Expression of Vimentin, α-SMA, and CD34 in Malignant Mammary Tumors in Dogs

Introduction. In modern oncology, the mammary tumor microenvironment is considered an important factor influencing the formation of the premetastatic niche in metastasis and contributing to development of a response to nutrient scarcity, lack of oxygen or medicines [1][2]. Studies [3–5] have established that mammary tumor stroma predominantly consists of cancer-associated fibroblasts (CAFs) — a heterogeneous group of cells participating in a number of processes in tumor microenvironment by fostering the extracellular matrix remodeling, depositing basement membrane components, and by stimulating angiogenesis, development, progression, invasion, and metastasis of cancer.

This cell pool originates from resident fibroblasts, mesenchymal and hematopoietic stem cells, as well as from endothelial cells (endothelial-mesenchymal transition) and epithelial cells (epithelial-mesenchymal transition) [3][4]. Vimentin, α-SMA and CD34 have been established to be the potential markers of the fibroblastic series cells (including CAFs) and can be used to determine the presence and localization of these cells in the mammary tumors.

Vimentin is an intermediate filament protein expressed in fibroblasts and endothelial cells. It regulates cell migration, differentiation, proliferation, adhesion, and invasion, and modulates the epithelial-mesenchymal transition induced by tumor initiation and development [6].

α-SMA is one of the actin isoforms that stimulates cell motility and is involved in maintaining shape and polarity of cells, as well as in regulating their transcription. Its expression is characteristic of myofibroblasts, known as myofibroblastic CAFs [7][8][9].

CD34 is a biomarker of hematopoietic and non-hematopoietic stem cells, including vascular endothelial progenitor cells, embryonic fibroblasts, and multipotent mesenchymal stromal cells [10][11].

Currently, the problem of expression of tumor marker proteins in the malignant mammary tumor stroma in animals has been insufficiently studied. At the same time, some studies indicate that in the stroma of many invasive malignant tumors with a low degree of differentiation, CD34+ cells are not detected. This is due to the fact that under the influence of soluble factors (especially TGF-β) secreted by tumor cells [12], they differentiate into α-SMA myofibroblasts associated with cancer [13][14].

The aim of the present study is to investigate the expression of fibroblastic differentiation biomarkers vimentin, α-SMA, and CD34 in the stroma of the canine mammary tumor microenvironment to determine their differentiation in CAFs.

Materials and Methods. The study was conducted in the period from 2022 to 2024 at the Scientific Diagnostic and Treatment Veterinary Center, at the Department of Parasitology and Veterinary-Sanitary Inspection, Anatomy and Pathoanatomy named after Professor S.N. Nikolsky of Stavropol State Agrarian University, and at Stavropol veterinary clinics (“Kolibri”, Sole Proprietor Shalamova E.V., “Pirogov Veterinary Center”, Sole Proprietor Zaichenko I.V.). Medical records of the oncological outpatients registered in the VetAIS, VetDesk, and Vetmanager veterinary practice management systems were analysed. The following aspects of the nosological profile were taken into account: sex, age, breed, castration performed, ridge of mammary glands affected, relapses, as well as the results of instrumental and laboratory examinations for having the pathological conditions indicating the presence of distant metastases.

The material for the study was collected during the unilateral/regional mastectomies and biopsies of mammary tumors in female dogs (n=26). For the study, 1 cm³ specimens were incised from the border of the altered regions of the mammary gland and healthy tissue and skin in compliance with the recommendations of the pathologists’ manual “Macroscopic examination of biopsy and surgical material”. After fixation in 10% neutral buffered formalin (“BioVitrum”, Russia) for 48 hours, the specimens were processed and embedded into the paraffin blocks, which were then sliced into 5-μm-thick sections and mounted on the adhesive slides with the PCI coating, and ground, color-coded edges (CITOTEST, China). To determine the localization of vimentin, anti-Vimentin antibodies (SP20), rabbit monoclonal, 1:25–1:50 (Richard-Allan Scientific Co., USA) were used; to determine the localization of α-SMA — anti-α-SMA antibodies (1A4), mouse monoclonal, 1:25–1:50 (Richard-Allan Scientific Co., USA) were used; to determine the localization of CD34 — anti-CD34 antibodies (134M 15), mouse monoclonal, 1:50–1:200 (Cell Marque, USA) were used.

Immunohistochemical studies were performed using a peroxidase polymer-based detection system in compliance with the manufacturer’s standard protocol (Dako, USA) by heat-induced antigen retrieval at 100°C in citrate buffer (pH 6.0) for 10 minutes, followed by nuclei staining with Mayer’s hematoxylin (“Biovitrum”, Russia). For negative control, the primary antibodies were replaced by a diluent (Spring Bioscience, USA).

Microscopic examination of the resulting histological preparations was performed using an Olympus BX45 upright light microscope (Japan) with a built-in C300 photo and video camera and CellSensEntry microscopy software (Japan). Ten digital images were taken from each preparation, taking into account histoarchitecture, for histological analysis at ×400 magnification, covering a HPF of 0.44 mm² of the studied tissue.

