Normal Structure,Function,and Histology of the Bone Marrow

Abstract

While a complete blood count provides information regarding possible treatment-related effects reflected in the peripheral blood, morphological evaluation of bone marrow cytology and paraffin sections provides information about bone marrow tissue architecture that otherwise would be missed by examination of peripheral blood alone. In decalcified, paraffin-embedded, hematoxylin and eosin (H&E)-stained sections of bone marrow, the more mature stages of the erythroid and myeloid cells, adipocytes, mast cells, and megakaryocytes can be identified, but lymphoid cells as well as immature progenitor cells can not be reliably identified. The quality of the marrow sections is governed by numerous variables related to specimen collection and processing and must be considered. In addition to discussing normal structure, function, and histology of bone marrow, methods for preparation and evaluation of bone marrow are presented.

Introduction

Blood and bone marrow is one of the largest organs in the body and is an important potential target organ of chemical exposure (Lund, 2000). For example, it was suggested that drug-related blood dyscrasias represented 10% of all blood dyscrasias reported in Sweden, and, 40% of those resulted in fatality (Bottinger and Westerholm, 1973). Since effects of a compound may be elicited in the circulating blood cell mass or the production of blood cells, evaluations of single or serial whole blood samples and smears, bone marrow aspirates, and marrow tissue sections are needed to understand the alterations in the leukon, erythron or thrombon that may occur in toxicity studies. Examples of blood and bone marrow toxicity can be found in Table 1.

Assessments of the blood and bone marrow have become routine procedures in the investigation of hematologic disorders in toxicology and safety assessment studies. Evaluation of blood has been extensively described (Jain, 1986a; Perkins, 1999; Ryan, 2001). The focus of this article will be evaluation of the bone marrow with the objectives of reviewing of some concepts regarding the bone marrow structure and function and review of qualitative and quantitative bone marrow evaluation methods. A review of various lesions of the bone marrow in laboratory rats, mice and dogs will be presented in a subsequent discussion (Travlos, 2006).

Bone Marrow Structure and Function

The bone marrow is found within the central cavities of axial and long bones (Figure 1). It consists of hematopoietic tissue islands and adipose cells surrounded by vascular sinuses interspersed within a meshwork of trabecular bone. It accounts for approximately 3% of the body weight in adult rats (Schermer, 1967), ~2% in dogs (Jain, 1986b) and ~5% in humans (Picker and Siegelman, 1999). The bone marrow is the major hematopoietic organ, and a primary lymphoid tissue, responsible for the production of erythrocytes, granulocytes, monocytes, lymphocytes and platelets. A brief discussion of bone marrow structure and function will be presented here; detailed descriptions can be found elsewhere (Jain, 1986b, Weiss and Geduldig, 1991; Wickramasinghe, 1992; Picker and Siegelman, 1999; Hoffman et al., 2000; Abboud and Lichtman, 2001).

The inner surface of the bone cavities and the outer surface of the cancellous bone spicules within the cavities are covered by an endosteal lining consisting of a single layer of flat “bone-lining cells” supported by a thin layer of reticular connective tissue; osteoblasts and osteoclasts are also found within the endosteal lining (Figure 2).

In long bones, one or more nutrient canals (containing a nutrient artery and 1 or 2 nutrient veins) pass through the cortical bone entering the marrow cavity obliquely. In flat bones, the marrow is served by numerous blood vessels of various sizes entering the marrow via large and small nutrient canals. After entry, the artery splits into ascending and descending branches that run parallel to the long axis in the central part of the marrow cavity, coiling around the primary venous marrow channel, the central longitudinal vein (Figure 3). These artery branches give rise to a multitude of small thin-walled arterioles (Figure 4) and capillaries that extend outwardly toward the cortical bone. Near the bone, the arterioles open up and anastomose with a plexus of venous sinuses. These venous sinuses drain via collecting venules that lead back centrally to the central longitudinal vein that then drains via the nutrient veins. The marrow has an extensive blood supply (Figure 5). Also, it appears that nutrient artery-derived capillaries extend into the Haversian canals, return to the marrow cavity then open into the venous sinuses. Thus, there is a circular pattern to blood flow within the marrow cavity, from the center of the marrow cavity toward the periphery of the marrow cavity then back toward the center. In long and flat bones, the blood supplies of the bone and bone marrow are interconnected through an endosteal network of vessels. The venous sinuses are thin-walled, consisting of a layer of flat endothelial cells with little to no basement membrane. The marrow does not have lymphatic drainage (Munka and Gregor, 1965).

