Association analyses of the JAK/STAT signaling pathway with the progression and prognosis of colon cancer

Single-cell transcriptomics reveals stage- and side-specificity of gene modules in colorectal cancer

BACKGROUND

An understanding of mechanisms underlying colorectal cancer (CRC) development and progression is yet to be fully elucidated. This study aims to employ network theoretic approaches to analyse single cell transcriptomic data from CRC to better characterize its progression and sided-ness.

METHODS

We utilized a recently published single-cell RNA sequencing data (GEO-GSE178341) and parsed the cell X gene data by stage and side (right and left colon). Using Weighted Gene Co-expression Network Analysis (WGCNA), we identified gene modules with varying preservation levels (weak or strong) of network topology between early (pT1) and late stages (pT234), and between right and left colons. Spearman's rank correlation (ρ) was used to assess the similarity or dissimilarity in gene connectivity.

RESULTS

Equalizing cell counts across different stages, we detected 13 modules for the early stage, two of which were non-preserved in late stages. Both non-preserved modules displayed distinct gene connectivity patterns between the early and late stages, characterized by low ρ values. One module predominately dealt with myeloid cells, with genes mostly enriched for cytokine-cytokine receptor interaction potentiallystimulating myeloid cells to participate in angiogenesis. The second module, representing a subset of epithelial cells, was mainly enriched for carbohydrate digestion and absorption, influencing the gut microenvironment through the breakdown of carbohydrates. In the comparison of left vs. right colons, two of 12 modules identified in the right colon were non-preserved in the left colon. One captured a small fraction of epithelial cells and was enriched for transcriptional misregulation in cancer, potentially impacting communication between epithelial cells and the tumor microenvironment. The other predominantly contained B cells with a crucial role in maintaining human gastrointestinal health and was enriched for B-cell receptor signalling pathway.

CONCLUSIONS

We identified modules with topological and functional differences specific to cell types between the early and late stages, and between the right and left colons. This study enhances the understanding of roles played by different cell types at different stages and sides, providing valuable insights for future studies focused on the diagnosis and treatment of CRC.

Combination of oxaliplatin and β-carotene suppresses colorectal cancer by regulating cell cycle, apoptosis, and cancer stemness in vitro

BACKGROUND/OBJECTIVES

Colorectal cancer (CRC) is the third most common cancer worldwide with a high recurrence rate. Oxaliplatin (OXA) resistance is one of the major reasons hindering CRC therapy. β-Carotene (BC) is a provitamin A and is known to have antioxidant and anticancer effects. However, the combined effect of OXA and BC has not been investigated. Therefore, this study investigated the anticancer effects and mechanism of the combination of OXA and BC on CRC.

MATERIALS/METHODS

In the present study, the effects of the combination of OXA and BC on cell viability, cell cycle arrest, and cancer stemness were investigated using HCT116, HT29, OXA-resistant cells, and human CRC organoids.

RESULTS

The combination of OXA and BC enhanced apoptosis, G2/M phase cell cycle arrest, and inhibited cancer cell survival in human CRC resistant cells and CRC organoids without toxicity in normal organoids. Cancer stem cell marker expression and self-replicating capacity were suppressed by combined treatment with OXA and BC. Moreover, this combined treatment upregulated apoptosis and the stem cell-related JAK/STAT signaling pathway.

CONCLUSIONS

Our results suggest a novel potential role of BC in reducing resistance to OXA, thereby enhances the anticancer effects of OXA. This enhancement is achieved through the regulation of cell cycle, apoptosis, and stemness in CRC.

