ultrastructure and lectin histochemistry of equine cutaneous histiolymphocytic lymphosarcomas

http://vet.sagepub.com/Veterinary Pathology Online

http://vet.sagepub.com/content/26/5/409The online version of this article can be found at:

DOI: 10.1177/030098588902600506

1989 26: 409Vet PatholP. G. Detilleux, N. F. Cheville and B. J. Sheahan

Ultrastructure and Lectin Histochemistry of Equine Cutaneous Histiolymphocytic Lymphosarcomas

Published by:

http://www.sagepublications.com

On behalf of:

Pathologists.American College of Veterinary Pathologists, European College of Veterinary Pathologists, & the Japanese College of Veterinary

can be found at:Veterinary Pathology OnlineAdditional services and information for

http://vet.sagepub.com/cgi/alertsEmail Alerts:

http://vet.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

by guest on July 10, 2011vet.sagepub.comDownloaded from

http://vet.sagepub.com/

http://vet.sagepub.com/content/26/5/409

http://www.sagepublications.com

http://vet.sagepub.com/cgi/alerts

http://vet.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav


Vet. Pathol. 26:409419 (1 989)

Ultrastructure and Lectin Histochemistry of Equine Cutaneous Histiolymphocytic Lymphosarcomas

P. G. DETILLEUX, N. F. CHEVILLE, and B. J. SHEAHAN US Department of Agriculture, National Animal Disease Center, Ames, IA, USA, and

Department of Veterinary Pathology, University College Dublin, Ballsbridge, Dublin 4, Ireland

Abstract. Tissues from subcutaneous lymphosarcomas and regional lymph nodes were examined by light and electron microscopy and by lectin histochemistry. Tumors were composed of two major cell types: small lymphocytes with few organelles and pleomorphic histiocytic cells with undulant surfaces, large numbers of cytoplasmic vacuoles, and many mitochondria with large crystalline inclusions. A large gram-positive coryneform bacterium was isolated from tumor nodules but was not identified morphologically in tumor tissues. Evaluation of sections of tumors with lectins as histochemical probes revealed three staining patterns: 1) lectin labeling histiocytic cells only (wheat germ, succinylated-wheat germ, Phaseolus vulgaris and soybean agglutinins); 2) lectins labeling histiocytic, interstitial and some lymphoid cells (concanavalin A, and Pisum sativum, Lens culinaris, and Ricinus communis I agglutinins); and 3) lectins failing to label any cell (peanut, Sophora japonica, and Ulex europaeus I agglutinins). In the lymph node, macrophages were labeled by lectins of groups 1 and 2; interdigitating reticular cells were labeled by group 2 lectins. Lectin staining of histiocytic cells in tumor tissues suggested that these were reactive cells and that lymphoid cells were the primary neoplastic component.

Lymphosarcoma is the most common tumor of the hematopoietic system of the h o r ~ e . l l , ~ ~ About 58% of the cases are leukemic, 25% have normal blood pic- tures, and 15% are 1e~kopenic.l~ Although atypical lymphocytes commonly appear in blood smears of leukemic animals, there is no precise cytologic classifi- cation system for lymphosarcoma in the horse. Equine lymphosarcoma has been subdivided into four clinical forms: multicentric, alimentary, thymic, and cuta- neomZ3 The cutaneous form is rare, there being two of 88 and one of 16 in two collections of equine lymphosarcoma.

In 1980, Sheahan et al. reported two cases of a unique cutaneous lymphoma termed “histiocytic lymphosarcoma.” l 9 Lesions that consisted oflarge histiocytes and lymphoblasts were limited to the subcutis. No retroviral virions were found by electron microscopy in tumor tissue or in cells cocultivated with tissue culture cell lines. No reverse transcriptase activity was dem- onstrated in tumor cell suspensions. Coryneform bacteria suggestive of the order Actinomycetales were present extracellularly and within histiocytes. They replicated after 14 days in cell culture medium but did not grow in solid media under aerobic or anaerobic environments. This bacterium produced no clinical effects when given subcutaneously to horses.

