The Impact of Parkin as a Possible Future Biomarker in Five Solid Pediatric Tumors

Castro EM^1,2,3, Rocha RMCM⁴, Taques GR², Zanoni A², Percicote AP², Pereira IRPD³, Poncio L², Werner B², De Souza SB³, De França N⁴, Benevides APK⁴, De Dalke DBZ⁴, Da Camargo CCVS⁴, Esposito SE⁴, Sotomaior VS⁴, De Noronha L⁴ and Souza CM^1*
*Correspondence: Souza CM, Instituto de Pesquisa Pelé Pequeno Príncipe - Av. Silva Jardim, 1632 – Curitiba-PR-CEP-80250-060, Brazil, Email:

Keywords

Parkin • Choroid plexus tumors • Astrocytoma • Medulloblastoma • Neuroblastoma • Nephroblastoma

Introduction

Childhood cancer can be considered infrequent when compared to adult tumors, and they correspond to nearly 3% of all malignant tumors [1]. The International Classification of Childhood Cancer [2,3] proposed the classification in 11 groups and in this article, we present five of them: Choroid plexus tumors (IIIa), Astrocytomas (IIIb), Meduloblastoma (IIIc); Neuroblastoma (IVa), and Nephroblastomas (Via) which are the main types of solid tumors in the Brazilian pediatric population.

Recent advances in molecular strategies and analytical platforms, including genomics, epigenomics, proteomics, and metabolomics, have identified an increasing number of potential biomarkers. Over the past 50 years, there has been a significant improvement in outcomes for children with cancer, driven mainly by better tumor stratification associated with the plurality of individualized treatment approaches [4]. In this context, the search for representative biomarkers can be challenging, especially concerning pediatric cancer. The use of biomarkers for prognosis has been the goal of many researchers to identify risk populations and help to predict unexpected outcomes.

Parkin is a protein encoded by the Parkin RBR E3 Ubiquitin Protein Ligase (PRKN), a gene discovered two decades ago [5]. The primary function of parkin is involved with proteasomal degradation, inducing ubiquitination of damaged proteins [6-8]. Parkin controls degradation in many cellular processes, such as cell cycle control through the degradation of cyclins (D and E). Other functions have already been described for this molecule such as regulation of cell proliferation and migration, protection against oxidative stress, mitochondrial homeostasis (mitophagy), xenophagy and tumor suppression [9-12]. Parkin also degrades the protein p21. p21 is essential in preventing the accumulation of mutations and the generation of genetic instability. High levels of parkin may result in low levels of p21, and its consequence would be an accumulation of genetic alterations in the tumor cells [13].

The tissue expression of parkin has been shown to have a classic protective role in some neoplastic contexts and this relationship still needs to be better established so that in the future it can be incorporated as a possible prognostic biomarker. Future research using parkin as a possible prognostic biomarker may allow the correct aspect in the association of this molecule in pediatric tumors, and the use of parkin as a prognostic biomarker may early indicate the association of individuals with different types of outcomes. The objective of this study is to investigate the association between parkin tissue expression and the clinical-pathological characteristics in five types of primary pediatric tumors to determine if there is any support for a potential use as a prognostic biomarker.

Materials and Methods

Databases

The cases selected for this study comprised Formalin Fixed Paraffin Embedded (FFPE) samples of five types of pediatric tumors, mostly obtained at the tertiary referral children's hospital (Pequeno Príncipe Children’s Hospital) in South Brazil. Thirty-three patients with choroid plexus tumors (1992 to 2010); one hundred and eight pediatric astrocytomas (2003 to 2015); twenty-nine meduloblastomas (1998 to 2009); ninety neuroblastic tumors (neuroblastoma/ganglioneuroblastoma/ganglioneuroma) (2001 to 2014) and seventy-seven patients diagnosed with nephroblastoma (1994 and 2012) were analyzed. This study brings together five ethics committee approvals. The Human Research Ethics Committee approved this study (Registration number: 3.573.221). Table 1 shows the main characteristics of these samples.

