MY WORK AS AN UNDERGRADUATE RESEARCHER
In this section you will be able to see the work in which I was able to collaborate throughout the 2020-2021 academic year in conjunction with Dr. Santiago and his second year Ph.D. student Sheila Valle.
Evaluation of Novel Proteomic Prostate Cancer Biomarker Panel to Distinguish Indolent Versus Aggressive Tumors
Objectives:
The objective for this project was to evaluate the stage predictive potential of a novel proteomic biomarker panel composed of the proteins E-cadherin, N-cadherin and Beta catenin, which are involved in epithelial to mesenchymal transition (EMT) and were selected for their relevance to prostate cancer progression.
Justification:
Prostate cancer is among the most common cancer types in men, and the incidence rates continue to rise [1]. In most cases, the cancer develops in a slow manner and does not present a significant initial risk for the patient, this form of PCa is known as indolent tumors [2], and is approached with active surveillance to monitor the development of the tumor. However, some tumors can be present as aggressive forms. In this case, when considering the available active treatment options, like radiotherapy or radical prostatectomy, it is clear that they drastically diminish quality of life for the patient. For this reason, stratifying the patients based on indolence or aggressiveness is of clinical importance. This stratification makes the difference between an over-treatment that produces permanent damage to a patient with indolent tumors or a correctly managed aggressive PCa patient. This sets the foundation for the needed development of clinically relevant biomarkers that can discriminate between indolent or aggressive type of tumors, to provide solid guidance towards the correct treatment options. Many biomarkers have been developed so far, ranging from genomic to proteomic, some have showed better results than others [3]. In addition, when taking into consideration the available and clinically used genomic biomarkers there are important limiting factors that have to be viewed, one being the possible inaccuracy of gene expression interpretation due to the fact that the expression of a gene does not necessarily correlate with the amount of active protein present in the sample. For this reason, the development of proteomic biomarkers promises to bypass this limitation and provide a more robust representation of the patient’s risk. For instance, the protein panel evaluated in this project is composed of three proteins: N- cadherin, E-cadherin and Beta catenin; all involved in cell-to-cell adhesion between normal prostatic cells. In PCa, the loss of E-cadherin expression is associated with metastatic progression, whereas N-cadherin acquisition appears to be a critical step in epithelial cancer metastasis and disease progression. This phenomenon is known as the “cadherin switch” and is associated with increased migration, invasive behavior, and poor prognosis. At the same time, B-catenin also helps in the maintenance of the cells to the epithelium in normal prostatic cells, same as E-cadherin, when there is metastasis the expression of this protein is lost.
Equipment and protocols:
Protein evaluation
Firstly, the basis for selection of the three proteins that compose the novel proteomic biomarker panel studied can be seen in Table 1. To proceed with the evaluation of the novel proteomic biomarker panel composed of the proteins E-cadherin, N-cadherin and Beta catenin, a total of 116 patients with PCa adenocarcinoma were evaluated on tissue microarrays (TMAs). From each patient we performed three individual immunohistochemistry (IHC) staining (Figure 1). To perform the IHC, we used antibodies to target the expression of the biomarkers on the tissues.
Table 1. Protein composition of the novel proteomic biomarker panel being evaluated, with their selection basis.
Figure 1. Immunohistochemically stained tissue samples for the three different proteins at 40X magnification. The samples shown are from a PCa patient with stage 2.
Statistical Analysis
Once all the IHC staining images where collected, they were analyzed using the ImageJ software, which provided a scoring based on the expression of the protein (Figure 2). For each scoring level a numerical value was assigned to correlate with the staging of the cancer, as follows: negative=1, low positive=2, positive=3 and high positive=4.
Figure 2. IHC image analysis using ImageJ software. The yellow arrow is highlighting the scoring assigned by the program.
After performing the tissue positive scoring evaluation, the Minitab Statistical software V.19 was utilized to develop a linear regression model to predict cancer staging based on the expression levels of the three biomarkers. In addition, this software was further used for the development of graphical models that presented the correlation between the expression levels of the three proteins individually with certain clinical parameters of the PCa patients. These parameters were: stage, lymph node invasion, tumor size, grade metastases, Gleason score and Gleason grade.