We examined a mare from Iowa that had tumors similar to the neoplasm observed by Sheahan et al. in two Irish horses. The lesions also contained mixtures of lymphocytes, histiocytes, and coryneform bacteria,

but no overt evidence of retroviruses, and had cells with striking mitochondria1 inclusions. We have com- pared the present case to previous cased9 and have characterized lymphoid and histiocytic components of the neoplasm using lectin histochemistry.

Lectins, which have specific affinity for the carbohydrate moiety of glycoconjugates, are increasingly used as histochemical probes for the study of neoplasms. lx5

In human tissues, peanut agglutinin is a useful tool to detect macrophage/histiocytes and Reed-Sternberg cells, while Ulex europaeus I agglutinin is a reliable marker for endothelial cells and tumors of vascular origin.5.1OJ6

Materials and Methods A 15-year-old Standardbred mare with clinical diagnosis

of lymphosarcoma had been progressively losing weight and was emaciated and anorexic. Subcutaneous nodules of var- ious sizes were on her head, neck, and back. During clinical evaluation, individual tumors increased in size, and none were seen to regress. The horse was killed by intravenous injection of sodium pentobarbital and succinylcholine, followed by exsanguination by severing the axillary blood vessels. Tissues were collected for virologic and bacteriologic analysis.

Eight subcutaneous tumor nodules were cut in midsection and fixed in 10% formalin. Tissues fixed in 10% formalin were dehydrated in graded alcohols, embedded in paraffin, cut at 6 pm, and stained with hematoxylin and eosin, Mas- son’s trichrome, phosphotungstic acid and hematoxylin,

409 by guest on July 10, 2011vet.sagepub.comDownloaded from


410 Detilleux, Cheville, and Sheahan

Table 1. Lectins and their inhibitors used in this study.*

Lectin Inhibitory Concen-

Sugar tration LectidSource Abbreviation Nominal Sugar Specificity (jdml)

I. Glucose/mannose group: Conavalia ensiformis (jack bean) Lens culinaris Pisum sativum

Triticum vulgare (wheat germ) 11. N-acetylglucosamine group:

Succinylated Triticum

111. N-acetylgalactosamine/galactose group: Glycin maximus (soybean) Sophora japonica Arachis hypogaea (peanut) Phaseolus vulgaris Ricinus communis I

Ulex europaeus I IV. L-fucose group:

Con A LCA PEA

WGA

S-WGA

SBA SJA PNA PHA-L RCA-I

UEA-I

aMan > aGlc > GlcNAc aMan > aGlc > GlcNAc aMan > aGlc = GlcNAc

GlcNAc(P1, 4GlcNAc),_, >

GlcNAcV 1,4GlcNAc) ,-2 > PGlcNAc > Neu5Ac

PGlcNAc

a and PGalNAc > a and @Gal a and PGalNAc > a and @Gal GalP1,3GalNAc > a and @Gal GalP 1,4GlcNAcP 1, 2Man B and aGal > GalNAc

aLFuc

aMetMan aMetMan aMetMan

GlcNAc

GalNAc GalNAc D-Gal GalNAc D-Gal

L-FUC

2.5 2.5 2.5

0.25 2.5

2.5 5.0 2.5 2.5 2.5

5.0 * Modified from Goldstein and Poretz.8

Giemsa stain, alcian blue at pH 2.5, and the periodic acid- Schiff technique.

For electron microscopy, four tumors were sectioned and sampled in 1-mm-square cubes. They were fixed in 2.5% glutaraldehyde, rinsed in cacodylate buffer, stained en bloc in 1% osmium tetroxide, dehydrated in alcohols, cleared in propylene oxide, and polymerized in epoxy resin. Ultrathin sections were stained with uranyl acetate and lead citrate. Four sections from each block were cut and examined with a Philips 410 electron microscope.

