Some Aids to Histological Recognition of Hyaline Droplet Nephropathy in Ninety-Day Toxicity Studies

Introduction

Hyaline droplet accumulation associated with alpha-2u globulin (α2u-g) nephropathy in the male rat kidney is a relatively frequent response to certain hydrocarbons when tested in subchronic assays for toxicity. In a recent histopathological survey of 42 ninety-day toxicity studies conducted by the National Toxicology Program (NTP), National Institutes of Health, from 1991 to 2001, hyaline droplet accumulation in proximal convoluted tubules (PCTs) was one of the most common renal findings (Travlos G. S. and Hard G. C., unpublished observations). The syndrome of α2u-g nephropathy is well known and characterized histologically in ninety-day studies by accumulation of eosinophilic-staining protein droplets in the P2 segment of the proximal tubule, along with granular casts in the outer medulla (Swenberg and Lehman-McKeeman 1999). Usually, the droplet accumulation in PCTs is easily recognized after hematoxylin and eosin (H&E) staining, but this is not always the case in studies where the chemical is inducing a low-level response. One reason for this is that control male rats present with a variable range of droplets, which can be quite extensive in some animals, obscuring an overall group difference from a low level of chemically induced droplet accumulation. Luz and Murray (1991) have drawn attention to a further complication for recognition of cytoplasmic droplet accumulation in rodent kidney tubules associated with H&E staining methodology. When an aqueous solution of eosin is employed in the staining procedure, as opposed to alcohol-based eosin, the droplets are much more difficult to visualize.

Through examining many subchronic studies of chemicals that induce α2u-g nephropathy in F344 and Sprague-Dawley male rats, the author has made certain observations that have proved to be of personal assistance in the recognition of a hyaline droplet response related to α2u-g nephropathy. The purpose of this brief communication is to describe these histopathologic aids in the hope that they will be considered of some use to toxicologic pathologists in general.

Hyaline Droplet Pattern

Although it has long been known that young male rats of conventional strains have a well-developed contingent of hyaline droplets in the P2 segment of proximal tubule representing secondary lysosomes (Logothetopoulos and Weinbren 1955; Maunsbach 1966), it seems not to have been generally recognized that the droplets assume a specific pattern in control males (at least in three- to five-month-old males) that is disrupted after treatment with chemicals inducing α2u-g nephropathy. The difference between this control male rat pattern and the renal response to an α2u-g-binding chemical is best determined using a protein stain such as Mallory Heidenhain (MH) or fluorescence microscopy. Chromotrope-aniline-blue (CAB) stain has been recommended as well for detecting hyaline droplets (de Rijk et al. 2003).

Young adult male rats normally have cytoplasmic droplets in PCT cells distributed in irregular tracts running from the deep cortex into the outer cortex (Figure 1a). This distribution probably traces the course of the P2 segments of proximal tubule (Goldsworthy et al. 1988). There is a distinct pattern in the distribution of the droplets within tubule cells. Some epithelial lining cells have a dense intracytoplasmic accumulation of droplets of uniform size, particularly filling the apical region. Other tubules have cells with a sparser, random scatter of intra-cellular droplets of various sizes, but the droplets are usually small (Figure 1b). In each case, the droplets have distinctly rounded outlines, and a crystal-like angular profile is seen only rarely. The cells densely crowded with droplets appear to bulge into the tubule lumen (Figure 1c), and are referred to here as “droplet-congested” (D-C) tubule cells for convenience. D-C tubule cells are scattered through the irregular tracts of droplets in the cortex but are particularly noticeable in the deep cortex. It appears that Uwagawa et al. (1992) referred to these D-C tubule cells as an “eosinophilic body type” of hyaline droplet, but most important, they showed that these cells (along with the scattered intracytoplasmic droplets in adjoining PCTs) were densely positive for anti-α2u-g immunohistochemical staining, confirming the droplets as lysosome related.

In chemically induced α2u-g nephropathy, not only is there an increase in proximal tubule hyaline droplets, and in the area of cortex involved, but there is also a disruption of the normal droplet pattern. Conspicuously, in all but the mildest cases of a hyaline droplet response, the D-C tubule cells are no longer present. Instead, in addition to round droplets ranging from large to small, the proximal tubule epithelium contains scattered polyangular crystal-like profiles or irregularly-shaped clumps of protein (Figures 2a, 2b).

Precursors Of Granular Casts

Granular casts are recognized as a hallmark feature of α2u-g nephropathy. These lesions occur at the junction of the outer stripe of outer medulla (OSOM) and inner stripe of outer medulla (ISOM), representing lodgment of cellular debris where the lumen of the P3 tubule narrows into the thin descending limb of the loop of Henle (Hard et al. 1993). Granular casts are characterized by marked dilation of solitary tubules with lightly staining, granular eosinophilic debris, which is derived from exfoliated cortical cells engorged with protein. The tubule lining is flattened and the basement membrane slightly thickened. Although considered to be a pathognomic feature of α2u-g nephropathy, granular casts are not always found in association with chemically induced hyaline droplet accumulation in ninety-day studies. This could be due to a weak effect representing the lower end of severity of the response. Alternatively, depending on the intensity of the response, the development of these lesions may not coincide with the ninety-day time point.

