SciELO - Scientific Electronic Library Online

vol.53 issue3Frequency of canine heartworm disease (D. immitis) in some regions of São Paulo State using a circulating antigen indicator testT-2 toxin and disturbed endochondral bone growth in broiler chicken author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Arquivo Brasileiro de Medicina Veterinária e Zootecnia

Print version ISSN 0102-0935On-line version ISSN 1678-4162

Arq. Bras. Med. Vet. Zootec. vol.53 no.3 Belo Horizonte June 2001 

A new methodology for the improvement of diagnostic immunohistochemistry in canine veterinary pathology: automated system using human monoclonal and polyclonal antibodies

[Uma nova metodologia para melhora do diagnóstico imunoistoquímico em patologia veterinária canina: sistema automático usando anticorpos humanos monoclonais e policlonais]


G.D. Cassali1, P. Silva2, A. Rêma2, F. Gärtner2, H. Gobbi3, W.L. Tafuri1, F.C. Schmitt2

1Laboratório de Patologia Comparada – Dep. Patologia Geral – ICB/UFMG
Caixa Postal 486
31270-901 – Belo Horizonte, MG

2Instituto de Patologia e Imunologia Molecular da Universidade do Porto (IPATIMUP), Porto – Portugal
Dep. de Anat. Patológica, Faculdade de Medicina da UFMG


Recebido para publicação, após modificações, em 16 de Janeiro de 2001




The authors describe their experience with an automated immunohistochemical system applied to canine tissue samples. Twenty human cellular markers specific monoclonal and polyclonal antibodies and two different antigen retrieval methods were used in normal and neoplastic breast tissue, as well as skin samples obtained from female dogs of pure and mixed breeds. The antibodies tested were the most frequently used in human and veterinary medicine studies, employed with diagnostic purposes in breast pathology, as well as in cancer research. Most of them may be used to study other normal and abnormal tissues and included cytokeratins, progesterone receptor, c-erbB2, p53, MIB-1, PCNA, EMA, vimentin, desmin, a-actin, S-100, pan-cadherin, and E-cadherin. The results demonstrated that using an automated staining system it is possible to use different human markers in veterinary pathology. The advantages of automated immunohistochemistry are improved quality, reproducibility, speed, and standardisation.

Key-words: Dog, immunohistochemistry, automated system, mammary gland



Os autores descrevem sua experiência com um sistema automático de imunoistoquímica aplicada à amostras de tecido canino. Foram utilizados 20 anticorpos humanos monoclonais e policlonais e dois diferentes métodos de recuperação antigênica em tecido mamário normal e neoplásico, bem como em amostras de pele obtidas de cadelas. Os anticorpos testados estão entre os mais usados em estudos de medicina humana e veterinária, com finalidade de diagnóstico em patologia mamária, bem como na pesquisa do câncer. Muitos deles podem ser usados para estudar outros tecidos normais e com alterações e incluem citoqueratinas, receptor de progesterona, c-erbB2, p53, MIB-1, PCNA, EMA, vimentina, desmina, a-actina, S-100, pan-caderina e E-caderina. Os resultados demonstraram que usando um sistema automático de imunoistoquímica é possível usar diferentes marcadores humanos em patologia veterinária. As vantagens da imunoistoquímica automatizada são melhora da qualidade, reprodutibilidade, rapidez e padronização.

Palavras-Chave: Cão, imunoistoquímica, sistema automático, glândula mamária




Labelled antibodies that permit visualisation of specific substances in human tissue sections have been used for more than 50 years. The majority of the antibodies to cellular markers were produced in the last decade, increasing the possibilities of the use of immunohistochemistry in diagnostic pathology. The fact that most antibodies were developed against human antigens may limit their use in veterinary pathology.

In spite of this, these antibodies have been used in cases where specificity for antigens is species-independent, such as intermediary filament proteins, progesterone and estrogen receptors, proliferation markers (MIB-1 and PCNA) and some prognostic markers such as tumoral gene supressor p53 and oncogene C-erbB2.

The recent introduction of automated systems in immunohistochemistry has allowed the use of antibodies to human markers to study different veterinary pathological processes.

In this paper, the authors describe their experience concerning standardisation of 20 monoclonal and polyclonal antibodies specific to human cellular antigens (mouse origin) using an automated staining system on canine species material.



Normal and neoplastic mammary tissue, as well as skin samples, were obtained from female dogs of pure and mixed breeds, following surgical resections and autopsies in the Veterinary Hospital, Federal University of Minas Gerais, Brazil, and in Oporto Veterinary Clinics, Oporto, Portugal and selected for immunohistochemistry.