Lectin histochemistry was done on formalin-fixed paraffin embedded specimens of three subcutaneous tumor nodules and one lymph node from the Iowa mare and one subcutaneous tumor from a previously reported case in an Irish mare.I9 Biotinylated lectins, avidin-D and biotin-labeled peroxidase (Vector Laboratories, Burlingame, CA) were used in this study (Table 1). Tris-buffered saline (TBS) (pH 7.4) containing 1 mM calcium, manganese and magnesium chlorides was used as a diluting and washing solution. Unless specified, all steps were done at room temperature in a humidified chamber.

Sections, 4 pm thick, were attached to acid washed poly- lysine coated glass slides and were deparaffinized for 15 minutes in xylene followed by 4 minutes in two changes of absolute ethanol and two changes of95% ethanol. Endogenous peroxidases were inhibited by immersing slides for 1 hour in absolute methanol containing 0.6% H,O, and 0.074% HCl. Following a 20-minute rinse in double distilled water and in TBS, sections were immersed for 10 minutes in TBS containing 1 .O% bovine serum albumin (TBS-BSA). Excess TBS- BSA was blotted from the sections, and the specimens were incubated for 1 hour with the biotinylated lectin brought to the optimal dilution using TBS. After a 10-minute rinse in TBS, the specimens were incubated for 1 hour with avidin-

biotin-peroxidase complex freshly prepared by incubating 20 p1 avidin-D and 20 pl biotinylated-peroxidase in 5 ml TBS. Following a 10-minute rinse in TBS the sections were incubated for 5 minutes with 3,3-diaminobenzidine diluted to 0.06% in Tris buffer (pH 7.6) containing 0.03% H,O,. Sec- tions were washed for 10 minutes in TBS, counterstained with Harris hematoxylin, dehydrated in graded ethanol and xylene, and mounted with paramount.

After rehydration, sections were incubated with neuraminidase (1 IU/ml, Sigma, type V) and/or galactose oxidase (5 IU/ml, Sigma) for 18 hours at 37 C and at room temperature, respectively. Following enzyme treatment, the sections were washed in TBS and stained with peanut agglutinin, wheat germ agglutinin, succinylated-WGA or soybean agglutinin.

Controls for lectin staining included oxidation with 1% periodic acid for 10 minutes prior to lectin staining and prein- cubation of the lectin for 10 minutes at room temperature with a 0.1 -M solution of its specific inhibitory sugar (Table 1).

Samples of tumor tissue were collected aseptically at nec- ropsy. For bacterial isolation, cores from these samples were homogenized in brain heart infusion and inoculated onto blood agar plates and into chopped meat broth. A 2-day culture of the isolated organism was sampled, fixed in glutaraldehyde, embedded in gelatin cubes, and processed for electron microscopy as described above. For viral isolation, samples were minced into 1- to 2-mm fragments and cocultivated with bovine embryonic spleen cells and rabbit kidney primary cell monolayers. Retroviral detection was attempted by cocultivation with transformed feline (F-8 1) cells for direct syncytial assay.24

Sera from the mare and three unrelated horses were examined for precipitating antibodies to the isolated organism in an agarose immunodiffusion system. Crude extracts of



Equine Cutaneous Histiolymphocytic Lymphosarcoma 41 1

Fig. 1. Subcutaneous nodules of tumor.

sonicated bacteria and supernatants of a 2-day broth culture were used as antigens.

