Correlation of cell cycle regulatory proteins ( p 53 and p 16 ink 4 a ) and bcl-2 oncoprotein with mitotic index and thickness of primary cutaneous malignant melanoma

Th e purpose of the study was to determine the frequency of expression p and pINKa proteins and bcl- oncoprotein in malignant skin melanoma and to determine their correlation with the proliferative index and tumor thickness. Th e study involved  patients:  () male and  () female. Mitotic index showed a correlation with p protein expression, a negative correlation with pINKa protein expression. Statistically signifi cant correlations were determined between the Breslow tumor thickness, Clark invasion level and p protein expression, as well as Breslow tumor thickness and bcl- oncoprotein expression (p<.), whereas there was no correlation between the pINKa protein expression and melanoma thicknes and Clark invasion level. Overexpression p protein and bcl- oncoprotein, with the loss pINKa protein of expression in the nodular melanoma, confi rms a frequent loss of function of these tumor suppressor gene and oncogene, and indicates a vertical tumor growth phase. Th e loss of tumor suppression function the p protein and bcl- oncoprotein overexpression in cutaneous melanoma correlates with larger tumor thickness, whereas the overexpression of mutated p protein and loss pINKa protein of expression indicate a higher proliferative tumour potential. Th erefore, these evaluated proteins may be the aggressive biological tumour activity markers. ©  Association of Basic Medical Sciences of FBIH. All rights reserved


Introduction
Malignant melanoma of the skin (MMS) is a melanocyte system tumor which is produced by the malignant melanocyte transformation generated by the neural crest [, , ].MMS is characterized by high malignant potential and high mortality rate, as well as a rapid occurrence rate in the late years, especially in women [].Multistep tumorigenesis is an evolving process that involves the inactivation of tumor suppressor genes, activation of oncogenes and defects of the reparatory genes which can induce inadequate gene reparation [].Th e most common oncogene defect occurs at the control site G/S which mediates the cell S phase onset [-].Melanoma is not an exception in the tumor biology and its occurrence is predominantly the result of the gene mutations accumulation whose key role is the regulation of cellular proliferation, diff erentiation, apoptosis or some other pathways of cellular death [, ].Many studies have shown a certain amount of damage and loss of tumor suppressor genes p and p with MMS patients, which is why they are considered to be responsible for the tumor carcinogenesis process [-].The tumor suppressor gene p has a central role in the cell cycle and apoptosis regulation.Th e mutation of the p gene, one of the most common genetic defects in human tumors, generates defects in the control site of the cell cycle and genetic instability of some cancerous cells [, , , -].Th e protein p, the product of cyclin-dependent kinase Na gene, is also known as p INKa, p-MTS or cyclin-dependent kinase /CDK inhibitor [, , , ].Th e frequent mutations in melanoma are on the CDKNA locus (typical target is p INK ) which suggests a key role for this protein in cell cycle control management in melanocytes [-].It has been observed that the p INKa protein expression in melanocytic lesions decreases from benign to malignant and metastatic lesions [].A bcl- oncoprotein overexpression can stop the apoptosis process which induces cell proliferation.Bcl- oncoprotein expression can be noticed in both benign and malignant melanocyte tumors, suggesting that it is not a key role in the malignant melanocyte transformation [, ].The purpose of this study was to determine the status of the tumor suppression proteins expression (p and p inka ) and bcl- oncoprotein in the primary ma-

Materials and Methods
The study involved  samples of the primary malignant melanoma of the skin.Th e tumors were surgically removed using the elliptic skin incision (excisional biopsy).