Tissue Microarray (TMA) and Immunohistochemistry (IHC)

From each case, hematoxylin-eosin (H&E) slides were prepared. Two different neoplastic areas free from tissue processing artifacts or necrosis were selected for TMAs block construction. Each TMA block included five to eight cases, with two 3 mm samples from each case. One conventional slide from each TMA was stained using H&E, and the other slides were separated for immunohistochemical study following the procedure described by da Silva-Camargo et al. [14]. The TMA slides were subjected to antigen retrieval and incubated with the monoclonal mouse anti-parkin antibody (1:100, Abcam^®, Cambridge, MA, USA). As a positive control, samples of high-grade colon cancer expressing large quantities of parkin were used. As a negative control, the primary antibody was omitted.

Morphological analysis of protein expression

Parkin immune expression was evaluated by quantitative and semi-quantitative analysis (Tables 2 and 3). Some adaptations were made in the different types of neoplasms addressed. The morphometric evaluation was used for quantitative analysis and the Allred score, initially described in breast carcinomas, for semi-quantitative analysis [15].

The quantitative analysis (morphometry) was based on images obtained by the slide scanner model Axio Scan Z1 (Zeiss; Germany), in high power field (HPF - 40x objective), for digital documentation in Tagged Image File Format (TIFF). After digitization, the software generated images in the format of photomicrography, which was selected and analyzed through color morphometry, using the Image Proplus™ analyzer program (Rockville, MD, USA). Immuno-positive areas in square micrometers of each photomicrograph were compiled and transformed into a percent by HPF.

The semi-quantitative analysis was made by the Allred score, and it was obtained by summing two scores (proportion and intensity of positivity), ranging from 0 to 8. The proportion score is subdivided according to the percentage of stained cells: score 0 - 0% stained cells, score 1 - < 1%, score 2 - 1-10%, score 3 - 11-33%, score 4 - 34-66% and score 5 - > 66%. While the intensity of positivity is evaluated: negative - score 0, weak - score 1, moderate - score 2, and strong - score 3 [15].

The results of the quantitative and semi-quantitative analyzes for all the tumors were compared to clinical-pathological variables of prognostic importance, such as the presence of metastases, local staging, treatment, histological type, presence of anaplasia, risk group, nodal involvement, and outcome.

Statistical analysis

Descriptive analysis was performed with the absolute and relative frequencies of the qualitative variables. The results of quantitative variables were described by medians, minimum values, maximum values. Student's t-test was used to compare the result obtained through quantitative and semi-quantitative analysis to two groups of qualitative variables and the Kruskal-Wallis test for comparison with three groups or more with quantitative variables. Kaplan-Meier’s curves were built to assess median survival with log-rank testing. Data were analyzed using the computer program The R Project for Statistical Computing. Data were analyzed by IBM^® SPSS Statistics v.20.0 software.

Results

Parkin immune histochemical tissue reactions were positive in all specimens analyzed, and tissue expression presented with ample variable intensity and proportion in cytoplasm and nucleus of neoplastic cells (Figure 1).

The comparison of the tissue immune expression of parkin (quantitative and semi-quantitative methods) in Choroid plexus tumor (IIIa Group; n=33; Table 1) showed no significant difference between carcinomas (n=29) and papilloma’s (n=4) (Tables 2 and 3). Outcome data was obtained for 23 of 29 choroid plexus carcinoma: 60.9% (14/23) died, 8.7% were alive for more than 1 year (2/23), 21.7% were alive for more than 5 years (5/23), and 2 (8.7%) were alive for more than 10 years (2/23) (Table 1). The outcome was compared with parkin immune expression (quantitative and semi-quantitative methods) and showed no significant results (Table 4).

In the Astrocytomas (IIIb Group; n=108), parkin tissue immunoexpression, when analyzed by both methods, resulted in a significant (P=0.011 and P=0.037) increased frequency in the low-grade group (Tables 2 and 3).

From the 108 patients included in this study, 67 had available follow-up and outcome data (Table 1). The follow-up ranged from 7.2 months to 12.34 years.

