Acinar micromechanics in health and lung injury: what we have learned from quantitative morphology

Abstract

Within the pulmonary acini ventilation and blood perfusion are brought together on a huge surface area separated by a very thin blood-gas barrier of tissue components to allow efficient gas exchange. During ventilation pulmonary acini are cyclically subjected to deformations which become manifest in changes of the dimensions of both alveolar and ductal airspaces as well as the interalveolar septa, composed of a dense capillary network and the delicate tissue layer forming the blood-gas barrier. These ventilation-related changes are referred to as micromechanics. In lung diseases, abnormalities in acinar micromechanics can be linked with injurious stresses and strains acting on the blood-gas barrier. The mechanisms by which interalveolar septa and the blood-gas barrier adapt to an increase in alveolar volume have been suggested to include unfolding, stretching, or changes in shape other than stretching and unfolding. Folding results in the formation of pleats in which alveolar epithelium is not exposed to air and parts of the blood-gas barrier are folded on each other. The opening of a collapsed alveolus (recruitment) can be considered as an extreme variant of septal wall unfolding. Alveolar recruitment can be detected with imaging techniques which achieve light microscopic resolution. Unfolding of pleats and stretching of the blood-gas barrier, however, require electron microscopic resolution to identify the basement membrane. While stretching results in an increase of the area of the basement membrane, unfolding of pleats and shape changes do not. Real time visualization of these processes, however, is currently not possible. In this review we provide an overview of septal wall micromechanics with focus on unfolding/folding as well as stretching. At the same time we provide a state-of-the-art design-based stereology methodology to quantify microarchitecture of alveoli and interalveolar septa based on different imaging techniques and design-based stereology.

Keywords: Alveolar recruitment; electron microscopy; imaging; interalveolar septa; micromechanics; pulmonary acinus; stereology.

FIGURE 1

Ultrastructure of the blood-gas barrier: Transmission electron micrograph of a rat lung fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 5 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). The capillaries (Cap) are open and free from blood cells. In (A) an example of the thin part of the blood-gas barrier can be seen. In this area the squamous extension of an alveolar epithelial type 1 cell (ACE1), the basement membrane (asterisk) and the endothelial cell (Endo) form the blood-gas barrier. In (B) an example of the thick part of the blood-gas barrier is illustrated. The interstitium between the AEC1 and the endothelium is widened and both AEC1 and the endothelial cell have a basement membrane (asterisk) of its own. Aside from collagen fibrils (Col), cell extensions of interstitial cells (IC), e.g. fibroblasts can be identified.

FIGURE 2

Three-dimensional model of an alveolar epithelial type 2 (AE2) cell: The mouse lung was fixed in situ by vascular perfusion via the right ventricle at an airway opening pressure of 2 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Ruhl et al., 2019). Tissue was processed for serial block face scanning electron microscopy (SBF-SEM) as described in Buchacker et al., 2019. The EM image stack was used for segmentation of an AE2 cell, located within a junction of three interalveolar septa. On a three-dimensional representation of the AE2 cell surface, apical portions of the plasma membrane are shown in light green and basolateral portions that are attached to the epithelial basement membrane are shown in dark green. Lamellar bodies located below the plasma membrane are shown in yellow. The AE2 cell is, further, shown inserted into a two-dimensional electron microscopic image of the surrounding environment. The AE2 cell shows multipolarity with four apical domains being in contact with three alveolar airspaces (Alv). The arrow points into a pleat filled with protein containing fluid (hypophase). The bottom of the pleat is formed in part by one of the apical domains of the AE2 cell.

FIGURE 3

The junction of inter-alveolar septa: A healthy rat lung was fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 10 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). In the middle of the septal junction (arrows), an alveolar epithelial type II cell (AEC2) is present. While most of the interalveolar septa contain a single layer of the alveolar capillary network, there appear to be two layers at this junction surrounding the AEC2, most likely due to pleating that causes piling up of the interalveolar septa.

FIGURE 4

Stereological test-systems: On two-dimensional (2D) sections, three-dimensional structures (3D) lose one dimension. Accordingly, a volume appears as an area (2D), a surface area as a line (1D) and a length as a point (0D). The numerical quantity of a structure is a dimensionless (0D) parameter in three-dimensional space. Since a negative dimension is not possible this parameter is not represented on two-dimensional images. Hence, quantity of any structure cannot be determined from single sections based on the principles of stochastic geometry (or any other method)—an unbiased test-volume (3D) generated by a disector is required.

FIGURE 5

Unbiased test-system for measurements of chord length of acinar air spaces. A healthy rat lung was fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 5 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). The parameter “chord length” is also known as the linear intercept length. It is based on simple, linear measurements of the dimension of the acinar airspaces from one border to the next. The left side of the randomly sampled image contains four straight line-segments, extended to the right by a dashed line, the so-called guard line. The line segments on the left serve to sample the starting point of the measurements. Each time the line segment intersects an interalveolar septum a measurement is performed from the intersection to the next surface of an interalveolar septum. The direction of the measurement follows the run of the test line and if needed also the dashed guard line. In order to locate the points of measurements exactly the top border of the line segment is used. The arrows label the measurements in this example. Some measurements are performed within an alveolus, others, however, travers via the alveolar opening through the alveolar duct airspace to the other side so that these measurements encompass both alveolar and alveolar duct airspaces.

