J. Comp. Path. 1997 Vol. 117, 83 89

SHORT PAPER

- j

C P

Canine C u t a n e o u s P r o t o t h e c o s i s : an I m m u n o h i s t o c h e m i c a l A n a l y s i s o f the I n f l a m m a t o r y

Cel lu lar Infi l trate

j. P6rez, P.J. Ginel*, R. Lucena*, J. HervAs and E. Mozos Departamento de Anatomia y Anatomia Patol6gica Comparadas and *Departamento de Patologla

Clinica Veterinaria, Facultad de Veterinaria, Avenida Medina Azahara 9, 14005 Cdrdoba, Spain

Summary

In this immunohistocbemical study, the distribution of the cellular in- flammatory infiltrate associated with cutaneous canine protothecosis (Prototheca wickerhamiz) was evaluated by consecutive biopsies taken before, during and after treatment. Antibodies specific to canine immunoglobulins (IgG, IgM, IgA), human CD3 antigen (pan T-lymphocyte marker) and human myeloid/ histiocyte antigen (macrophage/neutrophil marker) were used. Before treat- ment, cellular infiltrate was very scanty in the inflamed areas, but it increased during the treatment, whereas the number ofprotothecal organisms decreased. Statistical analysis revealed an inverse relation between the number of protothecal organisms and the number of infiltrating macrophage/neutrophils (P<0"004), T lymphocytes (P<0"001), and cells containing immunoglobulin G (P<0"001), M (P<0"001) and A (P<0"001) at different stages of the disease. These findings suggest either that protothecal organisms inhibit the migration or proliferation of cellular inflammatory infiltrate or that only dead protothecal organisms induce an effective local immune response.

�9 1997 W.B. Saunders Company Limited

Introduct ion

Protothecosis is an uncommon disease caused by several species of achloro- phyllous algae which are ubiquitous in the environment and are usually saprophytic (Sudman, 1974; Tyler, 1990). In man and in the cat, the mu- cocutaneous form of protothecosis is the most common, whereas in the dog the systemic form is more often reported (Tyler, 1990; Matsuda and Mat- sumoto, 1992). The pathogenesis of this disease remains unknown, but the minimal pathogenicity and low incidence indicate that concurrent diseases or altered host immune response may play a role in the onset and dissemination of the infection (Rakich and Latimer, 1984; Thomas and Preston, 1990; Tyler, 1990). To our knowledge, seven of 20 reported cases of canine protothecosis were in collie dogs (Migaki et al., 1982; Rakich and Latimer, 1984; Macartney et al., 1988; Thomas and Preston, 1990). An inherited immunodeficiency has been suggested to explain the susceptibility of this breed to the disease (Thomas and Preston, 1990; Tyler, 1990). However, the specific immune response has

0021 9975/97/050083 + 07 $12.00/0 �9 1997 W.B. Saunders Company Limited

84 J. P6rez et al.

been studied in only one dog suffering f rom systemic protothecosis (Rakich and Lat imer , 1984).

T h e purpose of this s tudy was to analyse the distr ibution of T lymphocytes , neutrophils , macrophages and immunoglobu l in -bea r ing (IgG, IgM and IgA) cells in the in f lammatory infiltrate o f pyogranu loma tous lesions f rom a collie dog with cutaneous protothecosis , to obta in in format ion on the local immune response before, dur ing and after t rea tment .