Thyroid gland cancer and insulin resistance: a modern view of the problem

Background. The impact of insulin resistance on the probability of increase in thyroid cancer risk has been drawing a lot of attention of researchers lately. This problem is far from being completely solved. Studying this interrelationship may influence the effectiveness of the treatment of the mentioned widespread pathology. Purpose – to review present literature sources on research of interrelationship between insulin resistance and thyroid cancer, and also possible mechanisms of this relationship. Materials and methods. Literature search was performed manually by the keywords (thyroid cancer, insulin resistance, IGF-1, IGF-2, abdominal obesity, increase in body mass index, metformin), and also literature sources from evidential data bases PubMed, Web of Science were reviewed. Metaanalyses, systematic reviews and cohort studies were also taken into account. 148 literature sources were studied in total. The sources, which had been published within the last 10 years, were preferably selected. Results. Insulin resistance is viewed as an important independent factor of development of numerous malignancies. The carcinogenic activity of insulin resistance is caused by the resistance itself, as well as by the metabolic disorders related to it. It has been established that excessive weight and obesity are to a great extent attributed to more aggressive clinical pathological signs of thyroid cancer. Recent research showed a larger volume of thyroid and higher risk of knot forming in patients with insulin resistance. Thus, thyroid cancer is one of the main factors of thyroid transformation. Therapeutic methods of eliminating metabolic syndrome and associated hormonal diseases for prevention and therapy of oncologic diseases are drawing ever-greater scientific interest. The anti-tumor features of metformin and its capability of retarding carcinogenesis are shown in the studies. Conclusions. The given literature analysis has proved that the problem of treating malignant thyroid tumors and their metastasis is caused not only by morphological, cellular and molecular-biological features of the tumor itself, but also by insufficient knowledge about the interrelationship between insulin resistance, abdominal obesity, increase in body mass index, high-calorie diet and reduction of consumption of polyunsaturated fats, harmful impact of environment with molecular changes, specific for thyroid cancer. It is confirmed by a significant increase in thyroid cancer rate, especially papillary histotype, alongside with an increase in obesity rate. The studying of possibilities of decreasing incidence and mortality rates of oncologic pathology when using medications, which stabilize insulin and contribute to a decrease in degree of hyperinsulinemia, one of which is metformin, generates profound interest

Extracellular vesicles from gastric epithelial GES-1 cells infected with Helicobacter pylori promote changes in recipient cells associated with malignancy

Chronic Helicobacter pylori (H. pylori) infection is considered the main risk factor for the development of gastric cancer. Pathophysiological changes in the gastric mucosa initiated by this bacterium can persist even after pharmacological eradication and are likely attributable also to changes induced in non-infected cells as a consequence of intercellular communication via extracellular vesicles (EVs). To better understand what such changes might entail, we isolated EVs from immortalized normal gastric GES-1 cells infected (EVHp+) or not with H. pylori (EVHp-) by ultracentrifugation and characterized them. Infection of GES-1 cells with H. pylori significantly increased the release of EVs and slightly decreased the EV mean size. Incubation with EVHp+ for 24 h decreased the viability of GES-1 cells, but increased the levels of IL-23 in GES-1 cells, as well as the migration of GES-1 and gastric cancer AGS cells. Furthermore, incubation of GES-1 and AGS cells with EVHp+, but not with EVHp-, promoted cell invasion and trans-endothelial migration in vitro. Moreover, stimulation of endothelial EA.hy926 cells for 16 h with EVHp+ promoted the formation of linked networks. Finally, analysis by mass spectrometry identified proteins uniquely present and others enriched in EVHp+ compared to EVHp-, several of which are known targets of hypoxia induced factor-1α (HIF-1α) that may promote the acquisition of traits important for the genesis/progression of gastric pre-neoplastic changes associated with H. pylori infection. In conclusion, the harmful effects of H. pylori infection associated with the development of gastric malignancies may spread via EVs to non-infected areas in the early and later stages of gastric carcinogenesis.

Advances in the research of plant-derived natural products against retinoblastoma

Retinoblastoma (RB) is a highly aggressive ocular tumor, and due to socioeconomic and medical constraints, many children receive treatment only in the metaphase and advanced clinical stages, resulting in high rates of blindness and disability. Although several approaches exist in the treatment of RB, some children with the disease do not have satisfactory results because of various factors. Plant-derived natural products have shown definite therapeutic effects in the treatment of various tumors and are also widely used in the study of RB. We review plant-derived natural products used in the study of anti-RB to provide ideas for the clinical application of these drugs and the development of new therapeutic drugs.

The Potential Mechanisms of Cinobufotalin Treating Colon Adenocarcinoma by Network Pharmacology

