‘Gel-Stacks’ gently confine or reversibly immobilize arrays of single DNA molecules for manipulation and study

Abstract

Large DNA molecules (>20 kb) are difficult analytes prone to breakage during serial manipulations and cannot be ‘rescued’ as full-length amplicons. Accordingly, to present, modify and analyze arrays of large, single DNA molecules, we created an easily realizable approach offering gentle confinement conditions or immobilization via spermidine condensation for controlled delivery of reagents that support live imaging by epifluorescence microscopy termed ‘Gel-Stacks.’ Molecules are locally confined between two hydrogel surfaces without covalent tethering to support time-lapse imaging and multistep workflows that accommodate large DNA molecules. With a thin polyacrylamide gel layer covalently bound to a glass surface as the base and swappable, reagent-infused, agarose slabs on top, DNA molecules are stably presented for imaging during reagent delivery by passive diffusion.

Method summary

Gel-Stacks technology provides multiple non-covalent molecular presentation modes, coupled with an unusually facile reagent delivery system designed for large-scale analytes, enhancing live imaging and manipulation. Enhanced further by modeling and software, Gel-Stacks technology becomes adaptable to a broad range of experimental applications.

Keywords: :

• confinement
• epifluorescence microscopy
• Fick's second law
• hydrogels
• large DNA molecule arrays
• reagent transport

The presentation of large, individual DNA molecules as arrays that facilitate manipulation and analysis is challenging [1]. Large DNA molecules are prone to breakage and benefit from workflows including as few, and as gentle, steps as possible. Although individual molecules have been locally confined within nanostructures [2,3], chemically tethered to surfaces [4], or adsorbed to charged glass surfaces [5,6], these approaches often require specialized fluidic devices and techniques or are not suitable for very large DNA molecules. Steric issues arise, especially when large DNA coils are confined, or covalently anchored to ‘hard’ surfaces for studies under conditions approximating near-free solution conditions [4]. Gel-Stacks (Figure 1A) deal with the antagonistic requirements for non-covalent localization of large DNAs onto a planar surface or in solution for critical imaging when using high numerical aperture objectives leveraging a shallow depth of focus (∼100's of nm). Importantly, Gel-Stacks use stacks of interchangeable agarose slabs to deliver reagents by diffusion in ways that simplify workflow steps and minimize steric effects. Agarose gel pore size (α) is concentration dependent, scaling as α = C^-γ, where γ is ∼0.6 [7,8], with pore diameters spanning 201–530 nm, meaning small molecules and even proteins enjoy nearly unhindered diffusion. In contrast, diffusion of large DNA molecules ‘into’ gels is highly constrained or controlled by size [9].

Figure 1. Gel-Stack preparation and reagent diffusion plots.

(A) Cross-sectional view (not to scale) of a completed Gel-Stack for epifluorescence microscopy showing individual DNA molecules (green squiggles) within an analyte fluidic layer sandwiched between a PAG base layer whose thickness is controlled by the size of polystyrene ‘PAG spacer beads’ and a blank agarose gel slab, topped by a reagent slab. Slabs are 1–2% low melting temperature agarose in high purity water, typically 15 mm × 15 mm × 1 mm in size. (B & C) Plots of the numerical solution of Equation 1(Equation 1) $\begin{array}{ll}\frac{\partial C}{\partial t}\hfill & =D\frac{{\partial }^{2}C}{\partial z^2}\hfill \\ C\hfill & =\{\begin{array}{l}{C}_0, 0<z\le L_{rl}\hfill \\ 0, L_{rl}<z<L_{rl}+L_{bl}\hfill \end{array}, t=0\hfill \\ \frac{\partial C}{\partial z}=0,\hfill & z=0,\hspace{1em}z=L_{rl}+L_{bl}\hfill \end{array}$(Equation 1) for a Gel-Stack with (B) L_rl = 1 mm, L_bl = 1 mm, (C) L_bl = 2 mm and, C₀ = 10 mM and D = ¿ the diffusion constant of the reagent (7.05 × 10^-10 m²/s; for Mg²⁺ [14]).

