Ewald R. Weibel

Stereological Principles for Morphometry in Electron Microscopic Cytology

Publisher Summary This chapter focuses on the stereological principles for morphometry in electron microscopic cytology. Morphometry of the cell serves the purpose of furnishing quantitative information on cellular fine structure with the aim of allowing quantitative correlation of biochemical or physiological data with morphological data obtained on structurally intact cells. the morphometric approach to cytology is not a purpose in itself, although admirable dimensional equilibria can be revealed which satisfy the esthetic needs. It is a means that serves the aim of structure-function correlation and derives its justification from the recognition that all orderly function must have an organized structural basis of a size that is adequate but not excessive. The chapter emphasizes on the possibilities of applying morphometric methods to correlative cell biology.

An official research policy statement of the American Thoracic Society/European Respiratory Society: standards for quantitative assessment of lung structure.

1.1. The Challenges To understand normal lung function, the processes of growth and development, and the mechanisms and effects of diseases, we need information about the 3D structure of the lung. Quantification of organ structure is based upon 3D physical attributes of tissues, cells, organelles, alveoli, airways, and blood vessels. When structures of interest are inaccessible or too small to be seen macroscopically, we rely on physical or optical sections through a few representative samples taken from the large heterogeneous organ. The resulting 2D images confer incomplete information about the 3D structure, and may not accurately represent true 3D properties, leading to possible misinterpretation when measurements are made on 2D sections. Because structural quantification is often considered the “gold standard” in evaluating experimental intervention, disease severity, and treatment response, it is imperative that these quantitative methods are (1) accurate to allow meaningful interpretation of results, (2) efficient to yield adequate precision with reasonable effort, (3) of adequate statistical power to encompass inherent variability, and (4) adherent to uniform standards to facilitate comparisons among experimental groups and across different studies. The lung poses special challenges, some of which are outlined below and discussed in later sections: (a) Heterogeneity of lung structure requires standardized preparation methods. The inflated lung consists of mostly air; only 10 to 15% of its volume consists of tissue (cells, fibers, and matrix) and blood. In vivo lung volume and relative volumes of air, tissue, and blood fluctuate widely, while gravitational and nongravitational gradients cause spatial heterogeneity in structure and function. Failure to standardize physiological variables or minimize tissue distortion introduces uncertainties or errors into subsequent measurements, to the point of their being meaningless (1). Careful selection of fixation and preparation methods that minimize shrinkage obviates this problem (Section 3). (b) Selected microscopic sections should provide a fair sample of the whole organ. The practice of picking specific samples or sections often fails to account for regional heterogeneity, leading to biased conclusions with respect to the whole organ. Deliberately choosing sections that contain a particular compartment (e.g., profiles of alveolar type 2 epithelial cells) overestimates their abundance within the whole lung. Using a sampling scheme that covers all regions with equal probability alleviates this problem (Section 4). (c) Measurements made on microscopic sections must be related to the whole organ or an appropriate reference volume. Studies continue to appear that report only relative measurements (i.e., volume and surface densities or ratios) without knowledge of the lung volume. These ratios are dependent on lung inflation, and must be multiplied by absolute lung volume to obtain accurate total quantities of the structures of interest. Uncertainties regarding lung volume can bias data interpretation. For example, enlarged mean airspace size need not signify emphysema or alveolar hypoplasia; the finding could also be caused by overinflation. Careful measurement of the lung volume eliminates this error (Section 5). (d) Lung structures are irregular and their geometry easily altered by pathology and intervention. Measurements on 2D images that rely on assumed geometry may misrepresent the 3D structure. Examples include estimating alveolar size from cross-sectional areas of alveolar profiles, and reporting alveolar surface area by the length of alveolar profile boundary. These measures can severely misrepresent the 3D structure of interest. Airspace size is often inferred from the mean linear intercept (Lm), which in fact measures airspace volume-to-surface