Search

Multi-model data integration has large scope in cancer research, in diagnosis, therapy andprognosis. The integration and analysis of different data types as one has wide impact inpersonalized medicine too. In papillary thyroid cancer, the most common but leastaggressive type of thyroid cancer, the integration of histopathology features to geneexpression profiles, survival data and tumor stage information can help improveprognosis of the condition. The study indicated that certain nuclear features extractedfrom slides are significantly associated with several genes. To be noted was the fact thatmany features related most significantly with genes involved in immune responses andgeneral immunity.

Abstract: Recurrent, metastatic forms of papillary thyroid cancer (PTC) continue to challenge the status quo therapy regime which combines surgical resection with radioactive iodine ablation of remnant thyroid and metastatic deposits. Repetitive use of these methods following the initial failure to control the disease impacts negatively on the quality of life and survival. Investigational use of targeted therapies like tyrosine kinase inhibitors (TKIs) have become a major part of the quaternary care given to patients with the recurrent, metastatic disease. Unfortunately, the varied outcomes and adverse side effects which accompany their empirical use underscores the need for a much better understanding of the cellular processes driving metastatic disease in PTC. We previously identified platelet derived growth factor receptor alpha (PDGFRα) as an important biological marker for the metastatic spread of PTC. By extension, in the current work, the prognostic implication of activated PDGFRα in papillary thyroid tumor samples was evaluated. We hypothesized that PDGFRα induces aggressive phenotypes such as lymph node metastases, dedifferentiation and compromised iodide transporting abilities, in papillary thyroid tumors. The signaling consequences of PDGFRα, and the effects of its blockade by tyrosine kinase inhibitors were also investigated. We demonstrate that PDGFRα creates an invasive and migratory follicular cell phenotype lacking the ability to transport sodium iodide in vitro and in vivo. The invasive and migratory behaviour are consequences of cellular epithelial to mesenchymal transition, and the formation of ECM degrading invadopodial protrusions. The unique ability of PDGFRα to mediate thyroid follicular cell dedifferentiation, with the ensuing suppression of iodide uptake in PTC cells, results from the downregulation of TTF1 expression. In clinical samples, PDGFRα is predictive of nodal spread, radioactive iodine resistance, as well as disease recurrence. We also observed that the expression of PDGF-AA, a PDGFRα-specific ligand, is associated with the metastatic PTC disease. Interestingly, our findings challenge the current canonical description of PTC as a MAPK/ERK driven malignancy, and clearly demonstrate a vital role for the STAT3 and AKT pathways in driving this aggressive disease phenotype. PDGFRα represents the first opportunity for targeted therapy in thyroid cancer that directly influences tumor differentiation, migration and ECM degradation. We propose that blockade of PDGFRα may be a useful adjunct therapy given its ability to dramatically reduce the severity of the aggressive phenotypes, where previous attempts with drugs that disrupt other tyrosine kinases have failed.

