Isolation of nucleic acids using liquid-liquid phase separation of pH-sensitive elastin-like polypeptides

Abstract

Extraction of nucleic acids (NAs) is critical for many methods in molecular biology and bioanalytical chemistry. NA extraction has been extensively studied and optimized for a wide range of applications and its importance to society has significantly increased. The COVID-19 pandemic highlighted the importance of early and efficient NA testing, for which NA extraction is a critical analytical step prior to the detection by methods like polymerase chain reaction. This study explores simple, new approaches to extraction using engineered smart nanomaterials, namely NA-binding, intrinsically disordered proteins (IDPs), that undergo triggered liquid-liquid phase separation (LLPS). Two types of NA-binding IDPs are studied, both based on genetically engineered elastin-like polypeptides (ELPs), model IDPs that exhibit a lower critical solution temperature in water and can be designed to exhibit LLPS at desired temperatures in a variety of biological solutions. We show that ELP fusion proteins with natural NA-binding domains can be used to extract DNA and RNA from physiologically relevant solutions. We further show that LLPS of pH responsive ELPs that incorporate histidine in their sequences can be used for both binding, extraction and release of NAs from biological solutions, and can be used to detect SARS-CoV-2 RNA in samples from COVID-positive patients.

Figure 1

Schematic illustration of the two-step LLPS procedure for extraction of SARS-CoV-2 RNA from a nasopharyngeal swab sample prior to RT-qPCR.

Figure 2

Gel retardation assays showing the binding of ELP fusion proteins and NAs in biologically relevant fluids. Agarose gels illustrate the binding activity of 0.1 mM E1.40COR30 (lanes 1, 3 and 5) and 0.1 mM E3.10 (lanes 2, 4, and 6) with either (A) 0.1 mg/mL tRNA or (B) 0.5 µM ssDNA1 in buffer (lanes 1 & 2), diluted artificial saliva solution (lanes 3 & 4), or diluted artificial nasopharyngeal fluid (lanes 5 & 6). Binding is indicated by retardation of NA electrophoresis.

Figure 3

Recruitment of ssDNA1 and tRNA into ELP coacervates upon LLPS at 50 °C in different biologically relevant fluids. (A) Illustration depicting the workflow followed to examine recruitment of NAs into ELP coacervates upon LLPS in different media. To liberate the NAs from the ELPs prior to running the gels samples were incubated in a stopping buffer. Agarose gels illustrate the recruitment of (B) 0.1 mg/ml tRNA and (C) 0.5 µM ssDNA1 RNA into the protein rich phase (PRP) of 0.1 mM E1.40COR30 and 0.1 mM E3.10 upon LLPS. Smearing in gel lanes may be due to NA degradation or transient protein association.

Figure 4

pH-sensitive charge and LCST behavior of H-20 and H-24. (A) Estimation by SnapGene of molecular charge vs pH for H-20 and H-24. (B) Characterization of the cloud point transition temperature (T_t) for LLPS of 0.5 mM H-20 and H-24. T_t was measured in the absence and presence of 0.5 μM ssDNA1 in pH 6 buffer (37 mM citric acid/126 mM Na₂HPO₄) and pH 9 buffer (100 mM Tris). ****: p < 0.0001; ns: not significant.

Figure 5

Gel retardation assays showing the binding of His-ELPs and NAs at pH 6. Agarose gels illustrate the concentration dependent binding activity of H-20 and H-24, but not of E3 with (A) 0.5 μM ssDNA1 and (B) 0.5 mg/ml tRNA in 37 mM citric acid/126 mM Na₂HPO₄ buffer at pH 6. (H-20 and H-24 concentrations are: 0.1, 1, 5, 10, 25, 50 and 100 μM. E3 concentrations are: 10, 100, 1000 μM).

Figure 6

Fluorescence and brightfield microscopy of phase-separated aqueous microdrops in oil containing 0.5 mM H-24 and 0.5 μM ATTO488-labeled ssDNA1. (A, left) pH 8 (100 mM Na₂HPO₄/300 mM NaCl); (B, right) pH 6 (37 mM citric acid/126 mM Na₂HPO₄). Fluorescence microscopy (top: A1,B1) and bright field-fluorescence overlay (bottom: A2,B2). Images are taken after 20 min at 62 °C in a fully phase-separated state. Scale bars = 50 μm.

Figure 7

Gel retardation assays showing the lack of binding of His-ELPs at basic pH (4 °C). Agarose gels illustrate the absence of appreciable binding at 4 °C of 100 μM H-20 and H-24 with 0.5 mg/mL tRNA (lanes 1–3) and 0.5 μM ssDNA1 (lanes 4–6) in 100 mM Tris buffer at pH 9.

Figure 8

Recruitment of ssDNA1 into H-24 coacervate from different physiologically relevant solutions and subsequent release upon LLPS after pH shift. (A) Workflow of two-step NA isolation assay. (B) Fluorimetry measurements of an ATTO488-labeled ssDNA1 in the supernatant (SN, circles, dark gray bars) and coacervate (squares, light gray bars) were taken after LLPS 1 and LLPS 2 for the three physiologically relevant solutions. ****: p < 0.0001; ***: p < 0.001; **: p < 0.01.