection from degradation. n = 3 distinct samples. Error bars reflect ±1 SE. Significance was measured using two-way ANOVA with Tukey’s multiple comparisons post hoc. ^∗p < 0.05, ^∗∗∗∗p < 0.0001; ns, not significant.

The difference in degradation of the DNA corona formed from stoichiometric and excess DNA for the 200 nm particles is notable and provides useful information to understand the mechanisms by which particles protect DNA from degradation. We hypothesize that the increase in degradation with an excess corona for 200 nm particles is due to increased aggregation of particles (Fig. 2 B; Tables 2 and 3). We propose that these large aggregates have more exposed DNA that is easily accessible by DNase 1. For the larger particles, which remain relatively monodisperse with both stoichiometric and excess DNA coronas, we suggest that the DNA on the particle surface is more highly associated with the particle surface and less accessible to DNase 1.

Coronas formed from sheared DNA (500–600 bp) are protected from DNase1 degradation

To better understand the effect of DNA length on the ability of particles to protect DNA from DNase 1-mediated degradation, we repeated the experiments with sheared DNA (DNA_S) (500–600 bp) and four particle diameters (40 nm, 200 nm, 1 μm, and 2 μm). Although 40 nm particles incubated with the longer DNA (∼10,000 bp) described above were not pipettable, we found DNA_S was usable. This may be due to its shorter length (∼150 nm) being closer in order of magnitude to the particle diameter. The 10 μm particles were not included because we found no significant difference in DNA degradation as a function of corona concentration for the longer DNA (Fig. 4).

As with the unsheared DNA, we found that using increased amounts of DNA_S (low to high molar ratios, as described in Materials and Methods) to interact with the particles led to increased amounts of DNA_S present in the corona (Fig. 5 A). In terms of resistance to DNase 1-mediated degradation (Fig. 5 B), the results for 1 and 2 μm particles are similar to unsheared DNA with increased amounts of DNA remaining on the particle surface for DNA-particle complexes formed with high concentrations of DNA_S and no extensive aggregation observed (Fig. 5 C; Table 4). Like the 200 nm particles and unsheared DNA, the 40 nm particles and DNA_S show extensive aggregation (Fig. 5 C; Table 4) and little protection from degradation (Fig. 5 B), which we attribute to particle aggregation and access of DNA by DNase 1. Interestingly, the 200 nm particles show a response to DNase 1 that is similar to the unsheared DNA (Fig. 4) with extensive degradation (74% ± 2% degraded) of the corona formed from high concentrations of DNA_S, but not low concentrations of DNA_S (15% ± 6% degraded). However, the shorter DNA_S does not lead to the same aggregation of particles observed for the longer DNA (Tables 3 and 4).

Figure 5.

Characterization of particles with a DNA corona formed using DNA_S. (A) Concentration of DNA_S adsorbed on the surface of particles. The particle concentration was normalized to 500 pM to allow for comparison between different particle diameters. (B) Concentration of DNA_S present on particles after treatment of DNA_S-particle complexes with DNase 1 (1 U/μg; 30 min, 37°C). The concentration of DNA_S present is normalized to the treatment of the same mass of DNA_S in the absence of particles. A value of 1 μg/mL DNA shows 100% protection from degradation. (C) TEM images of particles with a DNA_S corona formed using a high concentration (81 μg) of DNA. Error bars reflect ±1 SE. Significance for (A) was determined using multiple unpaired t-tests. ^∗p < 0.05. n = 3 distinct samples. Significance for (B) was measured using two-way ANOVA with Tukey’s multiple comparisons post hoc. ^∗∗∗∗p < 0.0001; ns, not significant. n = 3 distinct samples.

Table 4.

DLS of Bare Particles and Particles with a DNA_S Corona Formed Using a Stoichiometric (3 μg) or Excess (81 μg) Amount of DNA

Particle	DNA Amount	dh (nm)	Δ dh (nm)	PdI	Δ PdI
40 nm	low	342.2 ± 13.2	104	0.19 ± 0.05	0.02
40 nm	high	15,296 ± 16,234	15,058	0.73 ± 0.27	0.56
200 nm	low	1416.3 ± 151.3	1218	0.37 ± 0.03	0.37
200 nm	high	1323.0 ± 112.1	1125	0.38 ± 0.12	0.38
1 μm	low	1622.3 ± 144.4	−94	0.24 ± 0.05	0.06
1 μm	high	2171.0 ± 578.5	455	0.22 ± 0.17	0.04
2 μm	low	2482.0 ± 77.6	500	0.30 ± 0.12	0.15
2 μm	high	3250.3 ± 1118.2	1268	0.46 ± 0.26	0.31

Change (Δ) relative to bare particles is included for d_h and PdI.