Gel-Stack experiments are simple to compose: analyte molecules are first added to the base layer, followed by stacked agarose gel slabs harboring reagents (Figure 1A). The base layer comprises a flat, very thin layer of polyacrylamide hydrogel (PAG) covalently bound to an acid-cleaned glass surface [6,10] derivatized by trimethoxyvinylsilane [1,6]. Importantly, the PAG base layer must be kept under buffer to avoid damage by drying. Analyte molecules are confined within a thin fluid layer between the PAG base layer and a contiguous agarose gel slab. Fluid layer thickness is varied by the volume of liquid placed in the mount and used to modify degree of molecular confinement. Reagent addition is temporally controlled by placement of an agarose ‘blank’ slab, whose thickness can be varied, between the DNA analyte and ‘reagent slab.’

The dynamics of Gel-Stack action are described by a one-dimensional, Fick's second law model that considers reagent diffusion by concentration, C, time, t and distance from the top of the reagent slab, z,(Equation 1) $\begin{array}{ll}\frac{\partial C}{\partial t}\hfill & =D\frac{{\partial }^{2}C}{\partial z^2}\hfill \\ C\hfill & =\{\begin{array}{l}{C}_0, 0<z\le L_{rl}\hfill \\ 0, L_{rl}<z<L_{rl}+L_{bl}\hfill \end{array}, t=0\hfill \\ \frac{\partial C}{\partial z}=0,\hfill & z=0,\hspace{1em}z=L_{rl}+L_{bl}\hfill \end{array}$(Equation 1) where D is the diffusion coefficient [11], C₀ is the initial concentration in the reagent slab, L_rl is the thickness of the “reagent” slab and L_bl is the thickness of the ‘blank’ slab. The initial concentration profile takes the form of a step function, representing an abrupt change in reagent concentration at the boundary between the reagent and the buffer slabs, leading to an instantaneous release of the reagent. The Neumann boundary conditions represent a zero flux at the top and bottom of the Gel-Stack. A Jupyter notebook tool (Supplementary material) calculates Gel-Stacks reagent delivery profiles as varied by slab thicknesses and reagent diffusion coefficients. As illustrated in Figure 1B & C, we characterize subsequent changes in Mg²⁺ concentration as a function of time and Z-axis coordinates for two blank slab thicknesses by numerical solution of Fick's second law.

Gel-Stacks facilitate temporal regulation of reagent delivery and time-lapse imaging of individual DNAs, either as diffusing coils in a thick fluid layer (video 1, Supplementary material) or mildly constrained between adjacent hydrogel layers. Figure 2A–G shows several Gel-Stack time course experiments tracking DNAse I cleavage of YOYO-1 stained T4 bacteriophage DNA (169 kb) by fluorescence intensity. This was achieved using blank slabs of varying thicknesses, topped by a reagent slab, to modulate the lag time before cleavage: 1 mm (Figure 2A–C) and 2 mm (Figure 2D–F). Figure 2G compares these experiments with the modelled Mg²⁺ cofactor concentration profile essential for DNAse I digestion. The workflow entailed removing buffer from the PAG base layer via aspiration, then adding 10 μl of pre-stained T4 DNA and 2 U of DNase I in 10 mM Tris, covered by a blank agarose slab and a reagent slab pre-soaked in Mg²⁺. Minimization of DNA penetration into hydrogel layers was ensured by porosities significantly smaller than the hydrodynamic radius of intact molecules. Consistent with our Fick's law modeling (Figure 1B & C), the use of a 1 mm blank slab resulted in DNAse I activity triggered by the diffusion-controlled introduction of the Mg²⁺ cofactor [12] after ∼3 min. In contrast, the 2 mm blank initiated action after ∼10 min. Subsequently, DNA molecules started being cleaved into mildly fluorescent fragments, with fluorescence intensity (Figure 2G) estimating DNA degradation over time [13].

Figure 2. Tracking enzymatic action within Gel-Stacks and demonstration of reversible immobilization.

(A–G) Time lapse images and integrated fluorescence intensity (FI) measurements track progressive DNase I cleavage of T4 DNA molecules confined between a PAG base layer and 1 mm (A–C) or 2 mm (D–F) blank agarose slab: (A) 3 min, (B) 4 min, (C) 6 min, (D) 10 min, (E) 15 min, (F) 25 min. (G) Evolution of Mg²⁺ concentration (solid lines) immediately above the DNA analyte layer modeled for two slab thicknesses (Z = 2 mm and Z = 3 mm) over a 45-min duration compared with parallel experiments (dashed lines: fit to datapoints) reporting FI. (H) Box and whisker plots showing retention rates of surface-bound molecules per microscope field (67 fields per plot, area = 14,870 μm²) after: spermidine condensation, removal of agarose slabs, and spd-buffer wash. (I–K) Images of T7 DNA molecules before (I), after (J) condensation and immobilization to the PAG base layer, and after slab removal and washing with spd-buffer (K). Scale bars = 5 μm.