Network pharmacology, as a novel way using bioinformatics to explore drug targets and interactions in cancer, broadens our understanding of drug action, thereby facilitating drug discovery. Here, we utilized network pharmacology to explore the role and mechanism by which cinobufotalin functions in colon adenocarcinoma (COAD). We found that cinobufotalin represses the growth and proliferation of colon cancer cells, and integrated public databases for targets reported to be associated with COAD, together with those predicted to be targets of cinobufotalin. Targets overlapped between COAD-associated proteins and cinobufotalin target proteins were used to filter candidate targets of cinobufotalin in COAD. The following proteins were thought to occupy a key position in COAD-cinobufotalin target networks: SRC, PIK3R1, MAPK1, PIK3CA, HSP90AA1, CTNNB1, GRB2, RHO1, PTPN11, and EGFR. The networks regulated by cinobufotalin were involved mainly in extracellular signal stimulation and transduction, including MAPK signaling pathway, PI3K-AKT signaling pathway, and JAK-STAT signaling pathway. Besides, transcriptome sequencing results also indicated that cinobufotalin inhibits the response of colon cancer cells to extracellular stimulation and promotes cell apoptosis. Molecular docking results showed that cinobufotalin matches in the pocket of the top candidate cinobufotalin target proteins (SRC, PIK3R1, MAPK1 and PIK3CA). These findings demonstrate cinobufotalin can be developed as potential anti-cancer therapeutics.

Multi-Omic Data Improve Prediction of Personalized Tumor Suppressors and Oncogenes

The progression of tumorigenesis starts with a few mutational and structural driver events in the cell. Various cohort-based computational tools exist to identify driver genes but require multiple samples to identify less frequently mutated driver genes. Many studies use different methods to identify driver mutations/genes from mutations that have no impact on tumor progression; however, a small fraction of patients show no mutational events in any known driver genes. Current unsupervised methods map somatic and expression data onto a network to identify personalized driver genes based on changes in expression. Our method is the first machine learning model to classify genes as tumor suppressor gene (TSG), oncogene (OG), or neutral, thus assigning the functional impact of the gene in the patient. In this study, we develop a multi-omic approach, PIVOT (Personalized Identification of driVer OGs and TSGs), to train on experimentally or computationally validated mutational and structural driver events. Given the lack of any gold standards for the identification of personalized driver genes, we label the data using four strategies and, based on classification metrics, show gene-based labeling strategies perform best. We build different models using SNV, RNA, and multi-omic features to be used based on the data available. Our models trained on multi-omic data improved predictions compared with mutation and expression data, achieving an accuracy ≥0.99 for BRCA, LUAD, and COAD datasets. We show network and expression-based features contribute the most to PIVOT. Our predictions on BRCA, COAD, and LUAD cancer types reveal commonly altered genes such as TP53 and PIK3CA, which are predicted drivers for multiple cancer types. Along with known driver genes, our models also identify new driver genes such as PRKCA, SOX9, and PSMD4. Our multi-omic model labels both CNV and mutations with a more considerable contribution by CNV alterations. While predicting labels for genes mutated in multiple samples, we also label rare driver events occurring in as few as one sample. We also identify genes with dual roles within the same cancer type. Overall, PIVOT labels personalized driver genes as TSGs and OGs and also identified rare driver genes.

Differential expression analysis of genes and long non-coding RNAs associated with KRAS mutation in colorectal cancer cells

KRAS mutation is responsible for 40–50% of colorectal cancers (CRCs). RNA-seq data and bioinformatics methods were used to analyze the transcriptional profiles of KRAS mutant (mtKRAS) in comparison with the wild-type (wtKRAS) cell lines, followed by in-silico and quantitative real-time PCR (qPCR) validations. Gene set enrichment analysis showed overrepresentation of KRAS signaling as an oncogenic signature in mtKRAS. Gene ontology and pathway analyses on 600 differentially-expressed genes (DEGs) indicated their major involvement in the cancer-associated signal transduction pathways. Significant hub genes were identified through analyzing PPI network, with the highest node degree for PTPRC. The evaluation of the interaction between co-expressed DEGs and lncRNAs revealed 12 differentially-expressed lncRNAs which potentially regulate the genes majorly enriched in Rap1 and RAS signaling pathways. The results of the qPCR showed the overexpression of PPARG and PTGS2, and downregulation of PTPRC in mtKRAS cells compared to the wtKRAS one, which confirming the outputs of RNA-seq analysis. Further, significant upregualtion of miR-23b was observed in wtKRAS cells. The comparison between the expression level of hub genes and TFs with expression data of CRC tissue samples deposited in TCGA databank confirmed them as distinct biomarkers for the discrimination of normal and tumor patient samples. Survival analysis revealed the significant prognostic value for some of the hub genes, TFs, and lncRNAs. The results of the present study can extend the vision on the molecular mechanisms involved in KRAS-driven CRC pathogenesis.