ratio and can be converted to diameter or volume only by assuming a shape factor. Airspace distortion, or selective distortion of alveolar ducts but not alveolar sacs, can invalidate shape assumptions (Section 6). (e) The number of lung cells cannot be estimated by counting their profiles on random histologic sections because larger cells have a greater probability of being sampled. For example, if experimental intervention causes selective cell hypertrophy, the increased probability of counting cell profiles will lead to wrong conclusions. Again, using stereologic methods that are free of geometric assumptions eliminates this error (Sections 6–7). (f) In contrast to acinar structures that exhibit nearly random orientation (isotropy) and homogeneous distribution, conducting airways and blood vessels exhibit preferred directions (anisotropy) and inhomogeneous distribution, which alter their sampling probability on random sections. Specific sampling procedures that account for their nonrandom nature should be employed to ensure unbiased representation on 2D sections (Section 8). (g) Assessment of endobronchial or lung biopsy specimens is limited by their nonrandom nature and a lack of external reference parameter. Endobronchial biopsy specimens are also anisotropic with distinct luminal and basal sides and with respect to airway generations. To minimize potential errors in quantification, specimens should be processed with their orientation randomized and analyzed with respect to an internal reference parameter (Section 9). (h) The new imaging techniques CT and MRI offer the possibility of obtaining high-fidelity images of lung structure in vivo that can be used for quantitative assessment of structural changes. Since their images are sections of the organ, stereology can ensure accurate measurements (Section 10). Definition of terms (section of text where term is defined) Accuracy (Sec. 1.2); ALP-sector (Sec. 2.1, item a); Anisotropy (Sec. 1.1, item f); Apparent diffusion coefficient (ADC) (Sec. 10.4.1); Arithmetic mean thickness of air-blood barrier (Sec. 6.7); Bias (Sec. 1.2); Buffons needle (Sec. 1.3); Cavalieri Principle/Method (Sec. 1.3); Coarse nonparenchyma (Sec. 6.2); Coarse parenchyma (Sec. 6.2); Computer-aided stereology systems (Sec. 2.2, item c); Connectivity of airway branching systems (Sec. 8.1); Delesse principle (Sec. 1.3); “Design-based” (Sec. 1.2); Dichotomous branching of airways (Sec. 8.1, Fig. 9A); Disector principle: physical, optical (Sec. 2.1, items d and e); “Do more less well” (Sec. 2.2, item c); Sec. 4.4; Efficiency (Sec. 4.4); Euler characteristic (Sec. 6.4); Fine nonparenchyma (Sec. 6.2; Figure 5); Fine parenchyma (Sec. 6.2; Equation 12); Fractal tree (Sec. 8.1); Fractionator sampling (Sec. 4.2.5; Figure 4); Global estimators (Sec. 2.1); “Gold standard” in fixation (Sec. 3.1); Harmonic mean thickness of air–blood barrier (Sec. 6.7); Horsfield ordering system (Sec. 8.1; Figure 9b); Isector (isotropic orientation) (Figure 4); Isotropic uniform random (IUR) sampling (Sec. 4.2.3); Isotropy (Sec. 1.1, item f); Local estimators (Sec. 2.1, item e); Mean chord length or mean linear intercept (Sec. 6.6); Monopodial airway branching (Sec. 8.1); Morphometry (Sec. 2.1); Multistage stratified morphometric analysis (Sec. 6.1); Multistage stratified sampling (Sec. 4.2.6); Nucleator (Sec. 2.1, item e); Number-weighted mean particle volume (Sec. 2.1, items e and f); Orientator (Sec. 4.2.3); Point-sampled intercept (Sec. 2.1, item e); Precision (Sec. 1.2); Reference space (Sec. 5); Reference lung volume (Sec. 5.1); “Reference trap” (Sec. 5); Relative deposition index (RDI) (Sec. 7.2); Relative labeling index (RLI) (Sec. 7.2); Rotator (Sec. 2.1, item e); Sampling (Sec. 2.1, Sec. 4); Sampling fraction (Sec. 6.4; Figure 4); Sampling procedures (Sec. 4.2); Sampling rules (Sec. 4.1); “Silver standards” in fixation technique (Sec. 3.1; Sec. 3.3); Stereology (Sec. 2.1); Strahler ordering system (Sec. 8.1; Figure 9b); Stratified uniform random (StUR) sampling (Sec. 4.2.2); Surface density (Sec 2.1, item b; Sec. 6.3); Systematic uniform random sampling (SURS) (Sec. 4.2.1); Test probes, test systems (Sec. 2.1, item a; Sec. 6.9; Figure 6); Uniform random sections (Sec. 4.2.1; Sec. 4.2.2; Sec. 4.2.3); Vertical sections (Sec. 4.2.4; Figure 3); Volume density (Sec. 2.1, item b; Sec. 6.2); Volume-weighted mean particle volume (Sec. 2.1, items e and f). Open in a separate window Figure 3. Vertical sections. (A) An arbitrary horizontal reference plane, such as a cutting board, is considered fixed and the vertical section is perpendicular to this horizontal plane. Airways selected by microdissection can be sampled by this vertical section scheme, by bisecting the airway longitudinally and laying it flat with the luminal surface up. In this orientation, the arrow that runs from base to apex of the epithelium indicates the direction of the vertical axis, V. (B) Bisected airway can be cut into strips of tissue