Immunohistochemical analysis
Th e detection of the target markers -antigen, was performed using commercially available monoclonal anti-human antibodies (Table .) and highly sensitive and specifi cally marked streptavidin-biotin complex immunoassays (DAKO LSAB + -HRP kit).Th e chromogen was .'-diaminobenzidin (DAB), and the slides were lightly counterstained with Mayer hematoxylin (Merck, Germany).All the reactants were produced by the DAKO Company (Glostrup, Denmark).During the tissue staining, the positive and negative control samples were simultaneously stained in order to confi rm the specifi city and quality of the immunohistochemical analysis.
The mutated p and p and bcl- protein expression evaluation was performed based on the intensity of the immunohistochemical nuclear (p and p

Morphometric analysis
Tumor thickness measurements Th e tumor thickness was determined according to Breslow using an ocular micrometer and suitable computer software on a microscope with a digital camera.Th e thickness was determined at the widest tumor area starting from the upper line of the granulated epidermal layer or the bottom line of ulceration, vertically into the depth, to the last tumor coast or individual tumor cell.Th e thickness was measured in millimeters Mitotic activity The number of mitoses (mitotic index, MI) was determined in the tumor areas with the largest number of the mitotic figures at  fields of the highest microscopic amplification (x for each sample), (the total mitoses number/ HighPowerField -HPF).The MI was classified in the following manner: low index - mitotic figures/ HPF; high index >  mitotic figures/HPF.

Statistical analysis
The data obtained during the research was filed in a specially created data base.The results were statistically processed using the descriptive and analytical statistic methods based on the: mean value (x), standard deviation (SD), standard error (SE), Student's t-test in some groups in order to evaluate the statistical importance of mean value differences, MANOVA and Pearson's correlation.

Results
The mean age of patients was .±.years; the youngest patient was  years old, the oldest  years.There were  (.) male and  (.) female patients.There were two types of the tumor based on the cellu-  However, there were statistically significant correlations between the mitotic number and p positive cell percentage as well as the mitotic number and p reaction intensity.Moreover, there was a statistically significant negative correlation between the mitotic number and p INKa positive cells percentage and between the mitotic number and p INKa reaction intensity (p < .); whereas there was no significant difference between the mitotic number and bcl- positive cells percentage, and between the mitotic number and bcl- reaction intensity (Table ).Statistically signifi cant correlations were observed between the Breslow tumor thickness and p positive cell percentage and between the Breslow tumor thickness and bcl- positive cells percentage.Moreover, there were some significant between the Clark tumour invasion levels and p positive cells percentage and between Clark tumour invasion level and p reaction intensity (p < .).Other correlations were not signifi cantly relevant (Table ).