Protection of ssDNA in a DNA corona differs from dsDNA coronas

The smaller persistence length of ssDNA (0.7–6 nm) compared to dsDNA (∼50 nm) provides an interesting feature for comparison of the protective effect of DNA adsorbed on particles (18,19,20). Experiments were repeated using ssDNA (∼10,000 bp) generated from the same unsheared dsDNA as described above with three particles diameters (200 nm, 1 μm, 2 μm). Since DNase 1 is specific for dsDNA, we used S1 nuclease for degradation experiments. S1 is an endonuclease derived from Aspergillus oryzae that cleaves ssDNA but not dsDNA (21). All experiments were carried out with excess DNA with a focus on the protective effect of the particles.

We hypothesized that, due to the shorter persistence length, ssDNA will have greater flexibility than dsDNA in the interaction with particles, thereby allowing even greater resistance to enzyme degradation than dsDNA of identical length. We first measured the amount of ssDNA present in the corona (Fig. 6 A), which increased with increasing particle diameter as expected. Unlike dsDNA, we find that the protective effect of the particles is much greater for the 200 nm particles with no significant difference between S1 nuclease-treated ssDNA-particle complexes and an untreated control (Fig. 6 B). Interestingly, the 200 nm particles incubated with excess ssDNA show a similar level of aggregation (d_h = 2612 nm) observed with stoichiometric dsDNA (d_h = 2602 nm) (Fig. 6 C; Table 5). For the 1 and 2 μm particles with an ssDNA corona, DNA is much more efficiently degraded with S1 nuclease (Fig. 6 B). This is the opposite trend to that observed with 1 and 2 μm particles with a dsDNA corona. It is possible that the increased flexibility of ssDNA allows it to be degraded efficiently when adsorbed onto small clusters of particles such as the case with the 1 and 2 μm particles (Fig. 6 C; Table 5). When particles aggregate significantly, as is the case with 200 nm particles, the increased flexibility of ssDNA may allow ssDNA to wrap between large clusters making it inaccessible to S1 nuclease, preventing degradation (Fig. 6 C). The smaller radius of curvature of 200 nm particles may support DNA wrapping, thereby hindering nuclease activity. Together, these studies indicate that persistence length is a factor in the protective effect of DNA adsorbed on particles.

Figure 6.

Characterization of particles with a DNA corona formed using ssDNA. (A) Concentration of ssDNA adsorbed on the surface of particles. The particle concentration was normalized to 500 pM to allow for comparison between different particle diameters. DNA corona was formed using excess (81 μg) ssDNA. (B) Concentration of ssDNA present on particles after treatment of ssDNA-particle complexes with S1 nuclease (5 U/μg; 30 min, 37°C) in comparison to an untreated control. The concentration of ssDNA present is normalized to the treatment of the same mass of ssDNA in the absence of particles. A value of 1 μg/mL DNA shows 100% protection from degradation. (C) TEM images of particles with an ssDNA corona formed using an excess amount of DNA. Error bars reflect ±1 SE. Significance for (A) and (B) was determined using two-way ANOVA with Tukey’s multiple comparisons post hoc. n = 3 distinct samples. ^∗p < 0.05, ^∗∗∗∗p < 0.0001; ns, not significant. n = 3 distinct samples.

Table 5.

DLS of Bare Particles and Particles with an ssDNA Corona Formed Using an Excess (81 μg) Amount of DNA

Particle	DNA Amount	dh (nm)	Δ dh (nm)	PdI	Δ PdI
200 nm	excess	2610.7 ± 668.1	2612	0.58 ± 0.37	0.58
1 μm	excess	3079.7 ± 197.1	1363	0.51 ± 0.30	0.33
2 μm	excess	4757.7 ± 293.1	2775	0.52 ± 0.12	0.37

Change (Δ) relative to bare particles is included for d_h and PdI.