Modulating Properties of Piroxicam, Meloxicam and Oxicam Analogues against Macrophage-Associated Chemokines in Colorectal Cancer

The mechanisms underlying the antineoplastic effects of oxicams have not been fully elucidated. We aimed to assess the effect of classic and novel oxicams on the expression/secretion of macrophage-associated chemokines (RTqPCR/Luminex xMAP) in colorectal adenocarcinoma cells, and on the expression of upstream the non-steroidal anti-inflammatory drug (NSAID)-activated genes NAG1, NFKBIA, MYD88, and RELA, as well as at the chemokine profiling in colorectal tumors. Meloxicam downregulated CCL4 9.9-fold, but otherwise the classic oxicams had a negligible/non-significant effect. Novel analogues with a thiazine ring substituted with arylpiperazine and benzoyl moieties significantly modulated chemokine expression to varying degree, upregulated NAG1 and NFKBIA, and downregulated MYD88. They inhibited CCL3 and CCL4, and their effect on CCL2 and CXCL2 depended on the dose and exposure. The propylene linker between thiazine and piperazine nitrogens and one arylpiperazine fluorine substituent characterized the most effective analogue. Only CCL19 and CXCL2 were not upregulated in tumors, nor was CXCL2 in tumor-adjacent tissue compared to normal mucosa. Compared to adjacent tissue, CCL4 and CXCL2 were upregulated, while CCL2, CCL8, and CCL19 were downregulated in tumors. Tumor CCL2 and CCL7 increased along with advancing T and CCL3, and CCL4 along with the N stage. The introduction of arylpiperazine and benzoyl moieties into the oxicam scaffold yields effective modulators of chemokine expression, which act by upregulating NAG1 and interfering with NF-κB signaling.

The Role of Oxidative Stress and Its Counteractive Utility in Colorectal Cancer (CRC)

An altered redox status accompanied by an elevated generation of reactive oxygen/nitrogen species (ROS/RNS) has been implicated in a number of diseases including colorectal cancer (CRC). CRC, being one of the most common cancers worldwide, has been reported to be associated with multiple environmental and lifestyle factors (e.g., dietary habits, obesity, and physical inactivity) and harboring heightened oxidative stress that results in genomic instability. Although under normal condition ROS regulate many signal transduction pathways including cell proliferation and survival, overwhelming of the antioxidant capacity due to metabolic abnormalities and oncogenic signaling leads to a redox adaptation response that imparts drug resistance. Nevertheless, excessive reliance on elevated production of ROS makes the tumor cells increasingly vulnerable to further ROS insults, and the abolition of such drug resistance through redox perturbation could be instrumental to preferentially eliminate them. The goal of this review is to demonstrate the evidence that links redox stress to the development of CRC and assimilate the most up-to-date information that would facilitate future investigation on CRC-associated redox biology. Concomitantly, we argue that the exploitation of this distinct biochemical property of CRC cells might offer a fresh avenue to effectively eradicate these cells.

Identification of the Prognostic Value and Clinical Significance of Interferon Regulatory Factors (IRFs) in Colon Adenocarcinoma

BACKGROUND Colon adenocarcinoma (COAD) is one of the most common malignant tumors and has high incidence and mortality rates. The interferon regulatory factor (IRF) family is known as a key transcription factor in the IFN signaling pathway and cellular immunity. This research explored the relationship between the IRF family and COAD through use of bioinformatics technology. MATERIAL AND METHODS Using the UALCAN and GEPIA databases, we analyzed the transcription and prognostic value of IRFs in COAD, and GSCALite was used in cancer genomics analysis. TIMER, LinkedOmics, and Metascape were used to assess the potential function of IRFs in COAD. RESULTS The transcription levels of IRF3 were elevated in COAD tissues, while IRF2/4/6 were downregulated compared with normal patients in subgroup analyses of race, age, weight, sex, nodal metastasis, individual cancer stages, TP53 mutation status, and histological subtypes. IRF3 and IRF7 in COAD were significantly associated with a poor prognosis. Drug sensitivity analysis revealed that the expression level of IRF2/4/8 was negatively associated with drug resistance. A significant correlation was found between the IRF family and immune cell infiltration. Moreover, enrichment analysis revealed that the IRFs were associated with response to tumor necrosis factor, transcription misregulation in cancer, and JAK-STAT signaling pathway. We also identified several kinase and miRNA targets of the IRF family in COAD. CONCLUSIONS We identified IRF3 and IRF7 as prognostic biomarkers in COAD, and the IRF family was associated with immune cell infiltration and gene regulation networks, providing additional evidence showing the significant role of the IRF family in COAD.