Results Tumors were small, movable, subcutaneous nodules

distributed randomly on the head, neck, and flank (Fig. 1). Histologically, tumor tissues from all sites were similar and contained, in order of frequency, lymphoid cells, histiocyte-like cells, and scattered interstitial cells interwoven with fine collagen fibers (Fig. 2). Lymphoid cells varied widely in size, but small round cells with granular basophilic nuclei and indistinct cytoplasm predominated; mitotic figures were abundant, aver- aging more than 10 per high power field. Histiocyte- like cells, characterized by an abundant eosinophilic cytoplasm and pale vesicular nucleus, varied in shape from large round cells with eccentric nuclei to stellate cells with poorly outlined branching cytoplasmic processes. Some histiocyte-like cells contained a lymphoid cell in their cytoplasm (Fig. 2). Interstitial cells were elongate to stellate with cigar-shaped nuclei. Micro- scopically, lymph node architecture was normal. No lesions were detected in other organs.

Ultrastructural studies

Lymphoid cells had small nuclear and cytoplasmic volumes and few cvtodasmic organelles. Mitochon-

Fig. 2. Edge of subcutaneous tumor nodules; histologic section. Tumor mass has large histiocyte-like cells, small lymphoid cells and stellate interstitial cells (arrows). Lym- phoid cell within histiocyte-like cell cytoplasm (open arrow). Blood vessels are inconspicuous.

dria were small, and only one to ten were present per cross section of cell. Golgi complexes were large, but endoplasmic reticulum was absent or present in very small amounts (Fig. 3).

Histiocytic cells were uniform and had large cytoplasmic and nuclear volumes and their cytoplasm contained large amounts of intermediate filaments, vacuoles, and mitochondria with large lamellar protein inclusions (Figs. 4, 5) .

No mast cells were found in tumor tissue, and vascular tissue was not prominent. Small capillaries had endothelial cells that were high and contained large cytoplasmic volumes, many ribosomes, and dense granules.

Bacteriologic and virologic studies

No bacteria or viruses were seen in the sections of tumor examined by light and electron microscopy. At- tempts to identify retroviruses were unsuccessful. Ho- mogenates of tumor yielded no growth for 5 days when plated on blood agar. After a 1-week incubation, a pleomorphic Gram positive organism was isolated from chopped meat broth. Ultrastructurally, bacteria had heavy cell walls and thick capsules (Fig. 6) . Cultures of this organism, which had the biochemical and ul-

* - Y trastructural characteristics of a coryneform bacteri- by guest on July 10, 2011vet.sagepub.comDownloaded from


by guest on July 10, 2011 vet.sagepub.com Downloaded from


Equine Cutaneous Histiolymphocytic Lymphosarcoma 413

Fig. 4. Histiocyte-like cell adjacent to lymphoid cell (top left). Mitochondria1 matrix contains dense, rhomboid, laminar inclusions. Intermediate filaments (IF) occupy much of the cytoplasm. Cisternae of the nuclear envelope and rough endoplasmic reticulum are dilated and contain fibrillar material. Endoplasmic reticulum is irregular and ribosomes irregularly attached. Nuclear pore (NP). Golgi complex (G) bears many coated vesicles. Vacuoles (V) resemble mitochondria with cristolysis (M).




Fig. 5. Histiocyte-like cells with dense lamellar inclusions in mitochondria.

the staining pattern by WGA, i.e., round to oval cells without elongated cytoplasmic processes and stellate cells with cytoplasmic processes. Isolated lymphocytes surrounded by stained cytoplasm of a histiocytic cell were frequently seen, especially following concanavalin A (Con A) staining (Fig. 9). Most lymphocytes were stained by Con A (Fig. 9), but only half of these cells were positive following Pisum sativum agglutinin (PEA), Lens culinaris agglutinin (LCA), and Ricinus communis I agglutinin (RCA-I) labeling. Bundles of connective tissue at the periphery and throughout the tumor were intensely stained by the lectins of this group, especially PEA, LCA, and RCA-I.

All cellular types remained unstained with lectins of the third group (Fig. 1 1). Following neuraminidase in-

cubation, the staining pattern for peanut agglutinin (PNA) became essentially identical to WGA, with staining of all histiocyte-like cells and of numerous interstitial cells (Fig. 12). Incubation with galactose oxidase after neuraminidase treatment abolished the positive reaction with PNA.

