Introduction

There are breast lesions of variable biological behavior. Among the breast lesions with intimate relationship, the fibroadenoma and invasive duct carcinoma are biphasic lesions sometimes with a similar growth pattern. These tumors characteristically also show greater stromal cellularity (i.e., increased numbers of cells per unit area of tissue) and may behave in a benign or malignant fashion [1].

In general, fibroadenoma arises in the terminal duct lobular unit of the breast. It is the most common benign tumor of the breast and the most common form of breast tumor in women. Fibroadenoma is usually found as single lumps, but about 10–15% women have several lumps that may affect both breasts [2].

About the mammary tumors diagnosis, the immunohistochemistry (IHC) studies for protein expression have become the most attractive routine test due to its cost, convenience, and biological relevance. However, problems as to IHC staining variability and subjectivity in the interpretation of IHC conventional methods have been reported [3].

These findings suggest the necessity for more accurate studies in order to clarify this interpretation issue. Currently, the chemiluminescence biomarkers are applied for validation methods and screening [4].

Chemiluminescence, the generation of light via chemical reaction, has gained widespread use as a detection mode over the last several years. The method offers advantages in sensitivity and reagent stability, detection precise control, and low biohazard [5]. Historically, luminol and isoluminol were the chemiluminescent labels of choice, but they became obsolete by the advantages of acridinium ester (AE) such as high inherent sensitivity, low quenching, simplicity and easy of handling and disposal, and a long shelf-life [6]. Light from AE reaction is detected at 430 nm over a very short time frame, minimizing background noise and improving overall sensitivity exhibiting a linear response over an AE concentration range of more than 4 orders of magnitude, with a detection limit of approximately 5 × 10−19 mol [7]. Chemiluminescence has been used in clinical analysis routine and as a tool in clinical and biomedical research only for analytes detection in solution [8].

Preliminarily, in ours laboratories, we label acridinium ester molecule with a vegetal lectin to be used as auxiliary histochemistry tool to help the clinical–pathological evaluation of infiltrating duct carcinoma, a human mammary tumor of high incidence in Brazil [9]. Here, it is proposed a chemiluminescent approach for biomarker detection in tissues (histochemistry). The glycocode of normal, fibroadenoma, and invasive duct carcinoma tissues were investigated using lectin from Canavalia ensiformis (concanavalin A, Con A) and Arachis hypogaea (peanut agglutinin, PNA) labeled with AE as a model to investigate the feasibility of this methodology. Con A binds molecules containing α-d-mannopyranosyl, α-d-glucopyranosyl, and sterically related residues being the most studied and characterized legume lectin [10]. PNA has specificity for d-galactose and presents multimeric structure with four identical monomers binding to carbohydrates [11].

Materials and Methods

Human Mammary Specimens

Thirty-three samples of tumoral breast diseases were obtained (females, mean age 46.5 years old) and submitted to formalin-fixed and paraffin-embedded. The tumors were diagnosed as fibroadenoma (n = 22), and 11 biopsies diagnosed as invasive duct carcinoma were obtained from the Tissue Bank of the Clinic Hospital at the Federal University of the State of Pernambuco, northeast Brazil. All tissue samples were processed histologically and embedded in paraffin. Hematoxylin and eosin stained sections (4 mm) were examined to confirm the diagnosis. Two normal mammary tissues reduction mammoplasty specimen were collected from a plastic surgery clinic.

Con A and PNA Conjugations with Acridinium Ester

AE (DMAE-NHS/1966-1-53-2/Organic Lab kindly supplied by Dr. H. H. Weetall) was conjugated to Con A and PNA (Sigma-Aldrich, St. Louis, MO, USA) according to Weeks et al. [12]. Briefly, Con A and PNA (500 μL containing 2 mg of protein) were incubated with 15 μL of AE ester solution (0.2 mg diluted in 400 μL of N,N-dimethylformamide) for 1 h at 25 °C under mild stirring. Conjugates (Con A-AE and PNA-AE) were applied to a column of Sephadex G-25 (10 × 1 cm), previously equilibrated with 10 mM phosphate buffer, pH 7.2, and eluted with this buffer. Aliquots (1 mL) were collected and their protein contents were chemiluminescence-established [13]. The Con A-AE and PNA-AE conjugates were also evaluated regarding the maintenance of its carbohydrate recognition property (hemagglutinating activity) using glutaraldehyde-treated rabbit erythrocytes according to Beltrão et al. [14].

Fluorescence Measurements

Intrinsic fluorescence was performed on a spectrofluorometer (JASCO FP-6300, Tokyo, Japan). The fluorescence emission intensity of the Con A-AE and PNA-AE in phosphate buffer at pH 7.2 was measured at 25 °C in a rectangular quartz cuvette with a 1-cm path length. For intrinsic fluorescence measurements, the excitation was at 295 nm and emission was recorded from 305 to 400 nm, using 5-nm band-pass filters for both excitation and emission. Center of spectral mass (CM) were calculated according to equation: \( {\text{CM}} = \sum {I_{\lambda }}{F_{\lambda }}/\sum {F_{\lambda }} \), where F λ stands for the fluorescence emission at wavelength I λ and the summation was carried out over the range of appreciable values of F.