Cell Death by Gallotannin Is Associated with Inhibition of the JAK/STAT Pathway in Human Colon Cancer Cells

Background

Gallotannin (GT) is a polyphenol that possesses interesting anticancer properties. However, the mechanisms underlying its antitumor effects have not been well defined.

Objective

This study was designed to clarify the mechanisms underlying GT antitumor effects in colon cancer cell lines.

Methods

Three isogenic HCT116 cell lines (p53^+/+, p53^-/-, and p21^-/-) were treated with GT for different time points then Western blot, flow cytometry, and senescence analysis were performed to examine the effect of GT on Mitogen-activated protein kinase (MAPK) and Janus kinase (JAK)/signal transducer and activator of transcription (STAT) effectors, STAT3 downstream apoptotic targets, Sub-G1 phase, and programmed cell death induction. Transfection using Invitrogen Lipofectamine 2000 Transfection Reagent (Thermo Fisher Scientific, Waltham, Massachusetts) were used to identify the role of p53 and p21 in the p53^-/- and p21^-/- cell lines.

Results

Both low and high GT concentrations caused MAPKs activation marked by upregulation of extracellular signal-regulated kinase (p-ERK). The preincubation with the antioxidant Tiron (Sigma-Aldrich, St Louis, Missouri) showed that GT's antitumor effects were not mediated by reactive oxygen species. We then examined the effect of GT on the JAK/STAT pathway, which is known to be activated in colorectal cancer. GT totally inhibited the JAK/STAT pathway effectors JAK2, STAT1, and STAT3 and their downstream apoptotic regulators B-cell lymphoma-extra large (Bcl-x_L) and c-Myc in all 3 cell lines. HCT116 cancer cells exhibited differential sensitivity to GT with p21^-/- cells being the most sensitive and p53^+/+ cells that express p21 protein being the least sensitive. In p53^+/+ cells, GT induced senescence, whereas in p53^-/- and p21^-/- cells, GT induced apoptosis in a caspase independent manner marked by Poly(ADP-Ribose) Polymerase (PARP) cleavage, Bcl-2 downregulation, and upregulation of the Bcl-2 associated X (Bax) to B-cell lymphoma 2 (Bcl-2) ratio. In addition, the sub-G1 phase exceeded 50% in p21^-/- cells.

Conclusions

Considered together, our results indicate that GT is potent inhibitor of the JAK/STAT pathway in colon cancer irrespective of the p53 and p21 status, which provides insights into its mechanism of anticancer activities and future potential for clinical translation. (Curr Ther Res Clin Exp. 2020; 81:XXX-XXX).

Flaming the fight against cancer cells: the role of microRNA-93

There have been attempts to develop novel anti-tumor drugs in cancer therapy. Although satisfying results have been observed at a consequence of application of chemotherapeutic agents, the cancer cells are capable of making resistance into these agents. This has forced scientists into genetic manipulation as genetic alterations are responsible for generation of a high number of cancer cells. MicroRNAs (miRs) are endogenous, short non-coding RNAs that affect target genes at the post-transcriptional level. Increasing evidence reveals the potential role of miRs in regulation of biological processes including angiogenesis, metabolism, cell proliferation, cell division, and cell differentiation. Abnormal expression of miRs is associated with development of a number of pathologic events, particularly cancer. MiR-93 plays a significant role in both physiological and pathological mechanisms. At the present review, we show how this miR dually affects the proliferation and invasion of cancer cells. Besides, we elucidate the oncogenesis or oncosuppressor function of miR-93.

SPTBN1 suppresses the progression of epithelial ovarian cancer via SOCS3-mediated blockade of the JAK/STAT3 signaling pathway