Twenty-three of the Meduloblastomas (IIIc Group; n=29) were classic; three were nodular, two desmoplastic, and one anaplastic. Eleven meduloblastomas presented a standard molecular risk, five low risks, and three high risks (Table 1). In meduloblastomas, there were no significant differences between the tissue expression of parkin (quantitative and semi-quantitative method) and histological variants (Tables 2 and 3). Parkin tissue expression semi-quantitative analysis by Allred score (proportion of positivity) showed lower values in low molecular risk patients’ tumors when compared to high (P=0.01) and standard molecular risk tumors (P=0.01) (Table 4). Parkin's Allred score (semi-quantitative method) showed a significant association (P=0.033) with cell size, considering large cells with intensity 8 (7 to 8), and small cells with intensity 6 (4 to 8) of tissue expression of parkin. No significant results were found when parkin tissue immuno expression was compared to outcome in patients with medulloblastoma (Table 4).

The samples for Neuroblastic tumors (Group IVa; n=90) were previously published by Araujo et al., but clinical-pathological aspects can be seen in Table 1 [16]. Considering parkin tissue expression (quantitative and semi-quantitative method), there were no significant differences between histological variants of neuroblastic tumors (Tables 2 and 3). No significant results were found when parkin tissue immunoexpression was compared to the risk group and staging of patients with neuroblastic tumors (Table 4).

The outcome of neuroblastoma (analysis performed without ganglio neuro blastoma/ganglio neuroma tumors) patients was compared to parkin tissue expression (by Allred score), and surviving disease-free patients showed less tissue immune expression of parkin with P=0.033 (Table 4). A clinical analysis in neuroblastomas (analysis performed without ganglio neuro blastoma/ganglio neuroma tumors) involving parkin tissue expression, outcome (less/more than five years survival), and Shimada group (unfavorable and favorable) was performed. Outcome/survival curves that were performed for the Allred score (cutoff less than 4 or higher/equal 4) regarding parkin tissue expression in patients with neuroblastomas (analysis performed without ganglio neuro blastoma/ganglio neuroma tumors) showed that those who had parkin score higher than 4, had worse prognosis (P=0,031).

Table 1 shows some variables involved with Nephroblastoma (Group VIa; n=77). Further details can be accessed in Percicote et al. [17]. There were no differences in quantitative and semi-quantitative analysis for nephroblastoma variants concerning parkin immuno histochemical tissue expression, even when analysis of the three components was considered separately (Tables 2 and 3). No significant results were found when parkin immune expression was compared to risk group, staging, and outcome/survival of patients with nephroblastoma (Table 4).

Discussion

Parkin is a protein that was discovered around two decades ago in the context of juvenile Parkinson’s disease and proteosomal degradation studies [18].  In the normal brain tissue, parkin is expressed in the cytoplasm of neurons and glia cells [19]. More recently, parkin has been found in the nuclear compartment, where it appears to act as a transcription factor [20]. Today it is known to be a multifunctional protein involved in different types of neoplasia [21-24]. Evidence of parkin involvement in the indirect regulation of different tumorigenic phenotypes could be expected.

In view of these actions that involve parkin we have two possible hypotheses. One of them is that the high expression of parkin, in neoplastic tissue, could be associated with a favorable prognosis in patients with cancer, and the antagonistic idea also has strong biological plausibility [12,14,29]. In our analysis, parkin presents a protective and non-protective function in different pediatrics cancers, and this kind of variation reinforces the need for more studies concerning parkin and its role as a prognostic tool in different cancers.

In pediatric astrocytoma’s, our results may indicate a potential role of parkin as a protective biomarker, since low-grade pediatric astrocytoma’s showed higher parkin tissue expression, and less parkin tissue expression was associated with worse prognosis. Lin et al. established that less parkin expression indicated reduced survival rates [34]. Additionally, lower levels of parkin were associated with increased cell proliferation and poorer prognosis [35,36]. In astrocytoma’s, parkin could be considered as a possible tumor grade marker, being a tool to distinguish these high- and low-grade tumors, especially in the lack of classical morphological characteristics in small specimens. It is known that glioma genesis has different molecular mechanisms in adults and children [37,38] and for this reason we could infer that the association between parkin and the prognostic factors could indicate its classic protective function.

