Effect of X-ray irradiation on ancient DNA in sub-fossil bones – Guidelines for safe X-ray imaging

Sub-fossilised remains may still contain highly degraded ancient DNA (aDNA) useful for palaeogenetic investigations. Whether X-ray computed [micro-] tomography ([μ]CT) imaging of these fossils may further damage aDNA remains debated. Although the effect of X-ray on DNA in living organisms is well documented, its impact on aDNA molecules is unexplored. Here we investigate the effects of synchrotron X-ray irradiation on aDNA from Pleistocene bones. A clear correlation appears between decreasing aDNA quantities and accumulating X-ray dose-levels above 2000 Gray (Gy). We further find that strong X-ray irradiation reduces the amount of nucleotide misincorporations at the aDNA molecule ends. No representative effect can be detected for doses below 200 Gy. Dosimetry shows that conventional μCT usually does not reach the risky dose level, while classical synchrotron imaging can degrade aDNA significantly. Optimised synchrotron protocols and simple rules introduced here are sufficient to ensure that fossils can be scanned without impairing future aDNA studies.

through an optical device (e.g. microscope, photographic objectives, optical taper). For a constant X-ray spectrum, the X-ray dose necessary to perform a scan with a given voxel size depends mostly on the detector properties. By opposition, in conventional sources, the magnification effect is obtained only thanks to the conical geometry of the Xray source, by displacing the sample along the X-ray cone. In this case, for constant X-ray spectrum, the dose necessary for a scan of a given voxel size depends on the distance between the source and the detector. This major difference between synchrotron and conventional X-ray source makes possible to derive X-ray dose relatively easily for conventional sources from calibration points by geometric calculation, whereas it is necessary to test all the different detectors combinations (scintillator/optic/sensor) for the synchrotron configurations.
The second critical aspect for X-ray dose is linked to the X-ray spectrum. Low energy X-rays are more easily absorbed by the samples than high energy X-rays. Nevertheless, low energy X-rays can also bring higher contrast level than high energy ones. Tomography requires that a sufficient amount of X-rays goes through the sample (typically the lowest transmission has to be above 10%, but often results are better for minimum transmission above 20%). In case of broad X-ray spectrum, the low energies can be completely absorbed by the sample when higher energies can go through ( Supplementary Fig. 1). This very well-known effect is called beam hardening, which is leading to typical artifacts when scanning with such broad spectrum. Nowadays, very efficient algorithms can correct most of these artifacts 1 , but it remains that the low energy photons are depositing a large part of the dose as they can be totally absorbed. Adapted filtering of the source spectrum using adapted metallic filter allows removing the lowest energies in order that the Xrays used to scan the specimen are really useful to obtain the data. When performing X-ray tomography, it is therefore important to balance all these effects to find the optimal configurations allowing the narrower X-ray spectrum, with energy high enough to reach sufficient transmission through the sample without going to too high energy that would tend to reduce contrast.
Synchrotrons are well known to allow imaging using monochromatic beam, that remove de facto the beam hardening effect 1,3 . Nevertheless, narrow polychromatic beam can bring results very close, without visible beam hardening, while allowing less ring artifacts (better beam profile and stability), and faster scans.
Since 2011, all the fossil specimens scanned on the beamline ID19 at the ESRF are imaged using these high quality direct "pink" (meaning narrow spectrum polychromatic) beams 4-6 . It has to be noted that the use of high quality pink beam or of monochromatic beam does not really have an impact on the delivered dose since in both cases the average energies are similar as well as the dynamic level on the detector.
Imaging of fossils with synchrotrons also implies in nearly all the cases the use of propagation phase contrast. This technique can be up to 1000 times more sensitive to small density differences than X-ray absorption used in conventional systems, and is often giving better results with energies higher than those used for absorption. Phase contrast is nowadays the most important reason why using synchrotron sources to image fossils, especially for observation of small structures such as incremental lines in teeth or bones microstructures. It is also the key to reduce the X-ray dose for sub-fossils by using as much as possible the high sensitivity given by this approach. All the results presented in the present study for low dose synchrotron imaging are based on propagation phase contrast and would not be relevant for pure absorption imaging.
Classical sub-μm resolution configurations could even reach the level of total destruction of aDNA, but these configurations are restricted to small irradiated volumes ( Supplementary Fig. 2) thanks to the precise beam collimation of synchrotron sources, i.e. the possibility to adapt the beam size to the field of view using absorbing slits systems.