Bone marrow innervation occurs with myelinated and non-myelinated nerves that enter through the nutrient canals. Some innervation also occurs through epiphyseal and metaphyseal foramina. Nerve bundles follow the arterioles with branches serving the smooth muscle of the vessles or, occasionally, terminating in the hematopoietic tissue amongst hematopoietic cells.

The hematopoietic tissue consists of a variety of cell types including, the blood cells and their precursors, adventitial/barrier cells, adipocytes, and macrophages. The hematopoietic tissue cells are not randomly arranged but demonstrate a particular organization within the tissue (Weiss and Geduldig, 1991) (Figure 2). For hematopoiesis to occur it must be supported by a microenvironment that is able to recognize and retain hematopoietic stem cells and provide the factors (e.g., cytokines) required to support proliferation, differentiation and maturation of stem cells along committed lineages. The hematopoietic microenvironment consists of adventitial reticular cells (e.g., barrier cells), endothelial cells, macrophages, adipocytes, possibly, bone lining cells (e.g., osteoblasts) and elements of the extracellular matrix. A more detailed discussion of the organization and function of the hematopoietic microenvironment can be found elsewhere (Weiss and Geduldig, 1991; Hoffman et al., 2000; Gasper, 2000a, 2000b, 2000c; Abboud and Lichtman, 2001).

Hematopoiesis is a compartmentalized process within the hematopoietic tissue with erythropoiesis taking place in distinct anatomical units (erythroblastic islands); granulopoiesis occurs in less distinct foci and megakaryopoiesis occurs adjacent to the sinus endothelium. Upon maturation, the hematopoietic cells, regulated by the barrier cells, traverse the wall of the venous sinuses to enter the bloodstream; platelets are released directly into the blood from cytoplasmic processes of megakaryocytes penetrating through the sinus wall into the sinus lumen. Details of the hematopoietic process can be found elsewhere (Jain, 1986b; Hoffman et al., 2000; Gasper, 2000a, 2000b, 2000c; Abboud and Lichtman, 2001).

The production, differentiation, and maturation of blood cells are regulated by humoral factors (Table 2). Some factors (e.g., BPA/IL-3) act on the more primitive cells and have a general action, while others (e.g., erythropoietin) act on later progenitors of a specific cell line. The sources of hematopoietic factors vary. Erythropoietin is produced primarily in the kidney with minor amounts from the liver and stimulates proliferation of committed erythrocytic progenitors and release of immature red cells; high levels increase the rate of differentiation into erythrocyte progenitors. Burst promoting activity (BPA) is produced by T-lymphocytes and macrophages. IL-3 is produced by T-lymphocytes and myeloid cells and may be the same macromolecule as BPA. Colony simulating factors are produced by a variety of cells, including macrophages/monocytes, fibroblasts, endothelial cells, lymphocytes, and placenta. Most interleukins, B-cell growth factor, and B-cell differentiation factor are derived from T-lymphocytes. IL-1 is produced by macrophages. Hormones also play a physiological role (Jain, 1986c). For example, circulating erythrocyte counts increase or decrease following gonad removal in female and male rats, respectively; administration of the respective sex hormones abrogated the effects of gonadectomy. Additionally, bone marrow morphology was altered following ovariectomy in female rats (Benayahu et al., 2000). Hormones of the pituitary, adrenals, thyroid and gonads appear to participate in erythropoiesis by altering erythropoietin production and erythroid progenitor response to other factors (Jain, 1986c). For example, androgens, thyroxine and growth hormone increase erythropoietin production; estrogen has an inhibitory erythropoietic effect.