M a t e r i a l a n d M e t h o d s

Four scrotal biopsies from a 4-year-old male Rough Collie with cutaneous protothecosis were used for this study (Ginel et al., 1997). The first biopsy (biopsy I) was taken at the time of our initial examination of the dog, which had a 2-month history of scrotal swelling and ulceration, refractory to antibiotic treatment. Three nodules in the right carpus, several footpad fissures, haemorrhagic crusts affecting the trunk, nasal depigmentation and mild nasal serous discharge were also observed. The histo- pathological diagnosis of a diffuse granulomatous dermatitis associated with Prototheca wickerhamii infection was confirmed by culture in agar containing peptone 1%, glucose 2% w/v and chloramphenicol 100 Ixg/ml, and by species identification (API 20C yeast identification system) on the basis of biochemical features (Matsuda and Mat- sumoto, 1992). The microscopical examination of smears from fine-needle aspirates of cutaneous nodules of the carpus revealed the presence of protothecal organisms. Treatment was based on ketoconazole (10 mg/kg/24 h, per os) and clotrimazole topically applied once a day. Clinical response was slow but progressive and, after 3 months of treatment, the rhinitis and nasal discharge resolved, the footpad fissures and the trunk lesioits healed, and the scrotal skin showed improvement. After 4 months of treatment the scrotal granulomata were the only lesions still present and they were significantly reduced. A new biopsy (biopsy II) of the scrotum was taken to determine whether protothecal organisms were still present, and the treatment con- tinued. Histopathological examination revealed a chronic superficial and deep diffuse dermatitis with prominent fibrosis associated with periodic acid-Schiff-positive (PAS +) organisms. After 5 months of treatment the epidermis regained its normal appearance and the remaining indurated scrotal lesion was surgically excised (biopsy III). The ketoconazole and clotrimazole treatment was recommended for another month, but the owner failed to administer it. Five months after the surgical excision a relapse occurred and the dog was presented again for examination because of scrotal swelling, this time accompanied by marked nasal erythema and depigmentation. The histopathological examination of a new scrotal biopsy (biopsy IV) revealed a recurrence

o f the protothecal infection. Ketoconazole treatment was administered at the same dosage as before. Six weeks of treatment were necessary for resolution of all the lesions, but treatment was maintained for a further 4 weeks. After an 8-month follow- up there had been no recurrence and the dog was clinically healthy.

Tissue samples were fixed in 10% neutral buffered formalin for 24 h and embedded in paraffin wax. The haematoxylin and eosin (HE), PAS and Grocott's silver methenamine techniques were used for the histopathological study. The avidin biotin-peroxidase (ABC) method described by P~rez et al. (1996) was employed for the immuno- histochemical study. The primary polyclonal antibodies used and their dilutions were the following: rabbit anti-canine IgG, 1 in 200; rabbit anti-canine IgM, 1 in 300; rabbit anti-canine IgA, 1 in 500; rabbit anti-human CD3, 1 in 300. A mouse anti- human myeloid/histiocyte antigen (MAC387) monoclonal antibody diluted 1 in 1000 was also used. All primary antibodies were obtained from Dako, Glostrup, Denmark. In negative controls, primary polyclonal antibodies were replaced by isotypic rabbit non-immune serum and the monoclonal antibody was replaced by mouse non-immune serum. Normal canine and human lymph nodes were used as positive controls for

Canine Cutaneous Protothecos i s 85

Fig. 1. Scrotal dermatitis; biopsy II. Diffuse granulomatous dermatitis showing protothecal organisms phagocytized and partly destroyed by multinucleated giant cells (arrowheads). PAS. x 400. Inset: Characteristic internal septation of some protothecal organisms that are composed of several endospores (arrow). Silver methenamine staining, x 640.

antibodies raised against canine and human antigens, respectively. Normal canine scrotal skin was also analysed as a control.

Immunoreactive cells for CD3, MAC387, and canine immunoglobulins (lgG, IgM and IgA) antibodies as well as PAS § organisms were counted in 30 fields of x 400 magnification (randomly selected) in the inflamed areas of serial sections from the four biopsies analysed. The results were expressed as the mean___ standard deviation of immunoreactive cells or PAS § organisms per field. A comparative analysis between: (1) the number of PAS § organisms and phagocytic cells (macrophages and neutrophils); (2) the number of PAS + organisms and T lymphocytes; and, (3) the number of PAS § organisms and immunoglobulin (IgG, IgM and IgA) -containing cells, was made by means of a Pearson test of correlation and a Statistica 4"5 Computer Program (Statsoft Inc., Rockville, MD, USA).

