Skip to main content
Dryad

Dataset from: Lung evolution in vertebrates and the water-to-land transition

Cite this dataset

Cupello, Camila et al. (2022). Dataset from: Lung evolution in vertebrates and the water-to-land transition [Dataset]. Dryad. https://doi.org/10.5061/dryad.vdncjsxx1

Abstract

A crucial evolutionary change in vertebrate history was the Palaeozoic (Devonian ~400 million years ago) water-to-land transition, allowed by key morphological and physiological modifications including the acquisition of lungs. Nonetheless, the origin and early evolution of vertebrate lungs remain highly controversial, particularly whether the ancestral state was paired or unpaired. Due to the rarity of fossil soft tissue preservation, lung evolution can only be traced based on the extant phylogenetic bracket. Here we investigate, for the first time, lung morphology in extensive developmental series of key living lunged osteichthyans using synchrotron X-ray microtomography and histology. Our results shed light on the primitive state of vertebrate lungs as unpaired, evolving to be truly paired in the lineage towards the tetrapods. The water-to-land transition confronted profound physiological challenges and paired lungs were decisive for increasing the surface area and the pulmonary compliance and volume, especially during the air-breathing on land. 

Methods

Specimens of P. senegalus, L. paradoxa and S. salamandra were imaged at the PSICHÉ beamline of the SOLEIL Synchrotron (Saint-Aubin, France) while N. forsteri specimens were scanned at SPring-8 Synchrotron. The specimens were scanned isolated in a plastic tube filled with Phosphate-buffered saline (PBS) for P. senegalus and N. forsteri, ethanol for L. paradoxa and formaldehyde for S. salamandra. They were immobilized in a vertical position using gauze pads, and/or sank inside the tip of a plastic pipette in the case of tiny individuals, in order to benefit as much as possible from the available field of view and thus achieve the highest possible resolution.

At SOLEIL Synchrotron, imaging was performed using a monochromatic beam with an energy of 25 keV. A series of acquisitions with vertical movement of the sample was recorded to extend vertically the field of view and image the entire (or most of the) individual. Two distinct setups were used to accommodate the different sizes of the individuals (size variations occurring both between developmental stages and taxa). (1) Small individuals were scanned using a field of view of 2.6 x 2.6 mm2 (5x magnification) resulting in a projected pixel size of 1.3 μm, and a propagation distance of 148 mm. (2) Larger individuals were scanned using a field of view of ~12.6 x 3.3 mm2 (1x magnification) resulting in a projected pixel size of 6.17 μm, and a propagation distance of 500 mm. For individuals slightly wider than these field of views, the latter were extended horizontally by positioning the rotation axis off-centre and acquiring data over a 360° rotation of the sample. The volumes were reconstructed from the (vertically) combined radiographs using PyHST2 software (Mirone et al., 2014), with a Paganin phase retrieval algorithm (Paganin et al., 2002). The huge resulting volumes (from 70 Gb to 1.2 Tb) were reduced (crop, rescale 8-bit, binning) to facilitate 3D data processing.

Specimens of Neoceratodus forsteri (from 13.5 to 70 mm TL) were imaged at the SPring-8 Synchrotron, beamline 20B2. For specimens from 13.5 mm TL to 30 mm TL, beam energy of 15 keV was used with a double bounce Si (111) monochromator. Data were obtained at three different resolutions, and correspondingly used three combinations of two lenses and fluorescent material, as follows, 2.75 µm/voxel; 1st-stage lens: "beam monitor 2" f35 mm; 2nd-stage lens: Nikon 85 mm lens; GADOX thickness: 15 µm 4.47 µm/voxel; 1st-stage lens: "beam monitor 2" f35 mm; 2nd-stage lens: Nikon 50 mm lens; GADOX thickness: 15 µm 12.56 µm/voxel; 1st-stage lens: "beam monitor 5" f200 mm; 2nd-stage lens: Nikon 105 mm lens; GADOX thickness: 25 µm.

Datasets were acquired at propagation distances of 2.75 µm/voxel, 4.47 µm/voxel: 600 mm; 12.56 µm/voxel: 3 m and three different exposure times of 70 ms, 150 ms, and 200 ms per projection. Field of view were: pixel size x 2048 (2.75 x 2048 = 5632 µm; 4.47 x 2048 = 9154.56 µm; 12.56 x 2048 = 25722.88 µm) A total of 1800 projections were recorded per scan as the sample was rotated through 180°. A high-resolution computerized axial tomography scanning (CAT scan) was performed for the adult specimen of Neoceratodus (KPM-NI 11384) of 200 mm TL at the National Museum of Nature and Science (Tokyo, Japan) using the following scanning parameters: effective energy 189 kV, current 200 mA, voxel size 9.765 μm and 1000 views (slice width 0.1 mm).

Segmentation and 3D rendering were performed using the software MIMICS Innovation Suite 20.0 (Materialise) at the Laboratório de Ictiologia Tempo e Espaço of the Universidade do Estado do Rio de Janeiro.

Usage notes

This dataset includes sequences of images corresponding to the tomographic volumes segmented in the related article, shown in figs 1B, 1D, 1E, 2B, 2D, 4B, 4D, 4E, 4G, 4H, 4J, 4K, 5B, 5D, 6A, 6B, 6E, 6F, 6G, 6H, 6I, 6J of the article. The sequences of images are distinguished by their file name that indicates to which figure (and consequently tomographic volume) they correspond. 

Separate sequences of images can then be imported into any segmentation software or the freeware ImageJ/Fiji, for visualisation and processing.