Hematopoietic tissue is also sensitive to external influences and can become suppressed in response to dietary restriction, malnutrition, chronic inflammation, toxicity, and proliferative or neoplastic disorders (Jain, 1986d; Meierhenry, 1990; Wierda, 1990; Reagan, 1993; NTP, 1999; NIEHS, 1999, 2001; Lund, 2000; Weiss, 2000). In the rat, nutritional status is an important factor (Meierhenry, 1990). For example, diet restriction sufficient to halt weight gain in young rats decreased marrow erythroid elements by 50%, myeloid elements by 40%, and megakaryocytes by 20% (Brown, 1954). Complete restriction for 7 days reduced marrow cellularity by 30% (Furman and Gordon, 1955). Levin et al. (1993), demonstrated that a severe (25% of control) diet restriction for 2 weeks resulted in a relative erythrocytosis, lymphopenia, thrombocytopenia and bone marrow necrosis. And, it has been reported that, in the rat, protein intake, rather than total calories, is more important for maintenance of erythropoiesis (Bethard et al., 1958).

Hematopoiesis is a continuous process, but can be separated into distinct stages (Figure 6). The first stage involves uncommitted (pluripotent) stem cells contained in the bone marrow. These pluripotent cells have two primary functions. First, they maintain their numbers by a process of self-renewal and, secondly, they have the ability to give rise to all hematopoietic cells (erythrocytes, granulocytes, lymphocytes, monocytes, and platelets). They also appear to be found in greater numbers peripherally from the central axis, near the bone lining cells (Weiss and Geduldig, 1991; Picker and Siegelman, 1999; Gasper, 2000c).

Most of the understanding of hematopoietic proliferation and maturation has been derived using an irradiated syngeneic mouse model. Irradiated mice infused with donor cells give rise to hematopoietic foci in the spleen. In vivo, it was demonstrated that the stem cell pool could be measured in the rat and mouse (Till and McCulloch, 1961). Donor mouse cells injected into an irradiated mouse formed nodules in the spleen that could be visually counted. It was demonstrated that these splenic colonies were clones (Becker et al., 1963) and that the cells within these colonies were capable of self-renewal and differentiation into the major cell lines (Till et al., 1964). These splenic colonies have been shown to arise from a single pluripotent cell, which has been termed the colony forming unit-spleen (CFU-S). Depending on need, the bone marrow microenvironment and growth factors influence pluripotent stem cells to differentiate into committed stem cells of either the myeloid or lymphoid series (multipotential stem cells), or the second stage of hematopoiesis. They have a limited capacity for self-renewal, but have the potential to differentiate and develop mature progeny. Myeloid stem cells are the multipotential colony forming unit for granulocytes, erythrocytes, monocytes, and megakaryocytes (CFU-GEMM). The third stage is when committed stem cells, influenced by various growth factors, differentiate into lineage-specific progenitor cells. Progenitor cells exist in the bone marrow for megakaryocytes (CFU-Meg), lymphocytes, erythrocytes (BFU-E), eosinophils (CFU-Eos), and basophils (CFU-Baso). It appears neutrophils and monocytes arise from a common precursor (CFU-GM).

Lymphopoiesis occurs within the bone marrow microenvironment of adult mammals (Allen and Dexter, 1984). And B-lineage cells derived from the marrow can be identified by sequential changes in cell size and expression of immunoglobulin chains. Large pre-B cells are early precursors and contain heavy chains within the cytoplasm (Landreth and Kincade, 1984). Large pre-B cells (~10–13 microns) divide at least once to produce smaller (<9 microns) pre-B cells. With gene rearrangement, the small pre-B lymphocytes mature into B cells; they express K and A light chains in the cytoplasm (Wierda, 1990). The sequence of proliferation/maturation of B-lymphopoiesis is regulated by soluble factors secreted by stromal cells (Picker and Siegelman, 1999) and is sensitive to disruption by myelotoxic chemicals. For example, polyhydroxy metabolites of benzene (e.g., hydroquinone) have been shown to affect B-lymphopoiesis in the marrow causing maturation arrest of the B-cells at the pre-B cell stage (Wierda and Irons, 1982; King et al., 1988). T-cell lymphopoiesis occurs in the thymus that has been seeded with bone marrow-derived stem cells (Le Douarin et al., 1984). There is some evidence indicating that the prothymocytes have undergone some differentiation and/or commitment prior to relocating to the thymus (Picker and Siegelman, 1999). Morphologically, the rat or mouse bone marrow did not have, nor develop, lymphoid-cell aggregates or structures resembling follicles, even after immunization (Geldof et al., 1983). Further, while the marrow appeared to have a suitable microenvironment for immigrating antibody-producing cells, the cells dispersed singly, in a random arrangement, and did not appear to contribute to the immune response.