Nephroblastomas followed the same trend seen in astrocytomas. The blastematous component of nephroblastoma, which is the component responsible for a poorer prognosis, showed less expression of parkin in tumors of patients who died when compared to tumors of disease-free surviving patients. In meduloblastomas and neuroblastomas, higher values of parkin expression were associated with worse prognosis. The reduced sample size can be the explanation for the lack of association between the expression of parkin in choroid plexus carcinoma and papilloma’s and prognostic factors and outcome (patient death). Tissue biomarkers have been studied in these neoplasms, but no study has been carried out associating parkin with these tumors.

Genetic alteration in PRKN is common in many human neoplasms and in hereditary Parkinson's disease. In cancer, the PRKN gene is mutated or deleted, with copy number loss being the primary mode of alteration. Gong et al. examined the PRKN mutation pattern across 4934 tumors spanning 11 cancers, and PRKN deletions are the most common in tumors [26]. These data indicate that PRKN is one of the most deleted genes in human cancer. Due to these genetic aspects associated with parkin's participation in carcinogenesis, despite few references reporting expression of this protein in tumor tissue samples, our work aimed to evaluate the immune expression of this protein in five different types of solid pediatric tumors [39,40].

One of the limitations that could be attributed to our study would be the low sample size of two neoplasms (choroid plexus tumor and medulloblastoma). However, this condition reflects the low incidence that these neoplasms present in the Brazilian population of southern Brazil. Another potential limitation is that only the tissue expression of parkin in paraffinized samples (FFPE) was analyzed. The authors are aware that to consider a molecule as a biomarker, other analyzes should be performed. Our future goal will be to implement analyzes that include genetics (polymorphisms) and epigenetics (interference RNA) to complement and extend the deductive and predictive power in relation to parkin.

Conclusion

The tissue expression of parkin in the studied tumors showed an ambiguous action when considering its protective role to the patients. Parkin, acting as a protective prognostic biomarker, has been found in astrocytoma’s and nephroblastomas. In medulloblastomas and neuroblastomas it was observed as a biomarker unassociated with the matter of protection. Parkin is not always protective when it comes to its association with prognostic factors and consequently, outcome, but these results reinforce the importance of knowing parkin tissue behavior to identify new possible prognostic biomarkers.

Funding

No funding.

Conflict of Interest

The authors declare that they have no conflict of interest.

Data Availability

The data that support the findings of this study are available from the corresponding author, [CMS], upon reasonable request.

Acknowledgements

We are grateful to Complexo Pequeno Príncipe, Universidade Federal do Paraná and Pontificia Universidade Católica do Paraná for their support.