FIGURE 6

Distribution of chord lengths of acinar airspaces: Healthy and injured (ALI) rat lungs were fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure (Pao) of either 1 cmH₂O on expiration or 10 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). The distributions of chord lengths of acinar airspaces are illustrated as histograms and probability based on the Kernel probability distribution function. In both healthy and injured lungs fixation at higher Pao on expiration results in a right shift of the measurements. At Pao = 1 cmH₂O, hardly any differences can be identified in the histograms between healthy (A) and injured lungs (B). Accordingly, the Kernel probability distribution function shows hardly any differences (C). At Pao = 10 cmH₂O, the histograms suggest a right shift of the peak in the injured lung (E) compared to the healthy lung (D). The Kernel probability distribution function supports this right shift and indicates a second peak at larger chord length in the range of 150 µm (F). Note, the investigation was performed at a very early time point of bleomycin-induced acute lung injury development at which lung mechanical measurements were scarcely affected. In each group, 500 measurements were performed on randomized fields of view from two lungs.

FIGURE 7

Differentiation of acinar airspaces by point counting: Healthy rat lungs were fixed in vivo by vascular perfusion via the vena cava caudalis at airway opening pressures of either 1, 5, and 10 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). In order to determine the volume fractions of tissue, alveolar or alveolar duct airspaces within the lung, test points were superimposed on randomized fields of view. Lungs were treated by immersion in 4% OsO₄ before dehydration and embedded in glycol methacrylate to avoid shrinkage/tissue deformation after fixation. The probability of a test point hitting the profile of a structure of interest is directly proportional to the volume fraction of this structure of interest within the reference space. The ratio of test points hitting the structure of interest and the reference space provides the volume fraction of the structure of interest. In the examples, the probability of a test point placed on the randomly sampled fields of view depends on the volume fraction of tissue, alveolar airspace or alveolar duct airspace within the lung. The multiplication of the volume fractions with the lung volume will result in the absolute volumes of tissue, alveolar airspace or alveolar duct airspace per lung. With the goal to separate alveolar duct and alveolar airspace, the entrances into the alveoli were closed by drawing a straight line between the edges of the interalveolar septa. Points hitting alveolar duct airspace (green), alveolar airspace (red) and tissue (yellow) were labelled in the examples. At low Pao, the alveolar ducts were small and the inter-alveolar septa appeared to be at rest with a curvy or in part crumpled surface. At larger Pao, the alveolar ducts widened considerably, the inter-alveolar septa straightened (and appeared to be under tension) and the alveoli became larger. The number of test points hitting alveolar as well as alveolar duct airspaces increased on the expense of the tissue. If the images were representative for the whole organ, one would assume that the volume fractions of tissue deceases with inflation pressure. In order to be representative, however, it is advised to count 100—200 hits on a structure of interest from at least 60 randomized fields of view sampled from at least four randomized sections per organ.

FIGURE 8

Pleating of the blood-gas barrier: A healthy rat lung was fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 5 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). The capillary network is open and nearly free of blood cells. The empty arrowhead points to the entrance of a pleat that is filled with a protein containing fluid (asterisk). Pleats are generally filled with a small quantity of proteinaceous fluid but occasionally locations of direct epithelial-epithelial contact are observed. The pleat is limited to the blood-gas barrier, which invaginates into a capillary (Cap). Two filled arrowheads locate the blood-gas barrier. The image on the right shows the run of the basement membrane (black dashed line), shared by the endothelial cell and the alveolar epithelial type 1 cell, within the pleat.

FIGURE 9

Three-dimensional model of a pleat: The mouse lung was fixed in situ by vascular perfusion via the right ventricle at an airway opening pressure of 2 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Ruhl et al., 2019) and processed for serial block-face scanning electron microscopy (SBF-SEM) (Buchacker et a. 2019). The EM image stack was used to segment the shared basement membrane of the endothelial and alveolar epithelial type I cell (ebl, magenta) within a pleat. The arrow points at the slit-like entrance to the pleat, which is created by the blood-gas barrier and invaginates sickle-shaped into the capillary (Cap). The model of the pleat is put into the context of the EM stack by two-dimensional images.

FIGURE 10

Pleat involving more components of the interalveolar septa: A healthy rat lung was fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 5 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). The alveolar airspaces (Alv) and the alveolar capillary network (Cap) is open but contains red blood cells. The capillaries are lined by endothelial cells (Endo). The filled arrowhead points at the entrance to a pleat, the black dashed line marks the run of the epithelial basement membrane into the pleat. The pleat is partly bordered by the apical plasma membrane of an alveolar epithelial type 2 cell (AEC2) with its characteristic organelle, the lamellar body (LB). Moreover, the pleat contains some fluid and intraalveolar surfactant, represented by tubular myelin (TM). Underneath the AEC2, interstitial tissue is located, e.g. collagen fibrils (Col) and fibrobasts (FB) are visible. Note that two dark lines at left of left image are an artifact due to folding of the ultrathin section, not part of the tissue structure. The empty arrowheads point at pleats formed exclusively by the blood-gas barrier.