Methods for Evaluating the Bone Marrow

Evaluation of the hematopoietic system should be performed using a multi-pronged approach (i.e., peripheral blood exam including a CBC and differential, bone marrow smear exam or total femur counts and evaluation of cytocentrifuge preparations and/or bone marrow histopathology). Bone marrow histopathology and examination of peripheral blood are performed routinely in toxicity and safety assessment studies. Cytological preparations can be made routinely but evaluations are generally reserved for instances in which hematological changes are identified and determination of the underlying cause is needed.

Histopathology is a subjective assessment and is useful for evaluating bone marrow architecture, assessment of cellularity, estimation of M:E ratio (limited sensitivity), assessment of cell lineages, estimation of iron stores and other features (e.g., neoplasia, inflammation, pigment, infectious agents). Marrow smear evaluations and/or total femur counts provide quantitative results and better cellular morphology and determination of M:E ratios and maturation indices. While the following is a brief overview, more detailed information and techniques regarding histological and cytological evaluation of bone marrow have been described (Lewis and Rebar, 1979; Cline and Maronpot, 1985; Grindem, 1989; Tyler and Cowell, 1989; Wickramasinghe, 1992; Buckley, 1995; Andrews, 1998; Car and Blue, 2000; Freeman, 2000; Lanning, 2001; Valli et al., 2002).

When performing core marrow biopsies in adult dogs, it is usually necessary to take samples from the iliac crest, sternum, proximal humerus, trochanteric fossa of the femur or a rib, as the central femoral marrow cavity may be almost entirely replaced by fat. In the rat and mouse, however, the higher turnover of erythrocytes, due to a shorter circulating life span, means that the marrow space in most bones remains populated for life. And, in the rodent, it appears that the sternum and rib and, probably, humerus and proximal femur are important sampling sites, as the marrow at these sites remains hematopoietically active regardless of the animal’s age. For example, in the Fischer rat, from 4 months to 2 years of age, the sternum, ribs, humerus and proximal femur had a relatively similar marrow cellularity of approximately 68% (Cline and Maronpot, 1985). The distal femur and proximal tibia were similar at approximately 61% marrow cellularity. Regardless of age, the distal tibia had a noticeable lack of active hematopoiesis.

It is generally considered that, in the dog, normal bone marrow contains about 50% fat and 50% hematopoietic tissue; approximately 70 to 80% of the marrow being hematopoietic tissue in rats and mice (Valli et al., 2002). In the dog, however, marrow cellularity may range from 20% to 80% of the marrow space, depending on site and age (Weiss, 1986; Valli et al., 2002) (Figure 7). In a study evaluating Fischer rats, depending on the age and anatomic site, the average marrow space occupied by hematopoietic cells varied from 33–88% (Cline and Maronpot, 1985). In that study, the youngest animals had the highest marrow cellularity. For example, regardless of site, the mean marrow cellularity was ~80% at 2 months of age; by 2 years of age, the cellularity decreased to a mean of ~66%. Examples of normal rat and mouse bone marrow are shown in Figures 8 and 9.