R e s u l t s

The histopathological study of biopsies I and IV revealed small necrotic foci located in the superficial and deep dermis and surrounded by a scanty infiltrate of neutrophils, macrophages and occasional giant cells. Lymphocytes, plasma cells, fibroblasts and collagen fibres were also scanty at the periphery of the lesions. In biopsies II and III, a chronic diffuse granulomatous dermatitis composed of a moderate infiltrate of macrophages, giant cells (Fig. 1), some neutrophils, numerous lymphocytes, plasma cells and prominent fibrous con- nective tissue were seen. In tissue samples from the four biopsies, PAS + yeast-like organisms were observed within the cytoplasm of neutrophils and

8 6 J . P ~ r e z et al.

Table 1 Distribution of the cellular infiltrate and PAS + Proto theca w i c k e r h a m i i organisms in scrotal

granulomatous dermati t i s from the four b iops i e s analysed

Biopsy* Number'~ of protothecal organisms

Numberst of cells labelled by antibody to

MAC 387 CD3 IgG IgM IgA

I 50.5• 17"3• 9"2• 13"1• 2"25• 3"7• II 9"2• 32"4• 30"3• 34"9• 22"2• 24"9• III 16"0• 17"2• 32"9• 28"4• 14'2• 23-9• IV 32"8• 18"0• 15.0• 13"2• 4"4• 8"2•

* Biopsy I was taken before treatment; biopsies II and III were taken after 3 and 4 months of treatment, respectively; biopsy IV was taken 5 months after treatment was discontinued. ~'Results are expressed as the mean per x 400 magnification field• deviation. Statistical analysis revealed significant correlation (P<0"05) between the number ofprotothecal organisms and the number ofCD3 + T lymphocytes (P<0"001), MAC387 + cells (P= 0-004), IgG + cells (P<0-001), IgM § cells (P<0-001) and IgA + cells (P<0"001).

macrophages, or free in the inflamed areas. These organisms ranged from 8 to 30 gm in diameter and the larger ones showed internal septation, which is characteristic of Prototheca spp. (Fig. 1). In biopsies II and III, numerous destroyed protothecal organisms were located in the cytoplasm of giant cells and macrophages. The number of protothecal organisms in the four biopsies analysed and the number of cells labelled by the antibodies used for the immunohistochemical study are shown in Table 1.

With the CD3 polyclonal antibody, used as a pan T-lymphocyte marker in the dog (Ferrer et al., 1992), immunoreactivity was found in a variable number of lymphocytes located at the periphery of granulomatous lesions and in areas of diffuse dermatitis. Lymphocytes from interfollicular areas and medullary cords of control lymph nodes were also reactive with this antibody. MAC387 monoclonal antibody reacted with the cytoplasm of neutrophils and macro- phages located both in granulomatous lesions (Fig. 2) and in areas of diffuse dermatitis. Immunolabelling by MAC387 was also found in cellular debris located in the necrotic foci of biopsies I and IV. Macrophages and some neutrophils located in the medulla and paracortical areas of normal canine lymph nodes used as positive controls also showed immunoreactivity for this antibody. Immunolabelling of immunoglobulin G gave a diffuse cytoplasmic pattern and was found in numerous plasma cells of the cellular infiltrate of biopsies II and III (Fig. 3), whereas in biopsies I and IV the number of IgG + cells was low or moderate. Similar results were obtained with the anti-canine immunoglobulin M and anti-canine immunoglobulin A antibodies, which reacted with numerous infiltrating cells in biopsies II and III and with a small number of inflammatory cells in biopsies I and IV (Table 1).

Total white cell count decreased from 7100 cells/gl at initial examination to 3900 cells/gl when treatment was initiated and increased to 11 300 cells/gl after the cessation of treatment. Differential counts showed a consistent mild lymphopenia not affected by treatment (mean 786 cells/gl) and neutropenia (2418 cells/gl) that resolved after treatment. Plasma IgG and IgM immuno- globulin concentrations remained slightly reduced throughout the follow-up period, with a mean of 725"30 mg/d l (normal range 1327"50 to 1791 "54 mg/dl)

C a n i n e C u t a n e o u s P r o t o t h e c o s i s 87

Fig. 2.

Fig. 3.