References

1. http://www.inca.gov.br/cgi/notatecnica_11012007.pdf
2. Kramárová, Eva and Charles A Stiller. "The International Classification of Childhood Cancer." Int J Cancer 68 (1996): 759-765.
3. https://seer.cancer.gov/iccc/iccc-who2008.html
4. Siegel, Rebecca L and Kathleen D Miller. “Cancer Statistics’’. Cancer J Clin 66 (2016) : 7-30.
5. Shimura, Hideki, Nobutaka Hattori, Shin-ichiro Kubo and Yoshikuni Mizuno, et al. "Familial Parkinson Disease Gene Product, Parkin, is an Ubiquitin-Protein Ligase." Nat Genet 25 (2000): 302-305.
6. Martinez, Aitor, Ugo Mayor and Michael J Clague. "Multi-Story Parkin." Oncotarget 8 (2017): 50327.
7. Panicker, Nikhil, Valina L Dawson and Ted M Dawson. "Activation Mechanisms of the E3 Ubiquitin Ligase Parkin." Biochem J 474 (2017): 3075-3086.
8. Zheng, Xinde and Tony Hunter. "How Phosphoubiquitin Activates Parkin." Cell Res 25(2015): 1087-1088.
9. Drake, Lauren E, Maya Z Springer, Logan P Poole and Casey J Kim, et al. "Expanding Perspectives on the Significance of Mitophagy in Cancer." Sem Cancer Biol 47 (2017): 110-124.
10. Shires, Sarah E, Richard N Kitsis and Åsa B Gustafsson. "Beyond Mitophagy: The Diversity and Complexity of Parkin Function." Circul Res 120 (2017): 1234-1236.
11. Durcan, Thomas M and Edward A Fon. "The Three ‘P’s of Mitophagy: PARKIN, PINK1, and Post-Translational Modifications." Genes Develop 29 (2015): 989-999.
12. Gong, Yongxing, Steven E Schumacher, Wei H Wu and Fanying Tang, et al. "Pan-Cancer Analysis Links PARK2 to BCL-XL-Dependent Control of Apoptosis." Neoplasia 19 (2017): 75-83.
13. Park, Kyung-Ran, Jae Suk Yun, Mi Hee Park and Yu Yeon Jung, et al. "Loss of Parkin Reduces Lung Tumor Development by Blocking p21 Degradation." Plos one 14 (2019): e0217037.
14. da Silva-Camargo, Renata, Claudia Caroline Veloso, Rosimeri Kuhl Svoboda Baldin and Nayanne Louise Costacurta Polli, et al. "Parkin Protein Expression and its Impact on Survival of Patients with Advanced Colorectal Cancer." Cancer Biol Med 15 (2018): 61.
15. Harvey, Jennet M, Gary M Clark, C Kent Osborne and D Craig Allred. "Estrogen Receptor Status by Immunohistochemistry is Superior to the Ligand-Binding Assay for Predicting Response to Adjuvant Endocrine Therapy in Breast Cancer." J Clin Oncol 17 (1999): 1474-1481.
16. Araujo, Deli Grace de B, L Nakao, P Gozzo and CDA Souza. "Expression level of Quiescin Sulfhydryl Oxidase 1 (QSOX1) in Neuroblastomas." Eur J Histochem 58(2014): 2-4.
17. Percicote, Ana Paula, Gabriel Lazaretti Mardegan, Elizabeth Schneider Gugelmim and Sergio Ossamu Ioshii, et al. "Tissue Expression of Retinoic Acid Receptor Alpha and CRABP2 in Metastatic Nephroblastomas." Diag Pathol 13 (2018): 1-7.
18. Shimura, Hideki, Nobutaka Hattori, Shin‐Ichiro Kubo and Mutsuko Yoshikawa, et al. "Immuno-histochemical and Subcellular Localization of Parkin Protein: Absence of Protein in Autosomal Recessive Juvenile Parkinsonism Patients." Ann Neurol 45 (1999): 668-672.
19. Yin, Dong, Seishi Ogawa, Norihiko Kawamata and Patrizia Tunici, et al. "High-Resolution Genomic Copy Number Profiling of Glioblastoma Multi-form by Single Nucleotide Polymorphism DNA Microarray." Mol Cancer Res 7 (2009): 665-677.
20. Zarate‐Lagunes, Martin, Wen‐Jie Gu, Véronique Blanchard and Chantal Francois, et al. "Parkin Immunoreactivity in the Brain of Human and Non‐Human Primates: An Immuno-histochemical Analysis in Normal Conditions and in Parkinsonian Syndromes." J Comp Neurol 432 (2001): 184-196.
21. Cesari, Rossano, Eric S Martin, George A Calin and Francesca Pentimalli, et al. "Parkin, a Gene Implicated in Autosomal Recessive Juvenile Parkinsonism, is a Candidate Tumor Suppressor Gene on Chromosome 6q25–q27." Proc Nat Acad Sci 100 (2003): 5956-5961.