In general, decalcified, paraffin-embedded, hematoxylin and eosin (H&E)-stained sections of bone marrow, the more mature stages of the erythroid and myeloid cells, adipocytes, mast cells, and megakaryocytes can be identified, but stem cells, immature myeloid, erythroid, lymphoid, monocytoid and stromal cells cannot be identified consistently. An estimation of general hematopoietic activity and the myeloid/erythroid ratio can also be performed. The erythroid elements are smaller with round, dense, and deeply basophilic nuclei (Figure 10). The cytoplasm is basophilic in the blast forms with increasing eosinophilia as they mature. The granulocytes have large bean-shaped nuclei that are less basophilic and more vesicular than the erythropoietic cells (Figure 10). Megakaryocytes are easily recognized by their large size and multilobulated nuclei (Figure 10). While mature lymphocytes can be identified in bone marrow smears (Figure 12), unequivocal identification of lymphoid lineage cells in decalcified H&E-stained sections of bone with bone marrow is not readily accomplished.

For smears, marrow can be obtained from an exposed surface of, or extruded from, the sternum or rib using a sable brush (e.g., size 000) moistened with homologous serum or physiologic saline to which EDTA has been added. The tip of the brush is rolled gently in the exposed marrow and several stripes of marrow are made on glass slides; multiple slides may be prepared for special stains or techniques. For dogs, imprints of sternal marrow or aspirates from the iliac crest, sternum, proximal humerus, or trochanteric fossa of the femur may be collected and placed in EDTA/physiologic saline. When marrow granules have been obtained, crush preparations may be prepared by flattening and spreading the marrow flecks between two glass slides to produce a smear (Figure 11).

Preparations from samples obtained without EDTA may be used as long as the smears are prepared immediately after sample collection. Smears can be made from antemortem or postmortem samples. Postmortem collections should be performed as soon as possible following sacrifice (within minutes); in the dog, it has been indicated that postmortem samples should be taken within 30 minutes following the animal’s death (Tyler and Cowell, 1989). The air-dried smears are then stained using a general procedure for Romanowsky-type staining of blood films. Since marrow smears tend to be thicker than blood films, staining times must be modified (at least 2x longer) to ensure adequate staining. Also, formalin may affect the staining qualities of the marrow smears, care should be used to avoid contact of the marrow or marrow smears with formalin or its vapors prior to staining. Using the 100x objective, a morphological assessment of individual cell lineages and a differential bone marrow count is performed. Differential counts are best performed on 500-cell counts but lower counts (e.g., 250-cell counts) have been utilized. The differential should classify cells by type (e.g., myeloid, erythroid, megakaryocytic, lymphoid, macrophage, etc.) and stage (e.g., rubriblast, prorubricyte, rubricyte, metarubricyte, etc.) (Figures 12–17). Relative percentages of the cells, M:E ratios and maturation indices can then be calculated. Examples of differential counts of bone marrow from normal rats and dogs are presented in Table 3.

For histopathology, numerous variables regarding specimen collection and processing (e.g., fixation, decalcification, embedding, sectioning and staining) may affect the quality of the sample to be evaluated and must be considered. For a detailed discussion, the reader is referred to references regarding processing procedures (Bennett et al., 1976; Beckstead et al., 1981; Moosavi et al., 1981; Weiss, 1987; Wickramasinghe, 1992; Callis and Sterchi, 1998; Hedrich and Bullock, 2004; Fero, 2005). A brief description follows.

Fixation

If collected at necropsy, marrow should be collected as soon as possible for histological assessment. Fixative solutions stabilize tissue by cross-linking proteins and the degree of cross-linking may result in denaturation of proteins and affect morphological, cytochemical or immunohistochemical qualities of the specimen. Because its readily available, 10% neutral buffered formalin (10% NBF) is the most commonly used fixative. For routine H&E sections, no special handling is required prior to tissue processing. 10% NBF-fixed tissues may be used for frozen sections and evaluation of adipocytes. For specimens to be used for immunohistochemistry, a prolonged fixation in 10% NBF may be detrimental (e.g., loss of antigenic epitopes due to continued amino group cross-linking and alteration of tertiary protein structure). Transferring the tissue to 70% ethanol after a 12-hour fixation reduces fixation-related effects.