Twenty human cell markers antibodies were tested on 4m m sections of formalin fixed, paraffin embedded dog tissues. Streptavidin-biotin-peroxidase technique was used with antigen retrieval. Primary antibodies (Table 1) were applied using an automated slide staining system (Lab Vision Autostainer Model LV-1). Human breast cancer sections were used as positive controls. Negative controls were performed replacing primary antibodies with a non-immune serum. To test their efficacy three tests with different dilutions were initially performed for each antibody: the suggested dilution for human species, one below that limit (half of the concentration) and one above that limit (double of the concentration).



The antibodies were selected among those most used in human and veterinary medicine studies, with diagnostic purposes in breast pathology, as well as in cancer research. Most of them may be used to study other normal and abnormal tissues with the following specificity: cytokeratin (intermediate filament, epithelial marker): AE1/AE3, CAM 5.2 (Fig. 1A), MNF 116; progesterone receptor, c-erbB2, and p53 (prognostic markers); proliferation markers (stain protein involved on cell cycle): MIB-1 (Fig. 1B) and PCNA; EMA (epithelial membrane antigen, for epithelial differentiation); desmin (intermediate filament of muscle cells); a-actin (interacts with a-actin of smooth muscle); S-100 (protein found in several normal cells - melanocyte, histiocyte sub-population, chondrocyte, adipocyte, skeletal cardiac muscle, Schwann cells - can be expressed on correspondent tumors) (Fig. 1C); cadherin (adhesion molecules of epithelial cells): pan-cadherin, E-cadherin; Vimentin (intermediate filament of mesenquimal cells) (Fig. 1D).



Immunohistochemistry was performed in a open system - LabVision Autostainer – in the immunohistochemistry laboratory of IPATIMUP (Instituto de Patologia e Imunologia Molecular da Universidade do Porto). The system is composed by an automated slide processing which allows the use of non supplied reagents.

A universal staining kit applicable to monoclonal or polyclonal antibodies (UltraVision Large Volume Detection System Anti-Polyvalent, HRP – LabVision) was used. Reagents of this kit were a labeled streptavidin-biotin detection system. This technique involved the sequential incubation of the specimen with an unconjugated primary antibody specific to the target antigen, a biotinylated secondary antibody that reacts with the primary antibody, an enzyme-labeled streptavidin, and a substrate-chromogen (DAB Substrate System-LabVision). In spite of this, the same reagents and protocols used for manual staining may be used in the automated system. No additional material was necessary, such as disposable pipette tips, absorbent tissue etc.

The autostainer holds 48 slides, and it can be programmed identically or individually using any combination of primary antibody, detection system, staining time, reagent volumes and reagent delivery positions (3 per slide). A fully automated system must accommodate for a variety of tissue forms, including snap-frozen and paraffin-embedded sections, cytospin preparations and cytology smears.

The software provides rapid programming, and usually takes 10 to 15 minutes, for each step, for analyzing 48 slides. Up to 64 different reagents may be used for each batch.

A regular batch can run 48 slides and the incubation time for the primary antibody is 20 to 30 minutes. The complete run takes 2 to 3 hours enabling several runs per day. In order to simplify routine, all tests were performed with an incubation time of 30 minutes for primary antibodies and 10 minutes for all other reagents (DAB included).



Comparing with the suggested dilution for human tissues best results were observed for more concentrated primary antibodies; the same occurred when overnight incubation was performed.

The characteristics of the antibodies used in this study are listed in Table 1. Antigen retrieval was not necessary for 7 out of the 20 antibodies used. Enzymatic pre-treatment was performed for three antibodies, and heat induced epitope retrieval was used for the remaining antibodies, with a cytrate commercial solution (Retrieval Solutionä , Dako) for antigen unmasking, in a water-bath at 98ºC during 20 minutes. Enzymatic pre-treatment was performed using 0.4% pepsin for 30 minutes at 37ºC.

For nuclear antibodies (MIB-1, progesterone receptors, p53, PCNA) the best results were obtained using a brief counterstain (3 seconds) with hematoxilin. Extended time lead to misinterpretation concerning MIB-1 in which positivity was not dependent on staining intensity.

The concentration of Tween 20 recommended by the manufacturer (0.2%) did not permit nuclear markers stain; after several tests an ideal concentration of 0,1% was achieved. This concentration did not interfere neither with the staining nor with the normal functioning of the automated system.