Papillary thyroid cancer (PTC) is the most frequently diagnosed form of thyroid cancer, representing the majority of all cases. For unknown reasons it is more commonly diagnosed in women then in men and over the past three decades, its incidence has been rising in the United States. It is unclear why this is occurring while the incidence of other cancers is declining. Perhaps changes in diagnostic criteria, improvement in early detection and advances in technology are the reasons. However, the overall occurrence of all tumor sizes and the annual number of patients who die from thyroid cancer has also risen; contradicting the notion that these reasons alone can account for the entire increase. This fact has prompted thyroid research to focus on investigating the pathways that lead to pathogenesis with a goal to develop new therapeutics. If detected at an early stage, PTC is curable most of the time; however, for patients with aggressive and invasive forms of it, treatment options are limited and prognosis is poor. Predictors of death from PTC include the presence of gross local invasion and the development of distant metastases. Therefore, understanding the mechanisms underlying not only PTC development, but also local invasion and metastases are imperative to develop new therapies targeted for this group of tumors. In thyroid cancer, the most frequently upregulated signaling cascade is the mitogen-activated protein kinase (MAPK) pathway, which propagates signals from surface receptors to the nucleus via serine-threonine kinases. Cell processes that are regulated through this pathway include growth, differentiation, development, inflammation and apoptosis. In thyroid cancer, gene rearrangements and point mutations in MAPK signaling molecules are a common event. These oncogenic events cause constitutive activation and are generally mutually exclusive within individual tumors. These facts have prompted pharmaceutical development to focus on the MAPK pathway, for the past several years. Despite these efforts and multiple ongoing studies, to date, there are no effective FDA-approved therapies for patients with progressive metastatic thyroid cancer. One particular predictor of aggressive tumor behavior (particularly extensive local invasion) and unresponsiveness to treatment is the presence of the activating mutation BRAF V600E. From our prior studies, we determined that the invasive fronts of aggressive PTC exhibited a characteristic change in morphology known as epithelial-to-mesenchymal transition (EMT). EMT has been linked to aggressive or late stage tumor invasion and metastasis. It frequently occurs in PTCs with the BRAF V600E mutation; however, EMT has also been identified in its absence, suggesting it is a common endpoint for aggressive PTC behavior. After analyzing the major signaling nodes in the invasive fronts, evidence indicated that there was preferential activation of p-21 activated kinase (PAK) signaling. When PAKs are activated, they redistribute to the leading edge of a cell and direct the formation of lamellopodia and motility. In addition, PAK overexpression has been discovered and implicated in the progression of a number of other solid tumors, particularly breast cancer. This data combined with the previous facts, convinced us to focus on investigating the mechanism(s) by which BRAF mutant PTCs acquire an aggressive phenotype. Our hypotheses were that PAKs regulate thyroid cancer cell motility and were involved in BRAF-mediated tumor invasiveness. Before conducting functional studies, we characterized PAK isoform expression in six human thyroid cancer cell lines, normal human thyroid tissue, and thyroid cancer tissue by quantitative RT-PCR and Western blot (WB). All expressed PAKs 1, 2, 3, 4, and in most cases PAK6. A literature search suggested that Group I PAKs were the isoforms most likely involved, so we focused the functional experiments on this group. We transfected into the six thyroid cancer cell lines a group I specific molecular inhibitor called PID. It reduced trans-well filter migration by more than 50% without altering cell viability. To discern which of the three group I PAK isoform(s) were specifically regulating thyroid motility, two of the six cell lines were transfected with PAK 1, 2 or 3-specific siRNA. Only the PAK1 siRNA reduced migration significantly in both cell lines. To investigate this observation in vivo, we immunohistochemically stained invasive PTCs samples and demonstrated that there was a statistically significant increase in both pPAK and PAK1 expression in the invasive fronts vs. the tumor center, which correlates with the earlier profile studies. Therefore, we concluded that PAK isoforms 1-3, 4 and 6 are commonly expressed in human thyroid cancer tissues and cell lines. In particular, PAK1 regulates thyroid cancer cell motility, and pPAK and PAK1 levels are increased in PTC invasive fronts.Now that PAKs had an established role in thyroid cancer progression, we wanted to explore if there was a relationship between PAK activation and the mutational status of BRAF. We selected three human thyroid cancer cell lines (BCPAP, TPC1 and FTC133) with different mutations and transfected in BRAF siRNA. It reduced group I PAK function in all three-cell lines regardless of the mutation, while group I PAK suppression did not affect BRAF function. So we hypothesized that BRAF regulated thyroid cancer cell motility through the activation of group I PAKs. We tested this by performing migration rescue experiments with constitutive active (CA) PAK1, which did rescue the decrease in cell motility induced by the BRAF siRNA. In addition, because MEK1/2 is a well-established BRAF downstream effector, we verified using rescue experiments that it also plays a role in BRAF-mediated migration. We inhibited both MEK1/2 and group I PAKs and showed a lack of synergy. To determine whether one regulated the other, we inhibited MEK1/2 activity and examined PAK function, which was not affected by the suppression. Using both an overexpression and endogenous system, we established that BRAF and PAK1 are co-localized as well as physically bound. To further explore this relationship in vivo, thyroids from doxycycline inducible BRAF V600E mice were immunohistochemically stained for pPAK and total group I PAKs. The inducible BRAF V600E mouse thyroids showed an increase in pPAK expression compared to wild-type thyroid controls, with levels of total PAK expression unaltered. In conclusion, we have identified a new signaling cascade in which BRAF activates PAK1 in a MEK-independent manner. This pathway regulates thyroid cancer cell motility and may be an important novel BRAF-regulated pathway involved in thyroid cancer invasion.

TP53 does not fully comply with the Knudson’s two-hit model in that a reduction of constitutional expression of p53 may be sufficient for tumor predisposition. To determine whether the TP53 gene dosage affects the transcriptional regulation of target genes, we performed oligo-array gene expression analysis using isogenic cells with different TP53 gene dosage. We identified 35 genes whose expression is significantly correlated to the dosage of TP53. These genes are involved in a variety of cellular processes including signal transduction, cell adhesion, and transcription regulation. Among those genes, CSPG2 [chondroitin sulfate proteoglycan 2 (Versican)] was selected for further study because it contains a p53 binding site in its first intron and its expression highly correlates with TP53 dosage. Using in vitro and in vivo assays, we showed CSPG2 to be directly transactivated by p53. In search of tumor suppressor genes in Papillary Thyroid Cancer (PTC), we previously used gene expression profiling to identify genes underexpressed in tumor compared with paired unaffected tissue. While searching for loss of heterozygosity (LOH) in the genomic regions harboring candidate tumor suppressor genes, we detected LOH in a ~ 20 kb region around marker D9S176. Several ESTs flanking D9S176 were underexpressed in tumors, and for one of the ESTs, downregulation was highly associated with the activating BRAF mutation V600E, the most common genetic lesion in PTC. A putative novel gene, NAMA (noncoding RNA associated with MAP kinase pathway and growth arrest) containing the affected EST was cloned and characterized. Several characteristics of NAMA suggest that it might be a non-coding but functional RNA. NAMA is inducible by knockdown of BRAF, inhibition of the MAP kinase pathway, growth arrest, and DNA damage in cancer cell lines. We suggest that NAMA is a noncoding RNA (ncRNA) associated with growth arrest. Thus, this study identified a list of genes affected by TP53 gene dosage including a direct target, CSPG2. Further, a noncoding RNA, NAMA, associated with the activating mutation in the BRAF oncogene, was discovered. These results suggest that gene dosage variation in tumor suppressor genes or noncoding RNA may modify cancer risk and progression.