Results:
Novel proteomic biomarker panel prospective evaluation
The linear regression model developed for stage prediction with the IHC analysis can be seen in Figure 3. The model showed a stage predictive potential of 46% when considering the expression of the three proteins. In the same manner, when each protein was evaluated individually within the model; the predictive potential was 39% for Beta catenin, 47% for E-cadherin and 26% for N-cadherin (Figure 4).
Figure 3. Cancer Stage predicting linear regression model based on the biomarkers expression. The percent shown in red indicates the stage predictive potential of the model. The data shown in the table can be viewed fully by clicking here.
Figure 4. Individual stage predictive potential of each protein shown in red with the corresponding protein highlighted in yellow. (The data shown in the tables can be viewed fully by clicking here.
Finally, further analysis of the individual proteins through correlation graphs with clinical parameters (Fig. 5, Fig. 6, Fig. 7), showed that the three proteins had strong correlation with prognostic parameters like cancer staging as well as other important factors. A summary and representation of the results provided through the correlations graphs can be seen in Table 4. The p-value stablished for our data was of 0.05 or less to be considered significant.
Figure 5. Correlation graphs of E-cadherin with the patients clinical parameters.
Figure 6. Correlation graphs of N-cadherin with the patients clinical parameters.
Figure 7. Correlation graphs of Beta catenin with the patients clinical parameters.
Table 4. Result representation of correlation graphs.
**(-)= Inverse correlation, (+)= direct correlation and (x)= no strong correlation.
Conclusions:
After analyzing the results from the proteomic biomarker panel we concluded that the combination of the biomarkers (N-cadherin, N-cadherin, and B-catenin) has the potential to accurately predict 46% of patient staging. To confirm the accuracy of these results it is needed to increase the number of patients to be analyzed to reach 80% power analysis. However, when evaluating the expression of E-cadherin with the patient’s clinical parameters an inverse correlation with stage, lymph nodes metastasis and metastasis to distant sites, was observed. These results indicate that at higher the E-cadherin expression the less association with metastasis. Further, N-cadherin expression showed a correlation with tumor size, grade, metastasis to distant sites, Gleason grade, and Gleason score. These results suggest that at higher N-cadherin expression the patient is more likely to develop metastasis and higher tumor volume. Lastly, B-catenin expression showed a slight negative association with stage. Indicating that at higher B-catenin expression less staging and thus good prognosis for the patient. These results correlate with previous data that point out the importance of these proteins on EMT and cancer progression. Also, our results could be of importance when evaluating biopsies of patients in a clinical setting but needs to be further validated.
The future direction of this investigation is to perform a knockout for the genes that code for N-cadherin, E-cadherin, and B-catenin to evaluate more deeply the metastasis progression when these EMT markers are lost in PCa. Also, we will evaluate some candidates from the GPS test on on PCa tissue microarrays (TMAs) to increase the stage predictive capacity of our biomarker panel.
References:
-
Siegel, R.L., Miller, K.D. and Jemal, A. (2017), Cancer statistics, 2017. CA: A Cancer Journal for Clinicians, 67: 7-30. https://doi.org/10.3322/caac.21387
-
Chistiakov, D. A., Myasoedova, V. A., Grechko, A. V., Melnichenko, A. A., & Orekhov, A. N. (2018). New biomarkers for diagnosis and prognosis of localized prostate cancer. Seminars in cancer biology, 52(Pt 1), 9–16. https://doi.org/10.1016/j.semcancer.2018.01.012
-
Cucchiara, V., Cooperberg, M. R., Dall'Era, M., Lin, D. W., Montorsi, F., Schalken, J. A., & Evans, C. P. (2018). Genomic Markers in Prostate Cancer Decision Making. European urology, 73(4), 572–582. https://doi.org/10.1016/j.eururo.2017.10.036