Supplementary Figure 2.
Typical 3D dose deposition pattern of a sub-μm resolution scan for enamel microstructure as performed at the ESRF (3D simulation performed on the same lower first molar of the Engis 2 Neandertal child as in Supplementary Fig. 1). Only the central yellow part can reach the high dose level reported in the present study, the dose rapidly decreases for all the other parts crossed by the beam during the scan depending on the distance to the imaged part. Beam scattering (not simulated here) will also contribute to general dose level, but high resolution dosimetry experiments shows that it remains far less important than the dose deposition due to the direct beam, and its contribution decreases very rapidly with the distance to the direct beam (negligible after typically 200 μm in the geometry presented here).
Low resolution scans (voxel size larger than 20 μm, implying full irradiation of the specimens) were typically in the safe zone even with classical configurations, but the new configurations are well below the detection limit of 200 Gy.
The most detrimental configurations with the classical synchrotron scans were in the 5 μm range, where large areas were scanned while having substantial level of dose (typically complete teeth for dental development), but these scans would not have really endangered aDNA studies, except in case of multiple scans. The highest efforts for dose reduction were applied in this resolution range, as it was potentially the most dangerous one, and as it is a critical one for virtual dental and bones palaeohistology. The new configurations implemented on ID19 are now well below the detection limit.
Nowadays, only the sub-μm resolution scans typically used to observe enamel microstructures can still reach dose level that could have limited consequences on aDNA, but due to the small beam size, the concerned areas are very limited (typically small cylinders of 4*2 mm), and concern mostly enamel, where no sampling for aDNA would be done anyway. All in all, the complete set of configurations available at the ESRF for scanning of sub-fossils can be considered as safe for future aDNA studies, as long as good care is taken to perform the experiments (especially avoiding multiple scans whenever they are not necessary). Further efforts are ongoing to decrease dose for the sub-μm setup by further factor 2 to 3, and setups for voxel sizes larger than 10 μm by factor 2 to 10.
In the case of conventional X-ray imaging, it has to be noted that the high surface dose obtained for scans without any filters are due mostly to low energy X-rays. This effect was especially visible with the skyscan1273 scanner for which dose rates without filters were really higher than expected after the experiment on the BIR scanner. Even if the cause of this higher dose is not really clear, it appears to be due to the low energy part of the spectrum as even a thin aluminium filter can completely remove it. Hence, it could lead to substantial aDNA degradation in the sub-surface of a specimen, but not in depth, because the specimen itself would act as a filter and stop these low energy X-rays.

Supplementary Note 3 -X-ray device for luggage inspection in airports
The following information results from a correspondence with Dr. Andreas Frank (Director Technology Low Energy Systems Hardware), Dr. Arno Folkerts and Dr. Christian Rauth (Radiation Safety Officer) from Smiths Heimann GmbH, Wiesbaden, Germany. Hyperlinks in the following text will lead the reader to the website of the company as well as further technical details on the device (pdf to download).
Smiths Detection is a company that produces X-ray scanning devices for luggage inspection in airports. HiScan 6040 a-TiX is the current model in use by the security staff in most German airports (including Leipzig and Frankfurt) and many international airports as well.
The scanning device is equipped with four X-ray sources, each emitting a thin fan-beam which penetrates the luggage during security inspection. The luggage progresses through the tunnel of the device at a constant speed of 20 cm/s. The total exposure time is about 20 ms (5 ms for each of the 4 X-ray sources). The machines are configured in a way that the dose is always delivered in the same way, and this cannot be changed by the security staff at the airport. This results in a total delivered dose of 10 to 12 µGy per inspection process.
Smith Detection uses several dosimetry protocols for cross-validating their results, and although not directly and strictly obtained the same way as our values on conventional and synchrotron CT were acquired, the values provided above remain comparable, since being in the range of 10 µGy.
Occasionally, the security staff may repeat the procedure if the image quality is not satisfying enough, although this is not standard procedure in most airports. Further inspections would rather be conducted (e.g., opening and searching the luggage), possibly involving other technologies (e.g., trace detection).
Some airports in Europe still use older models of devices for luggage inspection (e.g., HiScan 6040i or HiScan 6046si) which only have one X-ray source, and for which the total delivered dose per inspection is 1.5 to 2.5 µGy (for 5 ms exposure time).