The use of human cellular markers specific antibodies to study other animal tissues is related to the diversity of species in veterinary medicine and lack of related species-specific cellular specific antibodies (Nouwen et al., 1990). Recently, several studies concerning female dog mammary glands were conducted using anti-human cell marker antibodies (Cassali, 1999; Gärtner et al, 1999; Mottolese et al., 1994). Mottolese et al. (1994) evaluated whether antigenic cross-reactivity occurs between canine and human breast tumors. They used a panel of 6 monoclonal antibodies recognizing distinct macromolecules that have shown previously to be expressed in benign and malignant human breast lesions. Their immunohistochemical analysis demonstrated very similar antigenic changes, suggesting that in both species, common pathogenetic pathways may characterize tumorigenesis of breast epithelium.

In human species the impact of immunohistochemistry has been most effective in surgical pathology in the diagnosis of "undifferentiated malignant neoplasm" (UMN). The anaplastic neoplasm, a tumor which the differential diagnosis includes lymphoma, carcinoma, melanoma or even a a high-grade carcinoma, is a common occurrence in the activity of the medical (Battifora, 1999) and veterinary surgical pathologist. Antibodies used in the interpretation of UMN include intermediate filaments (vimentin and cytokeratins) and antibodies to melanoma markers. The assessment of immunohistochemistry in human and veterinary infectious and parasitary diseases is also very important; as infectious agents possess antigens it is possible, based on their detection on cells and tissues, to make a ethiologic diagnosis. Nowadays, there are efficient antibodies to detect several virus, bacteria, fungi, and protozoa. The advent of immunohistochemistry allowed to a great improvement on scientific knowledge and also in the diagnosis of a variety of lesions and infectious diseases with a practical impact.

Most of cellular markers specific antibodies developed in the last years were prepared to recognize human antigens. The use of immunohistochemistry in veterinary pathology has had some limitations due to the lack of species specific antibodies to cellular markers in animals. In this paper it is demonstrated that using an automated system it is possible to use human markers in veterinary pathology.

The greatest advantage of automated immunohistochemistry is improved quality, reproducibility, speed and standardisation. It consequently leads to the achievement of new levels of quality with reduced labour, reagent costs, and also permits the improvement of intra and inter-laboratory reproducibility of assays (Grogan, 1992; Markin, 1992).

The autostainer system tested by us was designed to perform the immunostaining in a very similar way to the manual method, enabling the adaptation of technology from the established manual methods without significant changes related to the reagents used.

The high kinetic efficacy of the instrument through combined heating, mixing reagents, and evaporation control allows optimised signal formation with the minimum amount of antibody, and ensures an uniform environment from batch to batch, and from one laboratory to another, with highly reproducible results. Although most researchers think about cost savings, the consistency of staining patterns is perhaps the most important contribution of automation when applied to veterinary pathology. The automated system does not solve problems of tissue processing before immunohistochemistry that compromises antigen detection, and its goal is not to replace the technician, but improving technique performance.



Supported in part by FAPEMIG (Grant n° CBS838/96), CAPES/ICCTI (Grant n° 035/98) and CNPq.



BATTIFORA, H. Strategies for the immunohistochemical approach to the study of undifferentiated neoplasms. In: Congresso Brasileiro de Patologia, 22, 1999, Curitiba. Curso de Imunohistoquímica, 1999. p.4-13.        [ Links ]

CASSALI, G.D. Secretory carcinoma of the canine mammary gland. Vet. Pathol., v.36, p.601-603, 1999.        [ Links ]

GÄRTNER, F., GERALDES, M, CASSALI, G.D. et al. DNA measurement and immunohistochemical characterization of epithelial and mesenchymal cells in canine mixed mammary tumours: putative evidence for a common histogenesis. Vet. J., v.158, p.39-47, 1999.        [ Links ]

GROGAN, T.M. Automated immunohistochemical analysis. Am. J. Clin. Path. (Suppl. 1), p.S35-38, 1992.        [ Links ]

MARKIN, R.S. Laboratory automation systems. Am. J. Clin. Pathol., v.98, (Suppl 1), p.S3-10, 1992.        [ Links ]

MOTTOLESE, M., MORELLI, L., AGRIMI, U. et al. Spontaneous canine mammary tumors. Lab. Invest., v.71, p.182-187, 1994.        [ Links ]

NOUWEN, E.J., DAUWE, S., DE BROC, M.E. Occurrence of the mucinous differentiation antigen CA 125 in genital tract and conduticve airway epithelia of diverse mammalian species (rabbit, dog, monkey). Differentiation, v.45, p.192-198, 1990.        [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License