Innate Immune Stimulation Using 3D Wireframe DNA Origami

Abstract

Three-dimensional wireframe DNA origami have programmable structural and sequence features that render them potentially suitable for prophylactic and therapeutic applications. However, their innate immunological properties, which stem from parameters including geometric shape and cytosine-phosphate-guanine dinucleotide (CpG) content, remain largely unknown. Here, we investigate the immunostimulatory properties of 3D wireframe DNA origami on the TLR9 pathway using both reporter cell lines and primary immune cells. Our results suggest that bare 3D polyhedral wireframe DNA origami induce minimal TLR9 activation despite the presence of numerous internal CpG dinucleotides. However, when displaying multivalent CpG-containing ssDNA oligos, wireframe DNA origami induce robust TLR9 pathway activation, along with enhancement of downstream immune response as evidenced by increases in Type I and Type III interferon (IFN) production in peripheral blood mononuclear cells. Further, we find that CpG copy number and spatial organization each contribute to the magnitude of TLR9 signaling and that NANP-attached CpGs do not require phosphorothioate stabilization to elicit signaling. These results suggest key design parameters for wireframe DNA origami that can be programmed to modulate immune pathway activation controllably for prophylactic and therapeutic applications.

Keywords: 3D wireframe DNA origami; CpG; TLR9; immunomodulation; immunostimulation; multivalency.

Figure 1.. NANP design and characterization.

(A) DNA NANPs of varying sizes and geometries were fabricated using staples that were either unmodified or extended with ssDNA overhangs containing CpG motifs. The location and copy number of these CpG overhangs can be precisely controlled. Here we focused on two structures: (i) an icosahedron displaying up to 30 CpG overhangs, and a (ii) pentagonal bipyramid displaying up to 40 CpG overhangs. Red circles denote locations along NANP edges where CpG overhangs are displayed. (B) Design variations were used to explore the effects of CpG overhang copy number, inter-CpG distance, orientation, sequence, and oligonucleotide composition. (C) A fluorescent agarose gel shift assay was used to analyze the quality of NANP folding after spin column purification. Pentagonal bipyramids displaying 0, 10, 20, or 40 ssDNA overhangs were hybridized to complementary Cy5-modified oligos. The gel image of the SybrSafe channel shows gel shifts corresponding to the increasing molecular weight of each construct, while the gel image taken in the Cy5 channel exhibits an increase in band intensity due to the successive increase in Cy5-CpG copy number on each NANP. Similarly, the increase in Cy5-CpG valency is demonstrated by the shift from SybrSafe green to Cy5 red in successive lanes in the SybrSafe/Cy5 overlay gel image. Gel shift assays for icosahedral constructs are shown in Figure S1. (D) Dynamic light scattering (DLS) was used to evaluate NANP hydrodynamic diameter and polydispersity. Representative DLS measurements of an unmodified pentagonal bipyramid and an icosahedron are shown; measurements for all other NANP constructs tested are shown in Figure S2.

Figure 2.. Magnitude of TLR9 activation can be modulated by varying copy number of CpG overhangs (CpG-OHs) displayed on NANPs.

The strength of TLR9 activation corresponds to the valency of CpG-OHs for both (A) PB84 and (B) Ico42 NANPs. The overall concentration of CpG-OHs was held constant across samples by decreasing NANP concentrations correspondingly. For all assays, baseline absorbances were determined from the PBS control and subtracted from experimental samples. Absorbances were then normalized to the ODN2006 positive control (pos. con). Data show the average absorbance of samples in triplicate with standard error, where n = 3 biologically independent assays performed on separate days. P values are from a one-sided analysis of variance (ANOVA) with a post-hoc Tukey test to correct for multiple comparisons (*: P ≤ 0.05; **: P ≤ 0.01, ***: P ≤ 0.001, ****: P ≤ 0.0001). All unlabeled pair-wise comparisons are not significant.

Figure 3.. Magnitude of TLR9 activation is dependent on inter-CpG distance, CpG-OH nanoscale organization, and NANP geometry.

(A) TLR9 activation measured in response to stimulation with PB84 displaying 10 CpG-OHs at varying inter-CpG distances shows that the strongest activation is induced by the construct in which adjacent CpGs were displayed 7nm apart. (B) PB84 constructs were folded with either one or both sides of the NANP displaying 5 CpG-OHs. Total CpG-OH concentration was kept constant across all samples. TLR9 activation is significantly stronger for constructs in which both sides of PB84 were modified with CpG-OHs compared to their single-sided counterparts. (C) Significant differences in the magnitude of TLR9 activation were observed in response to stimulation by CpG overhangs displayed on different NANP geometries folded from the same phPB84 scaffold: a tetrahedron (Tet210), octahedron (Oct105), PB84, ICO42, and a dodecahedron (Dod42). Total CpG-OH concentration was kept constant across all samples. Baseline absorbances were determined from the PBS control and subtracted from experimental samples. Absorbances were then normalized to the ODN2006 positive control (pos. con). Data show the average absorbance of samples in triplicate with standard error, where n = 3 biologically independent assays. P values are from a one-sided analysis of variance (ANOVA) with a post-hoc Tukey test to correct for multiple comparisons (*: P ≤ 0.05; **: P ≤ 0.01, ***: P ≤ 0.001, ****: P ≤ 0.0001). All unlabeled pair-wise comparisons are not significant.

Figure 4.. Interferon expression levels in PBMCs can be modulated by engineered immunostimulatory NANPs.

Elevated levels of (A) IFNα, (B) IFNβ, (C) IFNλ, and (D) IFNω were observed in response to PB84 constructs displaying 20 or 40 CpG-OHs compared to unmodified 0 CpG-OH PB84, while the addition of 20 or 30 CpG-OHs to ICO42 constructs had minimal effect in some cases and reduced interferon expression levels in others. NANP concentrations were adjusted across samples to ensure constant CpG concentration for all modified constructs. A mixture of ODN2216, lipopolysaccharide, and phytohemagglutinin was used as the assay positive control (pos. con). Each bar represents averaged triplicate data from a single donor, where n = 3 donors. P values are from a one-sided analysis of variance (ANOVA) with a post-hoc Tukey test to correct for multiple comparisons (*: P ≤ 0.05; **: P ≤ 0.01, ***: P ≤ 0.001, ****: P ≤ 0.0001). All unlabeled pair-wise comparisons are not significant.