The normal human lung: ultrastructure and morphometric estimation of diffusion capacity

Abstract Eight normal human lungs, obtained from patients dying of causes not involving the lung. were totally fixed in situ by instillation of a glutaraldehyde solution into the airways shortly post mortem. The age range was 19–40 years, average body weight was 74 kg and the average lung volume 4300 ml. Stratified random samples from twelve regions were morphometrically studied by electron microscopy using stereological methods. The fine structure of the human lung parenchyma as seen by scanning and transmission electron microscopy is described. The alveolar surface area was found to be 143 m2 on the average (± 12); this value is 75% higher than previous light microscopic estimates mainly because of higher resolution of the electron microscope thus leading to a different definition of ‘alveolar surface’. Capillary surface area and volume were 126 m2 (±12) and 213 ml (± 31), respectively. The arithmetic mean thickness of the human air-blood tissue barrier was estimated at 2.2 μm (±0.19) and is thus considerably thicker than that found in other mammals: the same holds for the harmonic mean thickness of the barrier (0.62 μ±0.04). This appears to be related to a particularly large amount of connective tissue fibers found in the human alveolar septum. From this morphometric information total pulmonary diffusion capacity for 02 was calculated; using the set of largest and smallest physical coefficients we obtained respectively a maximal value, D L max = 263 ml O 2 /min ·mm Hg (±34) , and a minimal value, D L min = 125 (± 18) . These data relate to the totally inflated and unfolded lung; if they are corrected to account for ‘available’ gas exchange surface, reduced because of the presence of an alveolar extracellular lining, we obtain for ‘available’ diffusion capacity : D L max * = 130–190 and D L min * = 62–91 ml O 2 /min·mm Hg respectively. These corrected values seem to agree with physiological estimates of human dl in exercise.

Morphometric estimation of pulmonary diffusion capacity: I. Model and method

Abstract A model and method for estimating pulmonary diffusion capacity from morphometric information gathered on fixed lung specimens is presented. The model divides the diffusion resistance into four serial components: surface lining layer, tissue, plasma and red cells. The morphometric parameters defining these partial resistances, which can be determined on electron micrographs of lung sections by means of stereological methods, include the gas exchange surfaces, the capillary blood volume and hematocrit, and the harmonic mean thickness of the various layers. The merits and limitations of different methods of tissue preparation under “supravital” conditions are discussed and the basic stereological procedures necessary for deriving the parameters are outlined. The introduction of appropriate physical coefficients permits calculation of morphometric values of pulmonary (D l ) and membrane (D m ) diffusion capacities which can be used in attempts of quantitative structure-function correlation.

Sampling designs for stereology

The purpose of this paper is to propose the necessary sampling techniques for estimating a global parameter defined in a solid opaque specimen (e.g. the total volume of mitochondria in a given liver, the total capillary surface area in a given lung, etc.). The geometry of the specimen often suggests a multi‐level or cascade sampling design at different magnifications, whereby the object phase at one level becomes the reference phase in the next level. The final parameter is then estimated as the product of the intermediate ratios with the volume of the specimen, which is estimated independently. Each level can be regarded as an independent sampling design; a given stereological project may be planned in terms of one or more of these designs.

The ultrastructure of the normal human skeletal muscle

SummaryMuscle biopsies were taken from the middle part of the vastus lateralis muscle of 9 men, who were not regularly involved in endurance training (M, average

Response of Skeletal Muscle Mitochondria to Hypoxia

\dot V_{{\text{O}}_{\text{2}} }

Electron microscopic demonstration of an extracellular duplex lining layer of alveoli.

max=61.3 ml/min·kg), 3 sedentary women (W,