We further enhance Gel-Stacks by enabling conditional immobilization of DNA molecules onto the PAG base layer using spermidine-mediated condensation effects. Specifically, we found that spermidine tightly binds DNA molecules to the PAG base layer during condensation. Importantly, this interaction is reversible: placement of a reagent slab of 1× TE triggers release of bound DNA molecules to the fluid interface. Tracking DNA molecules during condensation and decondensation is shown in Figure 2I–K, where the setup is as described in Figure 1A. Before condensation, T7 bacteriophage DNA molecules (40 kb) are freely diffusing (video 1, Supplementary material) within the fluid interface (Figure 2I; ∼250 μm thick) formed by first adding 50 μl of TE buffer (10 mM Tris, 1 mM EDTA, pH 7.6) to the PAG base layer, then topped by a ‘blank’ agarose slab. Placement of the spermidine infused reagent slab created a final, equilibrated 2 mM concentration, tightly binding DNA condensates to the PAG base layer (Figure 2J) that remained immobilized after removal of all agarose slabs. The surface was then washed with ‘spd-buffer’ (1 ml 1× TE, 1 mM spermidine) ensuring that molecules remained condensed and immobilized (Figure 2K). After placement of a 1× TE reagent slab, the DNA molecules decondensed and re-entered the fluid interface as random coils. Figure 2H presents box-and-whisker plots of condensed molecules retained per field, before and after removal of the agarose slabs and after washing with spd-buffer. These data show that DNA condensation fixes molecules onto the PAG base layer to resist removal by washing or swapping agarose slabs. On average, 80% of DNA molecules were retained on the PAG base layer after slab removal; of these, 96% were retained after washing, resulting in an overall 77% retention rate. Given that large DNA molecules do not readily diffuse or mix, a broad distribution of molecules per field was expected, as reflected by the spread of molecule counts/field (Figure 2H). Last, after condensation and removal of all agarose slabs, a brief wash with 1× TE alone decondensed molecules, but retained most of them. In contrast, no retention was observable without a prior spermidine condensation step (not shown).

A Gel-Stacks workflow, enhanced by a Jupyter notebook tool, supports chemical/enzymatic steps requiring live, single molecule imaging by offering localization of large DNAs presented as arrays. The fluid interface sandwiched between the 7 μm thick PAG base layer and an agarose slab lightly confines and can minimize translational diffusion of analyte DNA molecules for enabling time-lapse imaging by epifluorescence microscopy offering minimal depth of focus. Multiple chemical, or enzymatic manipulation steps are greatly facilitated by swappable reagent cassettes (proximal and distal to analyte layer) consisting of reagent-infused slabs of agarose gel, and further enhanced by reversible immobilization of DNA molecules by spermidine condensation. These advantages distinguish Gel-Stacks technology from conventional surface-based analyte fixation methods. In summary, the versatility of Gel-Stacks technology is underscored by its potential application to a diverse range of large-scale analytes. This encompasses isolated cellular organelles, controlled cell lysis workflows, and an extensive array of single molecule assays, which are now progressively incorporating exceptionally large DNA molecules.

Author contributions

All authors had substantial contributions to the conception or design of the work; or the acquisition, analysis or interpretation of data for the work; contributed to drafting the work or revising it critically for important intellectual content; gave final approval of the version to be published; and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Financial disclosure

A Vasquez-Echeverri and S Calle-Casteñeda were partly supported by COLCIENCIAS, Minciencias, Colombia, scholarship program 783. The authors also thank the NHGRI for funding: R21 HG012281. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.

Acknowledgments

The authors thank the good people, past and present, associated with and of the Laboratory of Molecular and Computational Genomics, namely S Krerowicz and L Pape

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, stock ownership or options and expert testimony.

Additional information

Funding

A Vasquez-Echeverri and S Calle-Casteñeda were partly supported by COLCIENCIAS, Minciencias, Colombia, scholarship program 783.