Single-Molecule Biodosimetry

Abstract

Inferring characteristics of radiation exposure using biological molecules is extremely challenging. Current methods, in particular, lack a clear connection between dose and molecular response. Here, we demonstrate that resistive-pulse nanopore sensors enable single-molecule biodosimetry by quantifying the frequency of double-strand DNA scissions versus γ radiation dose. The resulting response curve shows an elongated Gaussian behavior, reminiscent of cell survival rates versus dose. We demonstrate that the competition of radical damage of DNA─i.e., single-strand lesions that lead to breakage─with bimolecular radical loss captures the form of the response. Our sensors and protocol provide a foundation for numerous technological advances. These include rapid dosimetry for triage in emergency situations and ex vivo monitoring of radiotherapy effectiveness in order to tailor treatment to patient- and tumor-specific response.

1

Nanopore-based single-molecule dosimetry. (a) An aqueous solution of 2.5 kbp DNA is irradiated with a ⁶⁰Co calibrated γ source. (b) Single–strand breaks in the sugar–phosphate backbone accumulate until two breaks occur close enough (here shown as l = 3 bases) to compromise the stability of the molecule and cause a double–strand break. (c) Irradiated DNA sample is quantified post–exposure with a glass nanopipette. (d), Histograms of the concentrations of the irradiated 2.5 kbp and unirradiated internal standards of 5 kbp and 10 kbp DNA as a function of dose (L ₀ = 1 bp). Absolute size and concentration are calculated from the ECD and capture frequencies by calibrating against these two internal standards.

2

Molecular standards as an internal calibration and ruler. (a) Ionic current versus time showing resistive pulses of the 5 and 10 kbp DNA internal standards and 1.0 Gy γ irradiated 2.5 kbp DNA. (b) Ionic current of an identical DNA mixture in which the 2.5 kbp DNA has been irradiated at 15.0 Gy. Characteristic current events for the irradiated length, 2.5 kbp (left), and the two internal standards, 5 kbp (center) and 10 kbp (right), are shown for 1.0 and 7.5 Gy enlarged in the boxes. The ECD is the area, A, shaded in orange or blue. (c) Log­(ECD) histograms for the 1 and 7.5 Gy samples without calibration (here, A ₀ = 1 pC). (d) The same data sets after calibration against the internal standard peaks (here, L ₀ = 1 bp). After calibration, the final 1.0 and 7.5 Gy histograms show a decreasing concentration of intact 2.5 kbp DNA with dose.

3

DNA dose response. The intact DNA concentration (blue circles and error bars) versus dose shows that the response is roughly Gaussian. At very low dose, one expects such behavior since bimolecular radical decay will be negligible. However, the best fit (red, dashed line) at low dose, ≤3 Gy, to a Gaussian model does not work well at high doses, overestimating the loss of intact DNA. A best fit to the whole range of data (not shown) similarly does not fit well (overestimating intact DNA at intermediate doses and underestimating at high doses), as the shape of the curve is not actually Gaussian in D but an elongated Gaussian. Including bimolecular decay yields a two parameter model and a high–quality fit to the data (orange, dotted line). The inset shows the measured concentration of three key molecules. The concentration of 2.5 kbp DNA decreases with increasing dose. Error bars are a scaled average of the standard deviation between the measured concentrations of the 5 and 10 kbp internal standards (that were fixed between runs). The scaling is based on the length of the data set and the number of 2.5 kbp events recorded at a particular dose.

4

Fragment concentration bias. (a) A comprehensive molecular ladder of ≤10 kbp DNA run through a pipet prepared identically to those used in the primary experiments. Separate ladders containing 100, 250, 600 bp, 1.5, 2.5, 5, and 10 kbp fragments (red), and 300, 900 bp, 2, 5, and 10 kbp fragments (blue) were recorded and aligned as described in the Methods to minimize overlap. (b) Molecular size distribution for all captured molecules after exposure to 15.0 Gy of radiation. Fragments between 100 bp and 2.5 kbp are visible below the 2.5 kbp peak indicated with the red arrow (L ₀ = 1 bp). (c) The probability, P _C , of successfully capturing a DNA fragment of a given size as calculated from the ladder shown in (a). Error bars are the standard deviation of the best–fit parameters of a logarithmic fit (black, dotted line) to the capture rates of the ladder run through three identically prepared pipettes. (d) The ratio of nucleotides in fragments (the difference between a “fragment” and the intact Gaussian fitted peak being >2 standard deviations) to nucleotides in the intact 2.5 kbp DNA after correcting for the deflated P _C of small molecules.