To the author’s knowledge, it has not previously been reported that granular casts are preceded by lesions of much smaller dimensions consisting of P3 tubules containing eosinophilic cellular debris at the OSOM/ISOM junction in H&E-stained sections. The lining epithelium of these tubules is usually basophilic and not as compressed as in the well-developed granular cast, but the luminal debris is similar to the contents of granular casts (Figure 3). These tubules undoubtedly represent precursors of granular casts, and in the author’s experience, may be seen as early as ten days after commencement of test compound exposure, that is, long before true granular cast development. In the absence of easily recognizable granular casts, checking for these tubular precursors can support the presence of hyaline droplet accumulation as a response to the chemical treatment in ninety-day or shorter-term studies.

Fluorescence Microscopy

It was recorded some forty years ago by Maunsbach (1966) that lysosomal bodies in proximal tubules of the rat were autofluorescent. This phenomenon occurs also when rat kidney sections are stained with H&E (Hard and Snowden 1991). Fluorescence microscopy was used to highlight hyaline droplets accumulating in rats bearing histiocytic sarcoma, where the accumulating protein represented lysozyme secreted by the histiocytic tumor cells (Hard and Snowden 1991). Since then, this technique has been used by this author to facilitate recognition of the hyaline droplet response in male rat kidneys with α2u-g nephropathy. H&E-stained kidney sections are examined under ultraviolet illumination at a wavelength of 450–490 nm. Regardless of how poorly hyaline droplets are visualized under brightfield microscopy, ultraviolet illumination facilitates their detection. As with MH staining, control kidneys of young adult male rats show the normal pattern of D-C tubule cells interspersed with tubules containing a random scatter of small droplets (Figure 4a). After exposure to chemicals inducing hyaline droplet accumulation in subchronic studies, not only is the disrupted pattern of autofluorescing intracellular lysosomes and crystalloid structures with loss of D-C tubule cells accentuated (Figure 4b), but the fluorescent intensity of the response is brighter than the rather dull pattern in control males. In addition, it is much easier to get an appreciation of the extent of cortical involvement by the hyaline droplet response than with brightfield examination of H&E sections. Thus, ultraviolet illumination can be used for checking the presence of a hyaline droplet response when MH- or CAB-stained sections have not been prepared.

Grading Of Hyaline Droplet Response

For the purpose of studies conducted under the standards of Good Laboratory Practice (GLP) it is necessary to grade the severity of a toxic response, usually on a scale of 0 to 4. Grading a hyaline droplet response, however, is not particularly easy, and it is best done using tissue stained with MH or by viewing H&E-stained sections under ultraviolet illumination.

It is the recommendation of this author that hyaline droplet accumulation in subchronic studies be graded on a 0 to 4 scale, but recognizing only three levels of abnormal response. In this scheme, 0 represents the absence of hyaline droplets in PCT, as is characteristic of female rats. Grade 1 (minimal) represents the normal (control) pattern and amount of droplets in young adult male rats. Grades 2 (mild), 3 (moderate), and 4 (marked) are the increasing levels of severity in cases where a chemical has caused an increase in hyaline droplet accumulation. In control male rats (grade 1), the droplets usually stain or fluoresce less intensely than those induced by chemicals, but in some control male rats, the droplets can involve an area of up to 25% or 30% of the cortex. The typical pattern is the mixture of small, rounded droplets together with D-C tubule cells as described above, which are distributed in tracts running an irregular course through the depth of the cortex. In a grade 2 response, approximately 30% to 50% of the cortex is occupied by proximal tubules that contain protein droplets, but there is a disruption of the normal pattern to include angular profiles or clumps suggestive of crystal formation (Figure 2b), together with a partial or complete disappearance of D-C tubule cells. For grade 3 (moderate), approximately 50% to 75% of the cortex is involved in hyaline droplet accumulation; and for group 4 (marked), 75% or more. In these two grades, the disruption of the normal pattern is even more distinct, and there is usually a total absence of D-C tubule cells.

Acknowledging the control male rat pattern of hyaline droplets as grade 1, and scoring just three grades of increased droplet accumulation (instead of four) appears to simplify the grading procedure for hyaline droplet accumulation to a level where results are relatively reproducible between pathologists. To assess the severity of hyaline droplet nephropathy rather than hyaline droplet accumulation alone, however, the grading system would need to account for the presence and severity of related pathologic features, for example, granular cast formation.

In conclusion, it should be stressed that the above observations concern chemicals that induce nephropathy consistent with α2u-g nephropathy. It is very possible that other forms of chemically-induced droplet accumulation exist, and these may not necessarily show the same features that are described here.