DNA damage – a novel method to measure the rate of single-stranded breaks from fragmenting dsDNA in nanochannels – theory and modeling

Abstract

Damage to DNA can cause death to an individual cell and serious harm to the host organism. Photosensitized reactions are one cause of DNA damage. It can lead to destructive chemical reactions targeted to a base of the DNA as well as a breakage along one or both of the DNA strands. Due to this, quantifying and understanding photosensitized driven DNA damage is an important topic of research.

From experimental data of fragmenting fluorescently stained linear double-stranded DNA in nanochannels, we will extract the non-observable single-stranded cleavage rate (nicking rate) from observable times of double-stranded cleaves (cuts), which lets us quantify the rate of DNA damage. To do this, we present a new probabilistic model that connects the cutting rate to the time of cuts. We present two distinct models for the cutting rate, the first one is analytical and the second is based on simulations.

We find through validation on synthetic data with known nicking rates that using the cutting rate from the simulation-based model yields more accurate estimates of the nicking rate compared to using the cutting rate from the analytical model. In addition, we manage to estimate the nicking rate for three experimental data sets with varying illumination strength. From these estimates, we conclude that the nicking rate, as expected, increases with increasing illumination strength.

We hope that this study will serve as proof of concept for our new methodology to estimate the nicking rate and provide a good starting point for other studies which want to add to the knowledge of nicking rate estimation under different conditions.

Popular Abstract

Claiming around 8 million lives yearly, cancer is one of the deadliest diseases world-wide. Due to its negative impact on human well-being, improving our knowledge about cancer and working towards more effective treatments in the future is essential.

Accumulation of damage to our genes can lead to cell death or severe harm to the host organism. To better understand the processes that result in gene damage, it is important to study deoxyribonucleic acid (DNA) damage. DNA damage can be studied in various ways, e.g., by exposing the DNA to, cutting enzymes, high energy radiation, oxygen or visible light. In this work, we study DNA damage caused by the exposure of oxygen and visible light.

We know that harmful reactants, such as free radicals, take part in chemical reactions that can cause damage to our DNA. We also know that light and oxygen can increase the creation of free radicals. The rate of DNA damage is crucial since our bodies need to keep up with the reparation processes and avoid accumulation of mutations. It is thus important to ask how oxygen and light change the rate of DNA damage, which are both present in our daily lives.

In this work, we will look at DNA damage through experiments of DNA in nanometer-sized channels captured using a fluorescence microscope. Even though the scale of the experimental setup is very small, it is not small enough to obtain the DNA damage directly. To better understand what limits our observations of DNA damage, we begin by picturing the DNA as a spiraling ladder structure. The ladder's two side-rails and rungs, correspond to the two main strands and base-pairs of the DNA, respectively. Free radicals may chemically react with one of the strands and break it, or analogously damage one of the rails on the ladder. When such a single-stranded break (nick) happens the DNA is still held together as one molecule due to the connection of base-pairs to the other strand, just as the ladder is still in one piece held together by the rungs. This process of single-stranded breaks is not observable with the microscope. On the other hand, if yet another single-stranded break occurs close enough on the opposite strand on the DNA we obtain a double-stranded break (cut), it has the effect of damaging both side-rails between the same pairs of rungs on the ladder. In this case, the DNA molecule divides into two molecules, a constellation which can be observed with the microscope, once the DNA fragments has diffused apart within the nanochannel.

Since we want to know at which rate the nicks occur but can only observe the cuts, we are challenged to measure something non-observable. To tackle this, we here introduce a new stochastic model, which can estimate the nicking rate given a series of observed cutting times using a functional form for the rate of cuts. As the microscope has a limited resolution, we cannot observe the exact time of the cuts in the experiments and the functional form for the cutting rate must account for this fact. To do this, we have simulated all parts of the experiments to obtain the observable functional form for the cutting rate. By performing simulations for different nicking rates, we obtain different functional forms for the observed cutting rate. Each of these observed cutting rate functions, together with the observed cutting times, is then used in the stochastic model to deduce the likelihood of the current nicking rate. The observed cutting rate function that gives the highest likelihood corresponds to our estimate of the nicking rate.

With this study, we hope that our new method to measure the rate of DNA damage can find its way into the hands of more researchers who can keep expanding the knowledge of DNA damage and use our method to measure the nicking rate under different experimental conditions.

• author: Brander, Magnus
• supervisor: Tobias Ambjörnsson; Jason Beech
• organization: Department of Physics; Computational Biology and Biological Physics - Undergoing reorganization
• course: FYTM04 20202
• year: 2020
• type: H2 - Master's Degree (Two Years)
• subject: Physics and Astronomy
• keywords: biophysics, DNA damage, single-stranded breaks, double-stranded breaks, Bayesian inference, Brownian motion, Gillespie algorithm, photosensitized DNA damage
• language: English
• id: 9031024
• date added to LUP: 2020-10-19 10:08:48
• date last changed: 2020-11-04 15:42:58