Discussion
MMS is one of the most studied tumors due to the vagueness of the tumor's etiological and biological aspects [, , , , ].In our research, the largest amount of the NM () consisted of epithelioid cells only, and then histologically combined tumors, epithelioid and spindle cells, whereas the smallest amount of the nodular melanoma consisted of spindle type cells.However, the largest amount of the SSM () consisted of epithelium cells exclusively.In general, most types of melanoma exhibit a combined histological image, with a particular dominant cell type, which was the case with our subjects -the dominant form being the epithelioid cell type [].Th e most aff ected anatomical area in women was the lower extremites area, lower leg, almost  of the cases, whereas in men it was the trunk , the back  of the cases.Generally speaking, the most common melanoma occurrence in men is on the trunk, in women on the lower extremites, which is confi rmed by the data from our research [, ].Tumor ulceration as an independent prognosis factor may also characterise cutaneous melanomas with a vertical growth phase in the late disease stage.Th e ulceration was a common characteristics of the NM in our group, and was observed in  () of the cases, while in the SSM group it was present in only  cases ().Our results confi rm a signifi cant diff erence between the NM and SSM compared to the present ulceration which other studies also confi rm [, , ].
In our group, the NM patients had a higher pathological stage (pT/WHO) compared to the SSM.This histopathological parameter also showed a statistically signifi cant diff erence [].Although it is known that the cutaneous melanoma is a cellular growth and development defi ciency, some genetic mechanisms responsible for its initiation and progression remain unclear [, , ].Loss of function of p tumor sup-pressor gene and mutation p tumour suppressor gene in malignant melanoma suggest the importance of their roles in the tumor pathogenesis [, , , ].Th e results of our study show a high level of nuclear expression of the p protein mutated form with thick skin melanoma, which is also confi rmed by the data from other studies [, ].Th e correlation between p protein overexpression and MI suggests a high proliferative tumor potential.Our data is confi rmed by a research by Soares de Sa et al. [] who believe that an overexpression of the p protein in correlation with a high mitotic rate probably points to the diminishing of the S-G control site in high proliferative melanoma.
We have encountered a frequent p INKa protein expression loss in patients with NM, as opposed to the SSM patients who have an overexpression of this protein.Th e permanent loss of the protein expression in NM suggests the lack of its function, late stage, and a possible role in the tumor progression, which is confi rmed by the results from other authors [, , , , ].Th e p INKa protein expression in melanocyte lesions decreases from benign to malignant and metastatic lesions [].We have observed a statistically relevant negative correlation between MI and p INKa protein expression, while Talve et al. [] do not show a correlation between p protein and MI in a group of  malignant melanoma.There is, however, a correlation between the number of Ki- positive cells and p INKa protein expression loss.Mihic-Probst et al. [] have also found a signifi cantly higher proliferative activity with the melanoma of low p protein expression.Th e diff erence in these correlations may be due to a diff erent p INKa protein expression which can lead to the existence of various p INKa protein mutations and their impact on the primary MMS proliferative activity.
Our study shows a high level of bcl- oncoprotein expression in MMS, which can also be confi rmed by the other research data [, , , ].Th is protein is expressed with both benign and malignant lesions, has no diagnostic signifi cance and is not a key aspect in the malignant melanocyte transformation [].Leiter et al. [] have found an increased bcl- and bcl- related genes bcl-xLm bcl-xS and Bax expression in malignant melanoma during the disease progression as the result of the apoptosis inhibition and higher tumor malignant potential.We didn't fi nd a correlation between bcl- oncoprotein expression and MI which suggests that this protein has no proliferative role in malignant melanoma of the skin.[].
Our study, as well as Yamamoto et al. [], also shows a high mutated p form expression in MMS and a positive correlation with tumor thickness.Compared to our results, other researches show no correlation between the p mutation overexpression and tumor size [].Although our study shows a high level of mutated p protein expression in MMS, the possible diagnostic signifi cance of this marker may need to be assessed with a larger subject group.Our results show that there is no correlation between the Breslow tumor thickness and Clark invasion level degree compared to p protein expression, which is confi rmed by other authors [, , ].Moreover, Straume et al. [] and Reed et al. [] in their studies have found a signifi cant connection of the p protein nuclear expression loss and vascular tumor invasion, but not so with Breslow tumor thickness or Clark invasion level.In contrast to our results and the studies mentioned above, some other authors suggest that the level of melanoma invasion after Clark is greatly increased with p negative melanoma [].We found a high level of bcl- oncoprotein expression with MMS and a positive correlation with Breslow tumor thickness, which may occur as a result of apoptosis inhibition which favorizes tumor cells survival and enables disease progression [, , ].In addition, we found a combined p and bcl- protein overexpression with cutaneous melanoma patients, which also causes a higher malignant tumor potential [].

Conclusion
Overexpression mutated p protein and bcl- oncoprotein, with the loss of expression p protein in the nodular melanoma, as shown by our study, confi rms a common loss of function these tumor suppressor genes and oncogenes, and indicates a vertical tumor growth phase.Our results show that the loss of tumor suppression function and the p protein and bcl- oncoprotein overexpression in malignant of melanoma of the skin correlate with larger tumor thickness value, whereas the overexpression mutated p protein and the loss of expression p INKa protein indicate a higher proliferative tumour potential.Th erefore, the evaluated proteins may serve as the aggressive biological tumour activity markers.