FIGURE 11

Lung injury and microatelectases: In (A) an image of the lung of a surfactant protein B knock out mouse is shown, fixed in situ by vascular perfusion via the right ventricle at an airway opening pressure of 10 cmH₂O on expiration two recruitment maneuvers (3 s pause at 30 cmH₂O) (Ruhl et al., 2019). The alveolar ducts are enlarged. Alveolar airspaces are rare and appear to be shallow. Instead, microatelectases (arrows) can be identified as seemingly thickened interalveolar septa characterized by a conglomeration of capillaries. In (B) similar findings can be observed in a rat lung 1 day after instillation of bleomycin to induce lung injury. The lung was fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 10 cmH₂O on inspiration, coming from 1 cmH₂O after two recruitment maneuvers (3 s pause at 30 cmH₂O).

FIGURE 12

Counting open alveoli. Top: A mouse lung was fixed in situ by vascular perfusion via the right ventricle at an airway opening pressure of 10 cmH₂O on expiration after two recruitment maneuvers (3 s pause at 30 cmH₂O). Tissue was sampled, treated with OsO₄ and embedded in epoxy resin (Ruhl et al., 2019). A tissue block was imaged by micro computed tomography (Nanotom M, Waygate Technology, Wunsdorf, Germany) at a voxel size 1 µm. A pair of images from the stack showing the same region is given. The distance from the top of the left image to the top of the right image is 4 µm. An unbiased counting frame of the area A_Frame is superimposed on the images. The counting frame area and the distance between the counting frames generates a test volume. In this test volume, the number of alveolar openings are determined as follows: alveoli open to the alveolar duct on one image but not on the other (arrow) are counted, provided that the opening to the duct is located within the counting frame and does not touch the red line (= forbidden line). Bottom: A healthy rat lung was fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 10 cmH₂O after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). After randomized sampling and embedding, serial sections of a thickness of 1.5 µm were cut. The 1^st and the 4^th section of a consecutive row of sections was collected and randomized pairs of fields of view showing corresponding regions from these 2 sections were images. An example is given here. The arrow points at a counting event which is defined above.

FIGURE 13

Micromechanics of the blood-gas barrier. Schematic of an interalveolar septum at end-expiration (bottom) and end-inspiration (top). The end-expiratory drawing is based on an electron microscopic image and shows in the upper blood-gas barrier (BGB) a pleat filled with fluid (light blue). The surface area of the involved alveolar epithelial cell is hidden in the pleat (yellow) and not exposed to air. After inspiration, the pleat has been opened and its surface area, although still covered by fluid is now exposed to the alveolar lumen. As a result, there is an increase in surface area and adaptation to changing alveolar size without stretching of the alveolar epithelial cell or the epithelial basement membrane (green). The BGB at the bottom of the septum however, does not have a pleat. The BGB is therefore stretched during inspiration which results in an increase in the surface area of both air-exposed apical membranes of alveolar epithelial cells and epithelial basement membrane. The BGB is thinned due to stretching. Capillaries (Cap) and the BGB are also subject to shape changes to adapt to an increase in alveolar volume during inspiration which do not result in an increase in the surface area of the basement membrane or apical plasma membrane of epithelial cells. This schematic is based on classical transmission electron microscopic images which are not able to visualize the complete liquid lining layer. Please note, that the fixation process with glutaraldehyde is based on cross-linking of proteins. Hence, only those parts of the liquid lining layer which contain proteins can be visualized and are given in this schematic.

FIGURE 14

Design-based stereology of the blood-gas barrier: A healthy rat lung was fixed in vivo by vascular perfusion via the vena cava caudalis at an airway opening pressure of 10 cmH₂O after two recruitment maneuvers (3 s pause at 30 cmH₂O) (Knudsen et al., 2018). In the center, a pleat filled with protein-containing fluid and bordered by the blood-gas barriers can be seen. On the left, the epithelial basement membrane (ebl) is delineated (dashed red line). While at the top right corner, the alveolar epithelium is exposed to air (Alv), the pleat hides epithelial surface area which might be recruitable on inspiration. The pleat is filled by a grayish material representing preserved parts of the liquid-lining layer. In order to quantify the surface area covered by air or hidden within pleats, test lines can be projected on randomized electron microscopic images for intersection counting as shown on the right. Line segments of a certain length are superimposed on the image. The probability of these line segments to intersect air covered or hidden alveolar epithelium is proportional to the collective length (L) of the line segments but also to the surface density of, e.g. air-covered or hidden alveolar epithelium. Hence, intersection (I) counting can be applied to determine the surface density (Sv) of the desired structures, given by the equation Sv = 2*I/L. Intersections are indicated as follows: blue arrow: air-covered alveolar epithelial surface area; yellow arrow: hidden alveolar epithelial surface area. The same line segments can be used to determine the surface area of the epithelial basement membrane, a parameter eligible to quantify stretch of the blood-gas barrier. The green arrows point at intersections of the line segments with the epithelial basement membrane.