Labeling of one tumor nodule from the Irish case of cutaneous histiocytic lymphosarcoma yielded similar results (Fig. 10). The staining reaction, however, was always weaker and showed more variability between different areas of the same section.

In the lymph node, a positive staining reaction was seen only in two cellular types: macrophage/histiocyte and interdigitating reticular cells. Staining of the macrophage/histiocytes was observed with lectins of the




Fig. 6. Section of coryneform bacterium. Thick cell wall and fibrillar capsule.

first and second groups. Positive macrophages were present throughout the organ but were more abundant in the medullary area (Figs. 13,14). Macrophages present in the subcapsular sinus and in the connective tissue surrounding the lymph node were also stained. Lectins of the second group and WGA stained reticular cells (Figs. 13, 14). Stained reticular cells were more

prominent in the medullary sinuses. Without prior neuraminidase incubation, PNA failed to react with any cellular type. After overnight incubation with neuraminidase, however, macrophage/histiocytes and most of the reticular cells were strongly labeled by PNA. Neuraminidase incubation also influenced WGA and s-WGA staining. Desialylation abolished the positive

Table 2. Lectin binding sites in cells of the subcutaneous tumor nodules and lymph node.

Subcutaneous Nodules Lymph Node

Histiocytic Interstitial Lymphoid Histiocytes/ Interdigitating Lymphoid Cells Cells Cells MacroDhaaes Reticular Cells Cells Lectins

Group I: WGA S-WGA PHA-L SBA

Group 11: PEA LCA

Con A Group 111:

PNA SJA

RCA-I

UEA-I

+++* +++ +++

+

+++ +++ +++ +++

0 0 0

+lo t 0 0 0

+++ +++ +++ +++

0 0 0

O* 0 0 0

+ / O + / 0 + / O ++

0 0 0

+++ +++

+ + / O

+++ +++ +++ + + II

0 0 0

+§ 0 0 0

+++ ++

+++ +++

0 0 0

0 0 0

* + + + = strongly positive. t +/0 = half the cells are positive. * 0 = no staining. 5 + = weakly positive. 11 ++ = positive.




Fig. 7. Subcutaneous nodule, Iowa horse. Treatment: succinylated-WGA (Triticum vulgure, Wheat germ) (binds only

Fig. 8. Subcutaneous nodule (section adjacent to Fig. 7). Treatment: WGA (Triticum vulgure, Wheat germ) (binds sialic

Fig. 9. Subcutaneous nodule, Iowa horse. Treatment: Con A (Conavuliu ensiformis, Jack Bean). Staining ofall cell types:

Fig. 10. Subcutaneous nodule, Irish horse. Treatment: RCA-I (Ricinus communis I). Staining of histiocytic and stellate

Fig. 11. Subcutaneous nodule, Iowa horse. Treatment: PNA (Aruchis hypogueu, Peanut). Negative. Binding of PNA

N-acetylglucosamine). Histiocyte-like cells lack elongated cytoplasmic processes.

acid and N-acetyl glucosamine). Some histiocyte-like cells have elongated cytoplasmic processes.

histiocyte-like, lymphoid and stellate interstitial cells. Lymphocyte within histiocyte-like cell cytoplasm (arrow).

interstitial cells.

prevented by sialic acid.




WGA staining of the reticular cells and increased the intensity of the histiocyte labeling by WGA and s- WGA.

Discussion The similarities between Irish and American cases

in histology, lectin-binding patterns, mitochondria1 crystalloids, and presence of coryneform bacteria in- dicate that the cases of equine histiolymphocytic lymphosarcoma are identical. The lymphoid cells appear to be the primary neoplastic component of this disease. The conclusion that these lesions do not represent a chronic inflammatory reaction was based on the following criteria: high mitotic index in the lymphoid cell population, multicentric distribution of the nodules, and lack of response to symptomatic treatment.