Declaration of interest
Authors do not have any commercial affi liations, or potential confl icts of interest associated with this work submitted for publication.
BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2010; 10 (4): 277-281 MILOŠ KOSTOV ET AL.: CORRELATION OF CELL CYCLE REGULATORY PROTEINS P53 AND P16 INK4A  AND BCL2 ONCOPROTEIN WITH MITOTIC INDEX AND THICKNESS OF PRIMARY CUTANEOUS MALIGNANT MELANOMA lignant melanoma of the skin and to analyse their correlation with the mitotic index and tumor thickness.
INKa ) staining and cytoplasmic staining (p INKa i Bcl-) and immunoreactive tumor cell percentage.Th e positive immunohistochemical reaction was observed as a light and dark brown colouring of the tumor nuclei.Th e p tumor cells were rarely visible -cytoplasmic colouring of light brown.Th e semiquantitative analysis is calculated as the percentage of the positive cells in relation to the total tumor cells in the fi eld.Immunohistochemical staining intensity determine as (+ = mild; + = moderate; + = intensive) and the tumor positive cell percentage (< no immunoreactive cells; - some positive cells; - many positive cells; > most positive cells).
The p protein mutated form expression was found in  () MMS.All  () NM showed a positive reaction to this protein, while in the SSM it was found in  () cases.A positive nuclear reaction was found in over  tumor cells in  () NM, moderate staining intensity (Figure), while SSM showed a positive reaction mild intensity of nuclear staining (percentage was - of tumor cells).Statistically relevant differences between NM and SSM were found for the percentage of p positive cells and p intensity of the immuno reaction (p<.).The loss of pINKa protein expression was found in  () MMS.A negative immuno reaction to p INKa protein was found in  () NM cases and in  () SSM cases.On the other hand,  () SSM showed a positive p INKa protein expression,  of tumour cells were immunoreactive, and the staining intensity was moderate and intensive.The P INKa immunoreactive tumour cells MMS also showed the cytoplasmatic light brown staining (Figure ).Th e immunoreaction intensity for the p INKa protein varied due to the different tumor segments; a moderate reaction was found in the upper parts of the tumor.Statistically relevant diff erences between the NM and SSM were found in the percentage of the p INKa positive cells (p<.) and p INKa reaction intensity (p<.).The bcl- oncoprotein expression was found in  () MMS.A positive immunoreactivity to this protein was found in  () NM and  () SSM.With  () NM, an immunoreaction was noticed in over  of the tumor cells, with signifi cant staining intensity (Figure ).A positive immunoreaction was observed in  () SSM with over  of the tumor cells of the moderate cytoplasmic staining level.There were no significant statistical differences between the NM and SSM for the percentage of the bcl- positive cells and bcl- immunoreaction intensity.

FIGURE 4 .
FIGURE 4. Diff use nuclear and cytoplasmic immunoreactivity for p16 protein of the tumor cells malignant melanoma of the skin, (IH, x20).

TABLE 1 .
Primary antibodies review used in the immunoassays lar type: in thirty-three patients (.) melanomas were nodular type (nodular melanoma, NM), (Figure ), and  (.) melanomas were of superficially spreading type (superficially spreading melanoma, SSM), (Figure ).According to the patient's sex, the most common MMS anatomical localizations in women were the lower extremities, lower legs - patients (.); the most common anatomical localization in men was the back - cases ().Statistically significant differences between NM and SSM were found for the mean age value (p<.),Breslow tumor thickness, invasion level by Clark, WHO tumor stage, ulceration, mitosis number (p<.), histological cellular tumor type (p<.) and maximum tumor diameter (p<.),whereas the sex variable was not statistically significant.
After the tumor thickness, MI is the second statistically relevant prognostic parameter for MMS.The melanoma prognosis is good if the mitotic index is low (<  mitoses/ mm  ), and bad if the MI is high (>  mitoses/mm  ).Many multivariate analyses have shown a relation of MI with tumor thickness in a way that mitosis remains a significant independent variable for the primary cutaneous melanoma

TABLE 2 .
Correlation of cell cycle regulation proteins (p53 and p16INK4a) and bcl-2 oncoprotein with mitotic index and Breslow thickness and Clark level MMS * statistically signifi cant correlation, (p < 0.05)[, , ].Most of the published studies show a correlation of the tumor thickness with other prognostic factors such as ulceration, anatomical localisation, sex and age of patients, inflammatory response, mitotic index, vascular invasion and microscopic satellitosis.Their results show that the tumor thickness is the most powerful prognostic parameter for localised cutaneous melanoma[, ].Both benign melanocyte tumors and malignant melanoma exprimate the p protein.Although the immunohistochemical reaction for this protein has no diagnostic relevance, numerous data have shown that the p protein expression increases in relation to the MMS thickness suggesting the presence of aggressive melanoma forms