Search

Imaging spectroscopy has been developed in the last two decades in the visible and infrared spectral range for detecting pigments and binders on paintings. The near-infrared (NIR) region has been proved effective for the discrimination of lipids and proteinaceous binders. More recently, the mid-infrared (MIR) range has also been tested on paintings. Reflection imaging prototypes already developed could be further optimized for cultural heritage analysis, for example by: enhancing the instrument configuration and performance; adopting compressive strategies to increase data processing speeds; using data validation to confirm that the processed image reflects the composition of a painted surface; and lowering price to enable more cost-effective analysis of large surface areas. Here, we demonstrate a novel hyperspectral Fourier transform spectrometer (HS FTS), which enables an imaging strategy that provides a significant improvement in acquisition rate compared to other state-of-the-art techniques. We demonstrate hyperspectral imaging across the 1400–700 cm−1 region in reflection mode with test samples and the painting ‘Uplands in Lorne’ (Acc. No.: GLAHA43427) by D.Y. Cameron (1865–1945). A post-processing analysis of the resulting hyperspectral images, after validation of reference samples by conventional Fourier transform infrared spectroscopy, shows the potential of the method for efficient non-destructive classification of different materials found on painted cultural heritage. This research demonstrates that the HS FTS is a convenient and compact tool for non-invasive analysis of painted cultural heritage objects at spatio-spectral acquisition rates potentially higher than current FTS imaging techniques. Ultimately, when combined with fast graphics processing unit-based reconstruction, the HS FTS may enable fast, large area imaging. Graphical abstract

The discovery1,2and the pioneering applications3of CRISPR-Cas effector complexes have provided powerful gene-editing tools. The effector complexes are guided to the targeted genomic locus by the complementarity of their CRISPR RNA (crRNA)4,5. Recognition of double-stranded DNA targets proceeds via DNA unwinding and base-pairing between crRNA and the DNA target strand resulting in the formation of an R-loop structure5,6. Full R-loop formation is the prerequisite for the subsequent DNA cleavage. While the CRISPR-Cas technology is easy to use, efficient and highly versatile, therapeutic applications are hampered by the off-target effects due to the recognition of unintended sequences with multiple mismatches7. This process is still poorly understood on a mechanistic level8,9. Particularly, the lack of insight into the energetics and dynamics of the R-loop formation hinders a direct modelling of the R-loop formation for off-target prediction.Here we set up ultrafast DNA unwinding experiments based on plasmonic DNA nanorotors to follow the R-loop formation by the Cascade effector complex in real time, close to base pair resolution. We directly resolve a weak global downhill bias of the energy landscape of the forming R-loop followed by a steep uphill bias for the final base pairs. We furthermore show a modulation of the landscape by base flips and mismatches. These data provide that Cascade-mediated R-loop formation occurs on short time scales in single base pair steps of sub-millisecond duration, but on longer time scales in six–base pair intermediate steps in agreement with the structural periodicity of the crRNA-DNA hybrid. We expect that the knowledge about the energy landscapes of R-loop formation of CRISPR-Cas effector complexes will pave the way for a detailed understanding and prediction of off-target recognition10.

CRISPR-Cas effector complexes enable the defense against foreign nucleic acids and have recently been exploited as molecular tools for precise genome editing at a target locus. To bind and cleave their target, the CRISPR-Cas effectors first have to interrogate the entire genome for the presence of a matching sequence. Matching is achieved by base-pairing between the crRNA of the complexes and the DNA target strand such that an R-loop is formed. R-loop formation starts at a specific PAM motif and progresses reversibly in single base-pair steps until mismatches stop further progression or until the full target is recognized and destroyed. The reversible nature of this process entails that even a fully matching target should only become recognized with a low probability per target encounter. The details of this process, which directly affect the effectiveness of the target search, remain unresolved.Here we dissect the target search process of the Type I CRISPR-Cas complex Cascade by simultaneously monitoring DNA binding and R-loop formation by the complex. We directly quantify the low target recognition probabilities and show that they increase with increasing negative supercoiling. Furthermore, we demonstrate that Cascade uses a combination of three-dimensional and limited one-dimensional diffusion along the DNA contour for its target search. The latter allows for rapidly scanning the PAM sequences in a given region and, importantly, significantly increasing the overall efficiency of the target search by repeatedly revisiting the sites. Overall we show that target search and target recognition are tightly linked and that DNA supercoiling and limited 1D diffusion need to be considered when understanding target recognition and target search by CRISPR-Cas enzymes and engineering more efficient and precise variants.

Spectrally-resolved imaging provides a spectrum for each pixel of an image that, in the mid-infrared, can enable its chemical composition to be mapped by exploiting the correlation between spectroscopic features and specific molecular groups. The compatibility of Fourier-transform interferometry with full-field imaging makes it the spectroscopic method of choice, but Nyquist-limited fringe sampling restricts the increments of the interferometer arm length to no more than a few microns, making the acquisition time-consuming. Here, we demonstrate a compressive hyperspectral imaging strategy that combines non-uniform sampling and a smoothness-promoting prior to acquire data at 15% of the Nyquist rate, providing a significant acquisition-rate improvement over state-of-the-art techniques. By illuminating test objects with a sequence of suitably designed light spectra, we demonstrate compressive hyperspectral imaging across the 700–1400 cm-1 region in transmission mode. A post-processing analysis of the resulting hyperspectral images shows the potential of the method for efficient non-destructive classification of different materials on painted cultural heritage.

Full-field imaging for non-invasive spectral analysis of pigments in painted cultural heritage is demonstrated from 1400–700cm-1. Acquisition without compressive sensing takes 4 minutes, with a further 2 minutes needed for spectral reconstruction and clustering.

We demonstrate compressive spectroscopy at 3.3 µm by projecting a sparse set of spectra shaped by a digital micromirror device. Acquisition at 280 Hz is achieved. Reconstructed spectroscopy shows good agreement with ground truth measurements.