All lectins that stained histiocyte-like cells also reacted with the lymph node macrophages and histiocytes. This correlation suggests that histiocytic cells in the tumor are reactive rather than neoplastic. Surface glycoconjugates on neoplastic cells are often altered, resulting in changes in the lectin staining pattern.l,5 Ricinus communis I agglutinin (RCA-I) can be used to distinguish reactive histiocytes from malignant cells in human tumors of monocytic and histiocytic origin.I6 Benign-appearing histiocytes stain strongly with RCA- I, whereas cytologically malignant histiocytes are RCA-I negative. 5 ~ 1 6

The absence of staining of lymphocytes in lymph node is surprising. Others using a similar panel of lectins observed a varying staining pattern in paraffin- embedded lymphoid t i s ~ u e s . ~ ~ ~ ~ The low lectin concentration used in our study may explain this difference. At a high concentration (1 00-1,000 pg/ml), concanavalin A was shown to bind all lymphoid cells, while at a low concentration (1 0 pg/ml) binding was restricted to macrophage/histiocytes.2' Because lymphoid cells have limited cytoplasm, lectin binding sites may be less abundant, explaining the more localized and generally weaker staining of these cell^.^.^' In addition, specific binding sites of the lymphoid cells may have been lost during fixation and paraffin embedding. Car- bohydrate residues in glycolipids and some small oli- gosaccharides are lost during tissue proces~ing.~,'~

The staining of lymphocytes in tumor nodules could be due to their neoplastic nature. Malignant transfor- mation can lead to a loss, overexpression, or alteration

of normal lectin binding sites.5 In malignant cells, terminal sugar residues are frequently deleted, resulting in the expression of normally masked subterminal residues. I

Some lectins of the same nominal monosaccharide specificity exhibited a different binding pattern, e.g., RCA-I belongs to the N-acetylgalactosamine/galactose group but reacted like lectins of the glucose/mannose group. Lectins are classified into five groups according to their preferential binding to D-pyranose sugars.8 Representative lectins from four of these five groups were used in this study (Table 1). Although some lectins are specific to a single glycosyl unit, most possess large binding sites reacting with two to six sugar residues.8 Lectins of the same nominal specificity may recognize different features of these complex carbohydrate~. '!~,~ In addition, lectins may also bind to non- carbohydrate l igand~.~

Our observation that sialic acid is abundant in the equine histiocyte/macrophage is confirmed by the loss of peanut agglutinin (PNA) reactivity when galactose residues, exposed by neuraminidase treatment, are removed by galactose oxidase. The lectin staining pattern combined with enzymatic treatment clarifies the reactivity of carbohydrate moieties of glycoconjugates. PNA binds preferentially to the disaccharide P-D-galactose( 1-3)GalNAc, but, due to the addition of terminal sialic acid, these PNA binding sites are generally masked and unreactive with PNA when present at the cell sur- f a ~ e . ~ . ~ These sites react to PNA when sialic acid is removed with neuraminida~e.~

Two types of histiocyte-like cells were discernible by the presence or absence of elongated cytoplasmic processes following lectin labeling. This difference in morphology is associated with a variation in expression of sialic acid residues. Wheat germ agglutinin (WGA) binds either to sialic acid or to N-acetylglucosamine (GlcNAc), even if present as an internal res- i d ~ e . ~ , * Succinylation of WGA (s-WGA) reduces its affinity to sialic acid without affecting the binding to GlcNAc residues.8 Without prior neuraminidase treatment, WGA stained both types of histiocyte-like cells, while s- WGA and WGA after neuraminidase incubation stained only the histiocyte-like cells lacking cytoplasmic processes. Similar morphologic variation among RCA-I positive stromal macrophage/histiocytes occurs in human malignant lymphoma^.^^ In malignant histiocytosis, cytoplasmic extensions of RCA-I

c Fig. 12. Subcutaneous nodule (section adjacent to that in Fig. 1 1). Treatment: Neuraminidase + PNA (Aruchis hypogueu,

Fig. 13. Lymph node, Iowa horse. Treatment: WGA (Triticum vulgure, Wheat germ). Strong staining of histiocytes/

Fig. 14. Lymph node, Iowa horse. Treatment: RCA-I (Ricinus communis I). Strong staining of histiocytes/macrophages

Peanut). Removal of the sialic acid residues permits labeling of galactose residues (to which sialic acid was attached).

macrophages and weak staining of interdigitating reticular cells.

and reticular cells.