SPTBN1 plays an anticancer role in many kinds of tumors and participates in the chemotherapeutic resistance of epithelial ovarian cancer (EOC). Here, we reported that lower SPTBN1 expression was significantly related to advanced EOC stage and shorter progression-free survival. SPTBN1 expression was also higher in less invasive EOC cell lines. Moreover, SPTBN1 decreased the migration ability of the EOC cells A2780 and HO8910 and inhibited the growth of EOC cells in vitro and tumor xenografts in vivo. SPTBN1 suppression increased the epithelial mesenchymal transformation marker Vimentin while decreasing E-cadherin expression. By analyzing TCGA data and immunohistochemistry staining of tumor tissue, we found that SPTBN1 and SOCS3 were positively coexpressed in EOC patients. SOCS3 overexpression or JAK2 inhibition decreased the proliferation and migration of EOC cells as well as the expression of p-JAK2, p-STAT3 and Vimentin, which were enhanced by the downregulation of SPTBN1, while E-cadherin expression was also reversed. It was also verified in mouse embryonic fibroblasts (MEFs) that loss of SPTBN1 activated the JAK/STAT3 signaling pathway with suppression of SOCS3. Our results suggest that SPTBN1 suppresses the progression of epithelial ovarian cancer via SOCS3-mediated blockade of the JAK/STAT3 signaling pathway.

Prophylactic Role of Lactobacillus paracasei Exopolysaccharides on Colon Cancer Cells through Apoptosis Not Ferroptosis

Background: Nowadays despite conventional methods in colon cancer treatment, targeting vital molecular pathways and induction of various forms of cell death by safe probiotic components like exopolysaccharides (EPSs) are of great importance and are considered as potential therapeutic agents. This study aimed to investigate the inhibitory effect of the EPS of L. paracasei on different colon cancer cell lines (SW-480, HT-29, and HCT-116). Methods: For this purpose, several cellular and molecular experiments including MTS assay, DAPI staining, Annexin V/PI assay, quantitative real-time PCR (qPCR) and some important ferroptosis-related assays were performed. Results: Based on the findings, L. paracasei EPS can induce apoptosis confirmed by all apoptosis related assays and could not act through ferroptosis pathways. L. paracasei EPS could hinder the Akt1, mTOR, and Jak-1 mRNAs, and induces apoptosis through down-regulation of the antiapoptotic gene (Bcl-2), up-regulation of pro-apoptotic genes (BAX, caspase-3, 8). Conclusion: The exploited EPS of an indigenous probiotic strain with anticancer potential with low/insignificant cytotoxicity to normal cells is proposed for future applications in molecular targeted therapy of colon cancer treatment. Furthermore, in vivo and clinical trials should be performed to evaluate the applicability of this component besides conventional methods to increase the survival rate of colon cancer patients.

Development and Clinical Validation of a Novel 4-Gene Prognostic Signature Predicting Survival in Colorectal Cancer

In this study, we collected genes related to energy metabolism, used gene expression data from public databases to classify molecular subtypes of colon cancer (COAD) based on the genes related to energy metabolism, and further evaluated the relationships between the molecular subtypes and prognosis and clinical characteristics. Differential expression analysis of the molecular subtypes yielded 1948 differentially expressed genes (DEGs), whose functions were closely related to the occurrence and development of cancer. Based on the DEGs, we constructed a 4-gene prognostic risk model and identified the high expression of FOXD4, ENPEP, HOXC6, and ALOX15B as a risk factor associated with a high risk of developing COAD. The 4-gene signature has strong robustness and a stable predictive performance in datasets from different platforms not only in patients with early COAD but also in all patients with colon cancer. The enriched pathways of the 4-gene signature in the high- and low-risk groups obtained by GSEA were significantly related to the occurrence and development of colon cancer. Moreover, the results of qPCR, immunohistochemistry staining and Western blot assay revealed that FOXD4, ENPEP, HOXC6, and ALOX15B are over expressed in CRC tissues and cells. These results suggesting that the signature could potentially be used as a prognostic marker for clinical diagnosis.

Network modeling of patients' biomolecular profiles for clinical phenotype/outcome prediction

Methods for phenotype and outcome prediction are largely based on inductive supervised models that use selected biomarkers to make predictions, without explicitly considering the functional relationships between individuals. We introduce a novel network-based approach named Patient-Net (P-Net) in which biomolecular profiles of patients are modeled in a graph-structured space that represents gene expression relationships between patients. Then a kernel-based semi-supervised transductive algorithm is applied to the graph to explore the overall topology of the graph and to predict the phenotype/clinical outcome of patients. Experimental tests involving several publicly available datasets of patients afflicted with pancreatic, breast, colon and colorectal cancer show that our proposed method is competitive with state-of-the-art supervised and semi-supervised predictive systems. Importantly, P-Net also provides interpretable models that can be easily visualized to gain clues about the relationships between patients, and to formulate hypotheses about their stratification.