To assess the expression of vimentin, α-SMA, and CD34 and the prognostic value of these markers, the method proposed by L.E. Gurevich (2003) was used:

— membranous or normal expression type: uniform distribution of immunoreactivity across the entire cell membrane;

— membrane-reduced expression type: distribution of immunoreactivity only in certain areas of the cell membrane;

— mixed membrane-cytoplasmic expression type: uniform cytoplasmic expression combined with membranous expression;

— fine- and coarse-granular cytoplasmic expression type: expression in the form of granules scattered in the cytoplasm of cells;

— clumpy expression type: expression in the cell cytoplasm in the form of randomly localized large clumps and conglomerates of immunoreactive material;

— absence of immunoreactivity.

To assess the prognostic value of the described expression patterns in the cells, immunoreactivity was assessed according to L.E. Gurevich (2003):

— 0 points;

— Membranous type — 1 point;

— Combination of membranous and granular-cytoplasmic — 2 points;

— Coarse-granular — 3 points.

Additionally, the intensity of expression of the immunoreactive material was assessed visually, taking into account the percentage of active cells and the total area of immunopositive structures in compliance with the recommendations of the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP, 2018):

1. Immunoreactivity is positive (IHC 3+), if complete intense peripheral membranous staining is observed in more than 10% of tumor cells.
2. Immunoreactivity is indeterminate (IHC 2+), if weak to moderate intensity, complete membranous staining is observed in more than 10% of tumor cells.
3. Immunoreactivity is negative (IHC 1+), if incomplete, weak/barely noticeable membranous staining is observed in more than 10% of tumor cells.
4. Immunoreactivity is negative (IHC 0) if no staining is observed or incomplete, weak/barely noticeable membranous staining is observed in 10% or less of the tumor cells.

Research Results. Preliminary, a routine histopathological examination was performed on 26 samples of mammary gland tumors, which revealed the solid carcinoma G3 to be the main histological tumour type (Table 1). Age-related data are presented in Table 2. All samples were obtained from females, two of which were neutered. The analysis revealed that the highest incidence of neoplasms was observed in small-breed dogs — 42.31% of all cases; in medium-breed and in outbred dogs  — 15.385%, each; in large-breed dogs — 26.92%. The mammary gland tumor affected both the left mammary ridge (2^nd mammary gland — 3.16% of the total amount; 5^th mammary gland — 15.79%) and the right mammary ridge (3^rd and 5^th mammary gland — 13.16%, each, 4^th mammary gland — 21.05%). Metastasis to the inguinal lymph nodes was revealed in 30.77% of all cases. One relapse case was identified among 26 patients.

Vimentin expression in canine mammary tumors is shown in Figure 1. Immunoreactive material was uniformly distributed throughout the cytoplasm in the form of coarse, dark-brown granules (score 3). Although, in some cells, it was visualised in perinuclear pattern as single large foci near the nucleus. Immunopositive areas occupied ≥50% of the total area of the material and biomarker expression was strong (+++). Vimentin-expressing cells had the fibroblastic features— stellate cells with the elongated nuclei. In mammary tumors, they were visualized as a dense network in both the interlobular and intralobular mammary stroma, and were localized around parenchyma in large amounts. This, in our opinion, indicates their paracrine effect on mammary tumor development.

High expression of alpha-smooth muscle actin (α-SMA) in canine mammary tumors is shown in Fig. 2. Dark brown immunoreactive material had perinuclear pattern in papillary, solid, and tubular carcinomas, whereas in mixed tumors, it was uniformly distributed throughout the cell cytoplasm (score 3). Immunopositive areas occupied ≥50% of the total area of the material, and biomarker expression was strong (+++). Cells of a fibroblastic phenotype — stellate cells with the elongated nuclei—were present in the form of cords in the interlobular and intralobular mammary stroma.

Expression of CD34 marker in canine mammary tumors is shown in Fig. 3. A small number of immunopositive cells had a scattered pattern of localization, they were detected in the interlobular stroma of mammary tumor. Areas with these cells demonstrated incomplete, barely noticeable expression, and cytoplasmic staining was negative (+) in more than 10% of tumor cells. The cells had round or elongated form, with nuclei replicating the form of the cells. Immunoreactive material was uniformly distributed throughout the cytoplasm and had the finely- and coarsely granular dark brown structure (score 3). Cells were mostly visualised in the areas where the stroma was subject to fibrosis and sclerosis.

Discussion and Conclusion. Based on the conducted research, it is possible to conclude that high expression of vimentin and α-SMA was observed in all neoplasms. As to CD34 marker, its expression in well-differentiated tumors (G1) was low, and in poorly differentiated neoplasms (G3) it was completely absent. Loss of expression of this marker occurred due to the differentiation of the cells in myofibroblasts of the stromal niche of canine mammary tumors. Our data confirm the results of studies [12][13][14] indicating that during the development and differentiation of malignant mammary tumors, there occurs a loss of CD34 expression and an increase of α-SMA expression.

Thus, immunohistochemical studies have shown that different types of canine mammary gland tumors have a cell line with a specific biomarker phenotype indicating their affiliation with CAFs. The data obtained by us can be used as basic applied data in the study of carcinogenesis in veterinary oncology.