A comparative study of recombinant and native frutalin binding to human prostate tissues

Abstract

Background: Numerous studies indicate that cancer cells present an aberrant glycosylation pattern that can be detected by lectin histochemistry. Lectins have shown the ability to recognise these modifications in several carcinomas, namely in the prostate carcinoma, one of the most lethal diseases in man. Thus, the aim of this work was to investigate if the alpha-D-galactose-binding plant lectin frutalin is able to detect such changes in the referred carcinoma. Frutalin was obtained from different sources namely, its natural source (plant origin) and a recombinant source (Pichia expression system). Finally, the results obtained with the two lectins were compared and their potential use as prostate tumour biomarkers was discussed.

Results: The binding of recombinant and native frutalin to specific glycoconjugates expressed in human prostate tissues was assessed by using an immuhistochemical technique. A total of 20 cases of prostate carcinoma and 25 cases of benign prostate hyperplasia were studied. Lectins bound directly to the tissues and anti-frutalin polyclonal antibody was used as the bridge to react with the complex biotinilated anti-rabbit IgG plus streptavidin-conjugated peroxidase. DAB was used as visual indicator to specifically localise the binding of the lectins to the tissues. Both lectins bound to the cells cytoplasm of the prostate carcinoma glands. The binding intensity of native frutalin was stronger in the neoplasic cells than in hyperplasic cells; however no significant statistical correlation could be found (P = 0.051). On the other hand, recombinant frutalin bound exclusively to the neoplasic cells and a significant positive statistical correlation was obtained (P < 0.00001). However, recombinant frutalin did not recognise all malignant prostate cases and, when positive, the binding to those tissues was heterogeneous.

Conclusion: Native and recombinant frutalin yielded different binding responses in the prostate tissues due to their differences in carbohydrate-binding affinities. Also, this study shows that both lectins may be used as histochemical biomarkers for the prostate cancer. Moreover, the successful use of a recombinant lectin in immunohistochemical studies of prostate cancer was for the first time demonstrated, highlighting the advantages of using recombinant systems in the preparation of pure lectin samples for diagnostic purpose.

Figure 1

Detection of frutalin specific glycoconjugates expressed in sections of human prostate tissues by immunohistochemistry. Samples of formalin-fixed, paraffin-embedded tissues from benign prostate hyperplasia (A, B) and prostate carcinoma (C to H) were immunostained with recombinant frutalin (A, C, E, G) and native frutalin (B, F, H). A negative control proves that the binding responses observed in the prostate carcinoma tissues corresponded to the lectins (D). No detectable recombinant frutalin binding was observed in hyperplasic cells (A), while weak staining was detected for native frutalin (B). Both lectins bound to cytoplasm cells of carcinoma glands (C, E, F, G and H, shown by the white arrows). The binding of recombinant frutalin to the carcinoma tissues was heterogeneous, with closed glands having strong and weak staining (C). The native frutalin staining was strong and homogeneous in carcinoma glands (F and H). Recombinant frutalin was able to specifically recognise carcinoma cells in middle of a benign lesion (E and G).