Scrotal granulomatous dermatitis; biopsy III. Intense immunoreactivity to human MAC387 monoclonal antibody is observed in numerous infiltrating cells located in a small granuloma surrounded by fibrous connective tissue. Avidin biotin-peroxidase method, Mayer's haematoxylin counterstain, x 400. Scrotal granulomatous dermatitis; biopsy III. Numerous plasma cells show intense and diffuse cytoplasmic immunoreactivity to canine immunoglobulin G polyclonal antibody. Avidin- biotin peroxidase method, Mayer's haematoxylin counterstain, x 400.

and 88"09 mg/dl (normal range 121"79 to 165-51 mg/dl), respectively. Mean plasma IgA concentration, although relatively low (21"62 mg/dl), was within the normal range (20"99 to 34-03 mg/dl).

D i s c u s s i o n

Altered host immune response has been considered a predisposing factor in naturally occurring protothecosis (Sudman, 1974; Tyler, 1990). In the dog, breed susceptibility may be a predisposing factor since a disproportionate number of cases have occurred in collie dogs (Tyler, 1990). However, although the dog in the present case showed mild neutropenia, lymphopenia and slightly reduced IgG and IgM plasma concentrations, these parameters are variable and there was no previous history of infectious or parasitic diseases, stressful conditions or concurrent symptoms suggestive of an immunodeficiency state. Its parents, and two littermates that were available for examination, were also normal. These observations suggest that the immune system defect needed for a successful infection is specific for the protothecal organisms.

Suppression of the cell-mediated specific immune response against Prototheca spp. has been described in human systemic protothecosis (Cox et al., 1974), whereas in the cutaneous form, neutrophil inability to kill the infecting organism (Venezio et al., 1982) or depressed natural killer activity (Tyring et al., 1989) have been reported. Rakich and Latimer (1984) reported that a collie dog with systemic protothecosis caused by Prototheca zopffi showed generalized

88 J. P~rez et aL

suppression of T-lymphocyte functions and inhibited chemotaxis of neutro- phils; both changes were attributed to a humoral factor, since the patient's serum inhibited chemotaxis of normal canine neutrophils, but it was not clear whether the depressed immune response was the cause or the effect of the disease.

In the case reported here, biopsies I and IV had a scanty cellular infiltrate associated with numerous protothecal organisms. This minimal inflammatory infiltration has often been reported in canine protothecosis (Migaki et al., 1982; Rakich and Latimer, 1984; Thomas and Preston, 1990; Tyler, 1990). After 3 and 4 months of ketoconazole treatment (biopsies II and III, respectively) most of the P. wickerhamii organisms had been killed and the inflammatory cellular infiltrate was more evident than in the lesions before treatment. This pronounced inflammatory reaction has been reported infrequently in canine protothecosis (Tyler, 1990). Statistical comparisons between the four biopsies revealed a significant inverse relation between: (1) the number of protothecal organisms and the number of infiltrating phagocytic cells (macrophages and neutrophils) (P=0"004); (2) the number of protothecal organisms and the number of T lymphocytes (P<0"001); and, (3) the number of protothecal organisms and the number of cells containing IgG (P<0"001), IgM (P<0"001) and IgA (P<0"001). These findings can be interpreted in the following ways.

(1) Protothecal organisms may inhibit the migration or proliferation of the inflammatory cellular infiltrate. The mechanism might be the suppression of the chemotaxis of neutrophils, previously demonstrated in man (Cox et al., 1974) and the dog (Rakich and Latimer, 1984). However, the infiltrate ofmacrophages, T lymphocytes and cells containing IgG, IgM or IgA was also low in lesions with a large number ofprotothecal organisms, suggesting that other mechanisms play a role.

(2) The lysis of protothecal organisms and release of antigens may induce an intense local immune response, which is not induced by intact organisms. In this way, the capsule of Cryptococcus neoformans inhibits the fungicidal activity and antigen-presentation process for human macrophages (Vec- chiarelli et al., 1994). The cell wall of Prototheca spp. may also play a role in the low local immune response developed in this disease.