22. Xu, Liang, De-chen Lin, Dong Yin and H Phillip Koeffler. "An Emerging Role of PARK2 in Cancer." J Mol Med 92 (2014): 31-42.
23. Veeriah, Selvaraju, Luc Morris, David Solit and Timothy A Chan. "The Familial Parkinson disease gene PARK2 is a Multisite Tumor Suppressor on Chromosome 6q25. 2-27 that Regulates Cyclin E." 12 (2010): 1451-1452.
24. Sun, Xiaodong, Min Liu, Jihui Hao and Dengwen Li, et al. "Parkin Deficiency Contributes to Pancreatic Tumorigenesis by Inducing Spindle Multi-polarity and Misorientation." Cell Cycle 12 (2013): 1133-1141.
25. Klimczak, Phamela Ferreira, Danielle Hornung Ventury, Fabio Rueda Faucz and Nikolaos Settas, et al. "Association of a PARK2 Germline Variant and Epithelial Ovarian Cancer in a Southern Brazilian Population." Oncol 91 (2016): 101-105.
26. Gong, Yongxing, Travis Ian Zack, Luc GT Morris and Kan Lin, et al. "Pan-Cancer Genetic Analysis Identifies PARK2 as a Master Regulator of G1/S Cyclins." Nat Genet 46 (2014): 588-594.
27. Lee, SeungBaek, Jun She, Bo Deng and JungJin Kim, et al. "Multiple-Level Validation Identifies PARK2 in the Development of Lung Cancer and Chronic Obstructive Pulmonary Disease." Oncotarget 7 (2016): 44211.
28. Lei, Zhong, Huijie Duan, Tengfei Zhao and Yuxiang Zhang, et al. "PARK2 Inhibits Osteosarcoma Cell Growth through the JAK2/STAT3/VEGF Signaling Pathway." Cell Death Dis 9 (2018): 1-13.
29. Bernardini, JP, M Lazarou and Grant Dewson. "Parkin and Mitophagy in Cancer." Oncogene 36 (2017): 1315-1327.
30. Villa, Elodie, Sandrine Marchetti and Jean-Ehrland Ricci. "No Parkin Zone: Mitophagy without Parkin." Trends Cell Biol 28 (2018): 882-895.
31. Braschi, Emélie, Rodolfo Zunino and Heidi M McBride. "MAPL is a New Mitochondrial SUMO E3 Ligase that Regulates Mitochondrial Fission." EMBO Rep 10 (2009): 748-754.
32. Yun, Jina, Rajat Puri, Huan Yang and Michael A Lizzio. "MUL1 Acts in Parallel to the PINK1/parkin Pathway in Regulating Mitofusin and Compensates for Loss of PINK1/parkin." Elife 3 (2014): e01958.
33. Lin, De-Chen, Liang Xu, Ye Chen and Haiyan Yan, et al. "Genomic and Functional Analysis of the E3 Ligase PARK2 in Glioma." Cancer Res 75 (2015): 1815-1827.
34. Yeo, Calvin WS, Felicia SL Ng, Chou Chai and Jeanne MM Tan, et al. "Parkin Pathway Activation Mitigates Glioma Cell Proliferation and Predicts Patient Survival." Cancer Res 72 (2012): 2543-2553.
35. Wang, Haiyang, Zhenfeng Jiang, Meng Na and Haitao Ge, et al. "PARK2 Negatively Regulates the Metastasis and Epithelial‑Mesenchymal Transition of Glioblastoma Cells via ZEB1." Oncol Letters 14 (2017): 2933-2939.
36. Dunham, Christopher. "Pediatric Brain Tumors: A Histologic and Genetic Update on Commonly Encountered Entities." In: Seminars in Diagnostic Pathology 27 (2010): 147-159.
37. Appin, Christina L and Daniel J Brat. "Biomarker-Driven Diagnosis of Diffuse Gliomas." Mol Aspects Med 45 (2015): 87-96.
38. Poulogiannis, George, Rebecca E McIntyre, Maria Dimitriadi and John R Apps, et al. "PARK2 Deletions Occur Frequently Sporadic Colorectal Cancer and Accelerate Adenoma Development in Apc Mutant Mice." Proc Nat Acad Sci 107 (2010): 15145-15150.
39. Letessier, Anne, Sarah Garrido-Urbani, Christope Ginestier and Gaelle Fournier, et al. "Correlated Break at PARK2/FRA6E and Loss of AF-6/Afadin Protein Expression are Associated with Poor Outcome in Breast Cancer." Oncogene 26 (2007): 298-307.
40. Toma, Marieta I, Daniela Wuttig, Sandy Kaiser and Alexander Herr, et al. "PARK2 and PACRG are Commonly Downregulated in Clear‐Cell Renal Cell Carcinoma and are Associated with Aggressive Disease and Poor Clinical Outcome." Genes Chromosomes Cancer 52 (2013): 265-273.