Conclusions

Overall, this work describes a complex relationship between DNA-particle interactions and nuclease resistance. We observe that DNA of all types (500–600 bp, ∼10,000 bp, dsDNA, ssDNA) forms a corona on synthetic particles that increases with the concentration of DNA used to form the corona (Figs. 2, 3, 5 A, and 6 A). The formation of a DNA corona protects the DNA from degradation by nucleases (Figs. 4, 5, and 6), which may have important implications for autoimmune disease (13). Particles of all diameters (40 nm to 10 μm) provide some protective effect for all types of DNA (500–600 bp, ∼10,000 bp, dsDNA, ssDNA), but the level of protection depends on a complex combination of DNA length and persistence length, particle diameter, and especially DNA-induced particle aggregation (Fig. 7). Our work suggests that smaller particles are more aggregated by the addition of dsDNA (Tables 2, 3, and 4) and that these DNA-particle aggregates are less protective for DNA degradation by DNase 1 (Figs. 4 and 5 B). In comparison, when DNA coronas are formed with ssDNA, our work shows that even ssDNA-particle aggregates are more resistant to degradation by S1 nuclease (Fig. 6 B), suggesting DNA persistence length is a factor in nuclease resistance.

Figure 7.

Schematic illustration of possible DNA coronas. (A) dsDNA corona on 200 nm and 2 μm particles formed using a stoichiometric amount of DNA. Particles are not fully covered with DNA when a corona is formed using a stoichiometric amount of DNA. The 200 nm particles have low levels of aggregation. (B) dsDNA corona on 200 nm and 2 μm particles formed using an excess amount of DNA. Particles are fully covered with DNA when a corona is formed using an excess amount of DNA. The 200 nm particles have high levels of aggregation. (C) ssDNA corona on 200 nm and 2 μm particles formed using an excess amount of DNA. Particles are fully covered with DNA when a corona is formed using an excess amount of DNA.

The use of polystyrene particles as a model system to address this question provides a well-controlled set of particles. In comparison, previous work examined the nuclease resistance of DNA on the surface of cationic liposomes and solid lipid nanoparticles (24,25,26), which included the additional factor of DNA-lipid interactions and possible encapsulation. This previous work also considered changes in DNA structure in response to interaction with lipid surfaces that were not considered here but that do provide directions for future work (27, 28). In addition, previous work with oligonucleotides conjugated to the surface of gold nanoparticles has pointed to the importance of packing density and salt concentration at the particle surface (29, 30), although these thiol-linked particles provide a different DNA-particle interaction than the electrostatic adsorption of DNA directly onto the surface particle leading to the DNA corona described here.

We hope this model system and these studies will aid in the understanding of cell-free DNA-particle interactions in autoimmune disease, including the possibility of targeted nuclease-mediated treatments capable of degrading DNA associated with particles (2,3,31). A relationship between particle size and SLE has been noted previously, but has proved difficult to unravel. For example, extracellular vesicles isolated from SLE patients and measured with DLS are larger, defined as the fraction between 306 nm and 1 μm, than those from healthy patients (32). Larger extracellular vesicles and microparticles in plasma isolated from SLE patients and synovial fluid from patients with rheumatoid arthritis have also been observed (33,34), but it is possible that the larger diameter is due to the inclusion of mitochondria in the particles or immunoglobulin binding to form immune complexes. We hope that the use of a well-controlled model system of polystyrene particles and exogenous DNA can help probe some of these highly complex aspects of the in vivo systems.

Data availability

All study data are included in the article and/or the supporting material.

Acknowledgments

This research was supported by NIH grant 1R21-AI175926. A portion of this work was performed at the Duke University Shared Materials Instrumentation Facility (SMIF), a member of the North Carolina Research Triangle Nanotechnology Network (RTNN), which is supported by the National Science Foundation (award number ECCS-2025064) as part of the National Nanotechnology Coordinated Infrastructure.

Author contributions

F.A., D.A.M., D.S.P., and C.K.P. conceived the experiments. F.A., D.A.M., T.K., and A.K. conducted the experiments. F.A., D.A.M., T.K., and A.K. analyzed the results. F.A., D.S.P., and C.K.P. wrote the manuscript. All authors reviewed the manuscript.

Declaration of interests

The authors declare no competing interests.

Editor: Jason Kahn.