It seems unlikely that, at the dosage used, ketoconazole/clotrimazole treatment influenced the immunohistochemical results. All the imidazoles act by inter- fering with the synthesis of ergosterol in the fungal membrane, and only at high dosages has ketoconazole shown immunomodulator effects, including suppression of neutrophil chemotaxis and lymphocyte blastogenic responses, inhibition of 5-1ipoxygenase activity and leukotriene production (Scott et al., 1995). All these actions would have resulted in a reduced inflammatory response, i.e., the converse of what was observed in this case.

In conclusion, this report shows an inverse relation between the number of P. wickerhamii organisms and the cellular infiltrate in canine cutaneous protothecosis, suggesting that this organism inhibits the local immune response, or that only dead protothecal organisms induce an effective inflammatory

Canine Cutaneous Protothecosis 89

response. Further studies should focus on the distribution of T-lymphocyte subpopulations and cytokines to clarify the pathogenesis of this disease.

R e f e r e n c e s

Cox, G.E., Wilson, J.D. and Brown, P. (1974). Protothecosis: a case of disseminated algal infection. Lancet, 17, 379-382.

Ferrer, L., Fondevila, D., Rabanal, R. and Ramis, A. (1992). Detection of T lym- phocytes in canine tissue embedded in paraffin wax by means of antibody to CD3 antigen. Journal of Comparative Pathology, 106, 311-314.

Ginel, PJ., Pdrez, J., Molleda, J.M., Lucena, R. and Mozos, E. (1997). Cutaneous protothecosis in the dog. Veterinary Record, in press.

Macartney, L., Rycroft, A.N. and Hammil, J. (1988). Cutaneous protothecosis in the dog: first confirmed case in Britain. Veterinary Record, 123, 494 496.

Matsuda, T. and Matsumoto, T. (1992). Protothecosis: a report of two cases in Japan and a review of the literature. European Journal of Epidemiology, 8, 397-406.

Migaki, G., Font, R.L., Sauer, R.M., Kaplan, W. and Miller, R.L. (1982). Canine protothecosis: a review of the literature and report of an additional case. Journal of the American Veterinary Medical Association, 181, 794-797.

P6rez, J., Mozos, E., Chac6n-M. de Lara, F., Paniagua, J. and Day, MJ. (1996). Disseminated aspergillosis in a dog: an immunohistochemical study. Journal of Comparative Pathology, 115, 191-196.

Rakich, P.M. and Latimer, K.S. (1984). Altered immune function in a dog with disseminated protothecosis. Journal of the American Veterinary Medical Association, 185, 681-683.

Scott, D.W., Miller, W.H. and Griffin, C.E. (1995). Dermatologic therapy. In: Small Animal Dermatology, 5th Edit., D.W. Scott, W.H. Miller and C.E. Griffin, Eds, W.B. Saunders, Philadelphia, pp. 221-224.

Sudman, M.S. (1974). Protothecosis. A critical review. American Journal of Clinical Pathology, 61, 10-19.

Thomas, J.B. and Preston, N. (1990). Generalised protothecosis in a collie dog. Australian Veterinary Journal, 67, 25-27.

Tyler, D.E. (1990). Protothecosis. In: Infectious Diseases of the Dog and Cat, G.P. Greene, Ed., W.B. Saunders, Philadelphia, pp. 742-748.

Tyring, S.K., Lee, P.C., Walsh, P., Garger, J,.u and Little, W.P. (1989). Papular protothecosis of the chest. Archives ofDen~o.to~ogy, 125, 1249-1252.

Vecchiarelli, A., Pietrella, D., Dottorini, M., Monari, C., Retini, C., Todisco, T. and Bistoni, F. (1994). Encapsulation of Cryptococcus neoformans regulates fungicidal activity and the antigen presentation process in human alveolar macrophages. Clinical Experimental Immunology, 98, 217-223.

Venezio, F.R., Lavoo, E., Williams, J.E., Zeiss, C.R., Caro, W.A., Mangkornkanok- Mark, M. and Phair, J.P. (1982). Progressive cutaneous protothecosis. American Journal of Clinical Pathology, 77, 485-493.

Received, October 21 st, 1996 1 Accepted, February 27 th, 1997J