Zenkers-type solutions (e.g., Zenkers-acetic acid, B5, Helly’s) are an often recommended, but, not commonly used fixative for bone marrow sections. These fixatives must be made fresh prior to use and tissues must be washed prior to processing or treated with Lugol’s iodine to remove pigments. Tissues are adequately fixed in 1–2 hours but then need to be transferred to alcohol; tissues become hard and brittle with prolonged fixation. Fine precipitates may form in the tissue and may become problematic when special stains are applied (e.g., silver stains). Additionally, the fixative constituent mercuric chloride is toxic and readily absorbed through skin. There are zinc chloride-based, Zenkers-like fix-atives (e.g., AZF); these have similar properties to Zenkers fluids but no mercuric chloride. Zenkers-type fixatives are useful for immunohistochemical procedures. And, they act as a mordant for the Giemsa stain resulting in improved nuclear detail.

Bouin’s solution is the preferred fixative by some investigators for bone marrow sections. Tissues are fixed overnight, followed by a 2 to 4-hour post-fixation wash (in water) and storage in 70% ethanol. Tissues become hard and brittle with prolonged fixation. The picric acid constituent stains tissues yellow. Since Bouin’s contains acetic acid, it may be used for decalcification; this may take an extended period of time and requires frequent (weekly) solution changes. Delicate morphological detail is not well preserved, but it does act as a mordant for basic aniline dyes (e.g., H&E) improving staining.

Decalcification

For paraffin-embedded sections, bone marrow specimens must be decalcified prior to sectioning; specimens for plastic embedding do not. There are numerous types of decalcifiers (e.g., acids, chelators, resins) and methods (e.g., immersion, sonication, microwave, ion-exchange) for the decalcification of bone marrow specimens. If preservation of enzyme reactivity or antigenic sites is required, selection of the method of decalcification is important. With larger specimens, some damage to tissue morphology, due to decalcification, is almost unavoidable (especially with the use of acid decalcifiers).

Both organic and mineral acids are commonly used for decalcification of bone marrow specimens. Organic acids (e.g., formic and acetic) decalcify slower than mineral acids (e.g., HCL and nitric). Mineral acids are commonly used in the rapid-type decalcification methods. Overdecalcification of tissue with acids, particularly mineral acids, results in tissue destruction. Thus, tissue in acid decalcifiers should not be left unchecked for extended periods (e.g., over a weekend). For proper decalcification, there must be even distribution of acid around bone sample. Thus, tissue suspension or mild agitation, such as gentle mixing (e.g., shaker/rocker) or air bubble percolation, may be useful. Additionally, acid decalcifiers may need frequent solution changes (e.g., daily) for adequate decalification. In general, acid decalcification is not recommended for samples needed for enzyme- or immunostaining. This caveat is particularly appropriate for the mineral acids like HCL or nitric; the organic acids appear to be somewhat better. All acid decalcifiers, especially mineral acids, reduce morphological quality and Giemsa stains may be unacceptable due to loss of basophilic-staining structures.

Chelating agents, such as EDTA, are also commonly used for decalcification of bone marrow specimens. Decalcification proceeds at a slower rate compared to the acid decalcifiers, especially compared to mineral acids. EDTA decalcification is recommended for enzyme- and immunostaining procedures. The action of EDTA is pH dependent (i.e., the higher the pH, the faster the decalcification). Since a high alkaline pH may also cause tissue destruction and resultant loss of cytological, enzyme or immunostaining qualities, however, the use of EDTA at a high pH (e.g., pH 10) must be avoided. With initial placement of tissue in a sufficient amount of chelator, solution replacement is usually not needed.

These decalcifiers may be applied individually or in combination using immersion, microwave, sonication or electrolytic methods. For immersion techniques, tissues are placed in adequate decalcifier at ambient temperature. This is the slowest of the aforementioned methods but causes the least artifactual tissue damage if the sample is overdecalcified and H&E staining is generally adequate. Microwave techniques utilize a microwave to heat a water bath in which a container with tissue immersed in a decalcifier is placed.