Lectin Histochemistry

Paraffin sections (1.5 × 1.5 × 8.0 10−4 cm) were cut, transferred to tissue cassettes, deparaffinized in xylene (1 × 1 min and 4 × 10 dips) and hydrated in graded alcohols (3 × 100% and 1 × 70%—10 dips each). Afterward, tissue slices were transferred to test tubes and then incubated with Con A-AE and PNA-AE (100 μL–100 μg mL−1) for 2 h at 4 °C, followed by washings (2 × 5 min) with 5 mL of 10 mM phosphate buffer, containing 0.15 M NaCl (phosphate-buffered saline, PBS), pH 7.2, and, then, transferred to polypropylene test tubs with a volume of 50 μL of PBS. Lectin binding inhibition assays were accomplished by incubating each lectin solution with 300 nM methyl-α-d-mannoside (Con A) and d-galactose (PNA) for 30 min at 25 °C prior to their incubation with tissues. The following steps were as described previously for the binding protocol.

Chemiluminescence Measurement

Chemiluminescence measurement was carried out using a luminometer Modulus Single Tube 9200-001 (Turner BioSystems, USA). The emission intensity was determined as relative light units (RLU) with a counting time of 5 s per sample. Duplicate measurements routinely exhibit precision rate lower than 5%.

Statistical Analysis

The software OriginPro 8 (OriginLab Corporation, One Roundhouse Plaza, Northampton, MA 01060 USA) was used for the statistical analysis, and data were expressed as mean ± standard deviation.

Results and Discussion

Con A and PNA lectin labeling with AE were carried out, and the conjugate (Con A-AE and PNA-AE) was collected by Sephadex G-25 chromatography (Fig. 1). The Con A-AE and PNA-AE were found around the 5th fraction (Fig. 1a, b, respectively) and were still capable to recognize the specific carbohydrates (Fig. 1c, d, respectively; hemagglutinating activities).

Fig. 1
figure 1

Con A-acridinium ester conjugate (Con A-AE) and PNA-acridinium ester conjugate (PNA-AE) purification profile from a Sephadex G-25 column (10 × 1 cm). Elution was carried out with 10 mM phosphate buffer, pH 7.2. Fractions (aliquots of 1 mL) were collected and chemiluminescence a Con A-AE and b PNA-AE and hemagglutinating activity c Con A-AE and d PNA-AE were assayed

Full size image

The combination of both lectins (Con A/PNA) to the acridinium ester may affect the atomic fluctuations in proteins, and it is possible to estimate these changes by fluorescence spectroscopy. Con A presents an emission spectrum (Fig. 2a) with a center of spectral mass located at 336 nm, suggesting that some of the four tryptophan residues are buried in the protein core. The intrinsic fluorescence emission did not change with combination of both lectins to the acridinium ester for hydrophobic residues (Fig. 2).

Fig. 2
figure 2

Intrinsic fluorescence emission spectra of a Con A/Con A-AE and b PNA/PNA-AE. Lectins excitation (295 nm) and emission (305–450 nm). Con A and PNA (solid line); Con A-AE and PNA-AE (dash line). Error in fluorescence determination is approximately ± 0.5%, smaller than the size of symbols

Full size image

Figure 3 presents the RLU emissions yielded by the glucose/mannose content recognized by Con A-AE in fibroadenoma (9,916.432 ± 634.387) and IDC (7,407.381 ± 605.879) compared to the normal tissue (1,120.217 ± 111.156). Also, it shows the RLU emissions related to the galactose content recognized by PNA-AE for the fibroadenoma (1,957.955 ± 611.786) and IDC (1,318.297 ± 293.124) compared to the normal tissue (427.568 ± 112.532). For both lectins, the mean values of RLU for fibroadenoma (n = 22) and IDC (n = 11) tissues were significantly different as observed using the non-parametric statistic test of Mann–Whitney U (p < 0.001). Furthermore, Con A-AE and PNA-AE inhibition binding assay using methyl-α-d-mannoside and d-galactose (300 mM), respectively, resulted in a decrease in RLU values indicating that unspecific binding between the conjugates and cell surface molecules did not occur. It is worthwhile to register that only two normal samples (tissue from normal breast submitted to plastic surgery) were used based on previous observation that Con A conjugated to peroxidase did not recognize the unaltered tissues [14]. It was preferred rather than the unaffected tissue borders of the transformed tissues as some studies report. Furthermore, it was easier to collect samples from patients with fibroadenoma (n = 22) than CDI (n = 11). However, this difference was considered in the statistical analysis.