418 Detilleux, Cheville. and Sheahan

positive cells were interpreted as an attempt of reactive histiocytes to contain tumor cells.16

The coryneform bacteria in these cases appear to be an equine saprophytic bacterium that grows in an im- munosuppressed host. Horses with lymphoid tumors commonly are immunologically deficient,6 and it seems likely that this underlies the persistence of the coryneform bacterium in tumor tissue. Similar organisms isolated from one of the two Irish cases were inoculated subcutaneously into three horses, but no significant changes developed within the subsequent 8-month pe- riod.19 Intra- and extracellular organisms were observed in impression smears of the tumor tissue from one of the Irish cases;I9 but, as in the American case, they were not detected either histologicaly or ultrastructurally.

This bacterium has not been characterized but is probably related to similar isolates reported elsewhere, i.e., a coryneform bacterium has been isolated in pure culture from uterine pus in chronic endometritis in a mare.2 Growth, morphologic, and biochemical characteristics of the coryneform bacteria fit descriptions of Corynebacteriurn rninutissimurn, the cause of human erythrasma, a bacterial infection of the skin.18

The significance of mitochondria1 lamellar inclusions is not clear. Similar crystalline inclusions have been reported within mitochondria of endoneurial fibroblasts of the sacral nerve roots of horses with neuritis of the cauda e q ~ i n a ~ . ~ and in Schwann cells of fibroblasts of a 14-year-old mare with lymphosarcoma involving peripheral nerves.’ Mitochondria1 inclusions are seen in normal tissues from dogs, monkeys, and human beings, and in pigs with oral lead intoxication, rats with ammonia intoxication and in monkeys given ethanol.20 Image analysis of electron micrographs by computer should reveal occult periodicities in these crystals that may aid in their identifi~ati0n.I~ S-100 protein and peripheral nerve myelin proteins (P2, PO) have been detected in granular cell tumors,12 and S-100 protein is found in histiocytosis X cells;22 mitochon- drial protein may be of this type.

Acknowledgements The authors thank Dr. James T. Meehan for help in de-

veloping lectin staining procedures, Dr. Martin J. Van Der Maaten for virologic studies, and Judi Stasko, Allen Jensen and Robert Kappmeyer for technical assistance.

References 1 Alroy J, Ucci AA, Pereira MEA: Lectins: histochemical

probes for specific carbohydrate residues. In: Advances in Immunohistochemistry, ed. DeLellis RA, pp. 67-88. Masson Inc, New York, NY, 1984

2 Blue MG, Hannwacker M A Endometritis in the mare

caused by a coryneform organism. A case report and experimental studies. Cornell Vet 4:33 1-343, 1984

3 Cummings JF, de Lahunta A, Timony J F Neuritis of the cauda equina, a chronic idiopathic polyradiculoneu- ritis in the horse. Acta Neuropathol 46: 17-24, 1979

4 Dahme E, Deutschlander N: Die Neuritis der Cauda equina beim Pferde im elektronenmikroskopischen Bild. Zentralbl Veterinmed [A] 23:502-5 19, 1976

5 Damjanov I: Lectin cytochemistry and histochemistry. Lab Invest 57:5-20, 1987

6 Dopson LC, Reed SM, Roth JA, Perryman LE, Hitch- cock P: Immunosuppression associated with lymphosarcoma in two horses. J Am Vet Med Assoc 182:1239- 1241, 1983

