Computational protocol: XTEND: Extending the depth of field in cryo soft X-ray tomography

Similar protocols

Protocol publication

[…] We proceeded analyzing XTEND performance in two experimental datasets acquired at the Mistral microscope. The first reconstruction was obtained of a Scenedesmus cell using ZP40 (Scenedesmus is a phototrophic microorganism isolated from Ebro delta (Spain) microbial mats in 2009), while the second case is a reconstruction of a HT-29 cell using ZP25. Note that the total photon dose used for experimental XTEND tilt series was below the damage limit at our achievable resolution. However, it was not possible to acquire an extra focal series at 0-shift to compare it with XTEND at same SNR levels without exceeding the damage limit as in the simulated tomograms.Previous analysis of a Sc. cell shaped a z-range of interest ~6 μm, while the DOF measured for ZP40 was 3.3 μm. We thus determined that, for a total of 3 focal series, a z-step ~4 μm should be used for correct application of XTEND (). Collection geometry was single tilt axis in the range [−70°, 68°] in 1° steps, with 1 s exposure time and 13 nm pixel size. Note that focal series projections were obtained by moving the FZP along the optical axis at each tilt angle and, although showed small shift misalignment around 1–2 pixels in x and y axes due to the ZP motion, it was corrected by comparison to the reference 0-shift tilt series using EFTEM-TomoJ. The average of the focal series projection at each tilt angle was aligned by IMOD. The in-focus z-range value of 6 μm and a SNR estimation for the 0-shift and XTEND projections, near 20 and 100, respectively, were used as input for deconvolution. compares slices of the three reconstruction procedures, 0-shift, 0-shift-dec and XTEND, with objective quality measures as for the simulations. Experimental data sections of the reconstructed XTEND volume already show a significant improvement, especially for internal features of cellular organelles, as those ones pointed by white arrows (). Density profiles extracted from slices at 2.4 μm are plotted in . They clearly show the highest contrast enhancement by XTEND compared to standard 0-shift and 0-shift-dec. The AQI measurement shows a clear quality increase for all planes, considering the comparison conditions we can achieve experimentally regarding dose limitation between 0-shift and XTEND cases (). In this case, 0-shift-dec did not show the greatest quality index at central slice, probably due to a slight misplacement of the central tilt axis close to −0.8 μm instead of 0 μm.The second experimental tilt series, shown in , is a sample of HT-29 cells. It was acquired with the high resolution zone plate ZP25. In this case, the estimated z-range of interest was ~5 μm, equivalent to its cytoplasm and combined with the experimental DOF of 1.6 μm, resulted in 3 focal series and z-step of ~3.3 μm. The tilt geometry in this experiment was in the range [−65°, 65°] in 1° steps, with variation of exposure time between 2 and 3 s and 10 nm pixel size. A clear improvement in the mitochondria cristae is noticeable in the XTEND reconstruction. Although the use of ZP25 is characterized by a higher degree of coherence than the one for ZP40, the results in demonstrate a similar behaviour when XTEND is applied, probably because both cases are partially coherent. […]

Pipeline specifications

Software tools TomoJ, IMOD
Application cryo-electron tomography