Fig. 3
figure 3

Chemiluminescence in fibroadenoma, CDI, and normal tissues treated with Con A-AE and PNA-AE. RLU (mean of triplicate) of normal (n = 2), fibroadenoma (n = 22), and IDC (n = 11) tissues

Full size image

Luminescence was also evaluated as a function of tissue area (Fig. 4). For Con A-AE, there was a hyperbolic relationship between RLU and the tissue area for fibroadenoma samples ranging from 0.125 to 1.0 cm2 while in IDC samples, this relationship was a linear correlation for the same tissue area range (Fig. 4a), as reported in our preliminary findings [9]. However, PNA-AE showed a hyperbolic relationship between RLU and the tissue area for both fibroadenoma and IDC samples (Fig. 4b).

Fig. 4
figure 4

Relationship between RLU and the area of mammary tissue diagnosed as fibroadenoma (open square) and IDC (closed square). a Con A-AE and b PNA-AE

Full size image

RLU from Con A-AE revealed that light emission was a tissue-area-recognition-limited process depending on the number of saccharide residues on the tissue surface. The reason for the proportional increase in RLU along with tissue area, which was observed for both benign and transformed tissues, is related to the proportional increase in the amount of lectin–carbohydrate complex formation in tissue areas evaluated. For fibroadenoma, the curve resembled the Langmuirian isotherm while for IDC, the increase in RLU was linear (r 2 0.9833) regarding tissue area. These results indicate that tissue area is an important parameter in the chemiluminescent reaction rate under the present conditions when lectins are used as histochemistry probes. It should be emphasized that the proportional relation between the amount of carbohydrates accessible in tissue and their recognition by lectins was abolished when Con A, in the conjugate, was inhibited by its specific sugar. The deviation from linearity observed for fibroadenoma would be a reflect of the glucose/mannose expression and/or disposition on the glycomoiety in the glycoproteins of cell surface in this pathology since tissue area, conjugate concentration, and reaction parameters were similar to IDC. Further studies focusing on the influence of tissue area on the RLU are in progress in our group using mammary tissue and AE conjugated antibodies. The results reflected that the RLU, for both pathologies, is related to the content of specific saccharide recognizable for an appropriate lectin.

The altered accessibility and/or content of glucose/mannose residues in cell membrane glycoproteins in fibroadenoma and IDC has already been observed [14] who used the isoform 1 of Cratylia mollis seed lectin and Con A, both conjugated to horseradish peroxidase, to study biopsies of human breast lesions (fibroadenoma, fibrocystic disease, and IDC) using lectin histochemistry and optic microscopy (qualitative method based in staining intensity). In that case, it was observed that cells of fibroadenoma presented a higher staining pattern than that observed for normal tissue but a lower than IDC.

Fibroadenoma of the breast is a common benign tumor composed of epithelial and stromal components that usually occurs in young women and has been associated with the development of malignancy. On the other hand, the histological diagnosis of carcinoma sometimes could be similar to fibroadenoma. This observation has been based on the criterion that the carcinoma cells were limited to a well-defined fibroadenoma or only focally extended into adjacent stroma or ducts [15].

Tumor progression is related to molecular changes in cells under malignant transformation [16]. The search for new neoplastic biomarkers in association with clinical and traditional histological parameters has been used to identify patients with high-risk and aggressive clinical course of disease [17, 18].

Glycans are central to many fundamental biological processes including cell–cell recognition, detection, and evasion of immune responses, cell attachment and detachment, cell fate, development, and morphogenesis. The importance of glycans for health is exemplified by their leading roles in cancer metastasis and in mediating immune responses [19]. The carbohydrates expression variation in many metabolic processes allows the use of lectins as molecular probes revealing the organization of cell surface glycoconjugates and their changes during aging and diseases [20]. Changes in the profile of cell surface carbohydrates may be involved with the emergence of tumor metastases and angiogenesis [21]. Lectins have been intensely used as histochemistry probes for cell surface saccharide profiling in normal and transformed tissues with specificity as high as the antibody–antigen recognition observed in immunohistochemistry [9, 14, 22, 23]. Here, the glycocode of breast tissue was characterized by using lectins that were revealed by a simple, specific, and sensitivity procedures based on chemiluminescence.

Conclusions

The results of this contribution showed that the tissue analyses using lectin-chemiluminescent-specific probes can be a valuable tool in biomarkers histochemical studies. Lectins, antibodies, antigens, and receptors labeled with acridinium ester can be used as a probe to specifically identify and quantify those biomarkers. This approach adds the notorious advantage of the use of methods based on staining assays. In the model tested using lectin chemiluminescent, it became apparent differences in the glycosylation profile according to the carbohydrates expression (glucose/mannose and d-galactose) into stromal and glandular cells of normal breast tissue, fibroadenoma, and ductal carcinoma.