7 Frankauser R, Bestetti G, Fatzer R, Straub R, von Tscharner C: Lymphosarkom mit Beteiligung periphe- rer Nerven beim Pferd. DTW 84:85-89, 1977

8 Goldstein IJ, Poretz RD: Isolation, physicochemical characterization, and carbohydrate-binding specificity of lectins. In: The Lectins: Properties, Functions, and Ap- plications in Biology and Medicine, ed. Liener IE, Sharon N, and Goldstein IJ, pp. 33-247. Academic Press, Lon- don, 1986

9 Hsu SM, Ree HJ: Histochemical studies on lectin binding in reactive lymphoid tissues. J Histochem Cytochem

10 Hsu SM, Yang K, Jaffe ES: Phenotypic expression of Hodgkin’s and Reed-Stemberg cells in Hodgkin’s disease. Am J Pathol 118:209-217, 1985

1 Jaeschke G, Rudolph R: Die Leukose des Pferdes. I. Nomenklatur, Klinik und Pathologie (Ubersichtsreferat). Berl Munch Tierarztl Wochenschr 98:88-94, 1985

2 Mukai M: Immunohistochemical localization of S- 100 protein and peripheral nerve myelin proteins (P2 protein, PO protein) in granular cell tumors. Am J Pathol 112:

3 Mukai M, Torikata C, Hisami I, Mikata A, Sakamoto T, Hanaoka H, Shinohara C, Baba N, Kanaya K, Ka- geyama K Alveolar soft part sarcoma. An elaboration of a three-dimensional configuration of the crystalloids by digital image processing. Am J Pathol 116:398-406, 1984

14 Neufeld JL: Lymphosarcoma in the horse: a review. Can J Vet Res 14:129-135, 1973

15 Ree HJ, Crowley JP, Leone LA: Macrophage-histiocytes in malignant lymphoma, small lymphocytic type (well-differentiated lymphocytic lymphoma). Cancer 56:

16 Ree HJ, Kadin ME: Lectin distinction of benign from malignant histiocytes. Cancer 56:2046-2050, 1985

17 Rittman BR, Mackenzie IC: Effects of histological processing on lectin binding pattern in oral mucosa and skin. Histochem J 15:467-474, 1983

18 Sarkany I, Taplin D, Blank H: Erythrasma: common bacterial infection of the skin. J Am Med Assoc 177:

19 Sheahan BJ, Atkins GJ, Russell RJ, O’Connor J P His- tiolymphocytic lymphosarcoma in the subcutis of two horses. Vet Pathol 17:123-133, 1980

31:538-546, 1983

139-146, 1983

11 17-1 123, 1985

130-134, 1961




20 Simpson CF, Bradley RE, Jackson RF: Crystalloid inclusions in hepatocyte mitochondria of dogs treated with levamisole. Vet Pathol 11: 129-1 37, 1974

21 Strauchen JA: Lectins as markers of lymphoid cells. I. Demonstration in tissue section by peroxidase technique. Am J Pathol 116:297-304, 1984

22 Takahashi K, Isobe T, Ohtsuki Y, Sonobe H, Takeda I, Akagi T: Immunohistochemical localization and distri-

bution of S- 100 proteins in the human lymphoreticular system. Am J Pathol 116:497-503, 1984

23 Van Den Hoven R, Franklin P: Clinical aspects of lymphosarcoma in the horse. Equine Vet J 15:49-53, 1983

24 Van Der Maaten MJ, Miller JM: Use of a continuous cell line for virologic and serologic investigations of bovine leukemia virus infections. Am J Vet Res 41:1785- 1788, 1980

Request reprints from Dr. P. G. Detilleux, Pathology Research Laboratory, National Animal Disease Center, PO Box 70, Ames, IA 500 10 (USA).



ultrastructure and lectin histochemistry of equine cutaneous histiolymphocytic lymphosarcomas

Documents