Decalcification is enhanced (especially with the mineral acids) but it is easy to heat-damage the tissue (especially at >45°C). Using 70% power for 20 minutes with 10-minute cool-down intervals help diminish the effects of heat. H&E staining is generally adequate, but one may see dark marrow components with sparse, smudged nuclei (probably related to heat damage). Sonication techniques involve immersing the tissues in a sonicator containing decalcifier and sonicating the tissues. The speed of decalcification is enhanced and H&E staining is adequate, but there can be cytological alterations similar to what occurs using the microwave method. The electrolytic method involves suspending a single tissue in an acid decalcifier between 2 electrodes and passing a weak electrical current to enhance decalcification time. This method is slightly faster than immersion and the staining is comparable to the immersion method. However, the setup is elaborate and not amenable to a high-throughput operation; this requires frequent solution changes and heat damage can occur.

Decalcification by ion-exchange varies from the above methods in that a calcium-sequestering resin is used in combination with an acid decalcifier. This technique appears to be faster than immersion methods and appears to provide the best morphology for tissues stained with H&E. A resin (e.g., Win-3000) is placed in the bottom of a container, which is then filled with an acid decalcifier (typically an organic acid such as formic acid); the tissue(s) are immersed in the decalcifier. Since the resin sequesters the calcium, a chemical precipitation method for determination of demineralization endpoint cannot be performed. The resin may be reused and daily solution changes are unnecessary. However, this method is not amenable to high throughput operations due to the availability and cost of the resin, limited tissue capacity in the decalcification chamber and the time required for thorough washing of the resin for reuse.

Demineralization Determination

Determination of the demineralization endpoint is an important step. Overdecalcification results in tissue destruction and underdecalcification results in poor dehydration and infiltration and, ultimately, sectioning. The X-ray method is the most accurate but requires the appropriate equipment. Also, tissues treated with fixatives containing mercury or other heavy metals are rendered radiopaque; thus, X-ray cannot be used for these tissues. The chemical precipitation method (e.g., calcium oxalate precipitation) is reliable and easily applied for acid decalcifiers. A modification of this method also may be used for EDTA decalcified tissues. Mechanical bending or probing is a subjective assessment and is by far the easiest method and is often done. But, it is not recommended as it is not reliable and the tissue manipulation may disrupt architecture or dislodge soft tissue components from the sample. Cutting the sample has also been used and appears to be an accepted practice.

Tissue Processing and Staining

Because of its routine use, H&E-stained sections of paraffin-embedded bone marrow tissue are typically evaluated histologically. However, H&E does not provide consistent hematopoietic cell differentiation. Section thickness is an important factor and 3-micron sections provide better cellular morphological detail compared to thicker (≥5-micron) sections (Figure 18). Cytological quality is not, however, as good compared to Romanowsky-stained cytology preparations. Giemsa–stained sections provides better morphological detail compared to H&E-stained sections and are more easily compared to Romanowsky-stained smears.

For the Giemsa stain, however, acid-decalcified tissue results in loss of basophilic-staining structures. The use of chelating-type decalcifiers improves Giemsa staining. Plastic embedding provides a significant improvement in cellular morphology. Additionally, there is no need for decalcification, thus, reducing shrinkage artifact and loss of cellular detail related to decalcification methods. Thin sections (≤3 microns) can be easily produced, improving cytological quality. However, plastic embedding is time, labor and cost intensive and, thus, not justified for a high-throughput operation. Prussian Blue stain is used for the assessment of iron stores; acids in decalcifiers or fixatives may washout iron stores. Thus, tissue iron could be underestimated.

Summary

Since hematopoietic system is a potential target organ of chemical exposure, evaluation of the blood and bone marrow is important component of any toxicity or safety assessment study. Evaluation of the hematopoietic system should routinely include a CBC and differential and bone marrow histopathology. While the CBC provides information regarding possible compound-related effects demonstrated in the peripheral blood, morphological evaluation of bone marrow provides information about bone marrow tissue architecture (e.g., cellularity, cell linages, vascular or stromal alterations, inflammation, necrosis), estimation of iron stores and identification of other features (e.g., pigment, infectious agents, proliferative or neoplastic disorders) that otherwise would be missed by examination of peripheral blood alone. The quality of the marrow sections, however, is governed by numerous variables related to specimen collection and processing and must be considered.