The workshop focuses on the clinical application of comprehensive data and approach toward better clinical outcome of patients with cancers. The outstanding basic and clinical researchers will give a talk about their research and perspectives.

Distinguished researchers in a filed of cancer research and proteogenomics will give a talk about their research and perspectives.

Henry Rodriguez, NCI, USA

Title: International Collaboration in the Age of Precision Medicine [Cancer Moonshot Proteogenomics]

Summary:

Massive analysis of cancer genomics data from thousands of tumors from all major cancer types has facilitated the molecular classification of cancer in modern medicine, guiding a precision oncology approach for patients. However, advances in precision oncology have limitations as the needs of many cancer patients remain unaddressed. Integration of proteomics with genomics (proteogenomics) facilitates to advance our understanding of the causes of cancer and narrow target selection for potential therapeutic intervention by providing granularity regarding cancer-relevant pathways. Proteogenomics presents an opportunity to generate new insights in cancer biology by melding genomics with proteomics, with the potential to fast become an essential part of laboratory medicine. This seminar will discuss how genomics, transcriptomics, and proteomics are being combined in the quest to understand the etiology of cancer – in basic clinical sample studies and translational research (clinical trials).

Jong Bae Pak, NCC, Korea

Title: Proteogenomic analysis of longitudinal trajectory of glioblastoma evolution 

Summary:

The evolutionary trajectory of glioblastoma after therapy cannot be associated with discrete alterations of driver cancer genes. Here, we performed a proteogenomic analysis of 123 longitudinal glioblastoma pairs, temporally separated by the standard-of-care treatment. Integrative analyses identified the prevalence of a highly proliferative and neural progenitor-associated state at diagnosis and its replacement by recurrent tumors sharing activation of neuronal differentiation and synaptogenic pathways. Proteomic and phosphoproteomic analyses revealed that the molecular transition from the proliferative/progenitor subtype to the neuronal at recurrence is marked by coherent post-translational activation of the WNT/PCP signaling pathway and the BRAF protein kinase. Multi-Omic analysis of Patient-Derived Xenograft (PDX) model demonstrated similar patterns of evolutionary trajectory, marked by activation of neuronal signaling programs that were consistently observed in longitudinal patients. Inhibition of the BRAF kinase with small molecule inhibitors impaired both neuronal differentiation and invasive capability of recurrent glioblastoma cells, which are the phenotypic hallmarks of glioblastoma progression after therapy. This work provides comprehensive insights into the biological mechanisms of glioblastoma evolution and treatment resistance and highlights new therapeutic opportunities to effectively counter them in the clinic.

Koichi Matsuda, University of Tokyo Japan

Title: Genomic characterization of sarcomas for novel therapy

Summary:

Bone soft tissue tumor is a rare cancer showing remarkable diversity. We constructed Japan Sarcoma Genome Consortium (JSGC) in 2014 through the collaboration of 30 medical and 12 research institutes. By the integrated whole exome and RNA sequencing of 115 dedifferentiated liposarcoma (DDLPS), we identified CTDSP1/2-DNM3OS fusion genes in 8% of tumors, that are specific for dedifferentiated components of DDLPS. DDLPS are clustered into three groups based on SCNAs that are associate with clinical outcome. For the analysis of Tenosynovial giant cell tumor, we found frequent CBL mutation (35%) which was associated with poor prognosis and JAK2 activation. RNA sequencing revealed fusion transcripts involving CSF1, including novel CSF1-VCAM1, CSF1-FN1 and CSF1-CDH1 fusions, in 13/18 (72%) cases. Analysis of extraskeletal osteosarcoma revealed frequent TP53 mutations in 5/10 (50%), deletion of CDKN2A locus, amplification of 1p32, 1q21, and 12q15, a kataegis-like pattern in 4/7 cases. We are also conducting genomic analysis of more than 500 samples from 13 tumor types. These studies would contribute to elucidate the pathogenesis of bone and soft tumors as well as drug development.

Koichi Ogura, NCC, Japan

Title:Genomic Landscape of Actionable Alterations in Bone and Soft Tissue Sarcomas: Report from the Center for Cancer Genomics and Advanced Therapeutics(C-CAT) database

Summary:

Sarcomas are heterogenous tumors comprising >150 subtypes according to the 2020 WHO classification. The rarity of each subtype of sarcoma poses challenges for diagnosis, understanding biology, and generating clinical evidence to support new drug development. This background urged us to investigate the actionable genomic landscape of bone and soft tissue sarcoma using the big data from the Center for Cancer Genomics and Advanced Therapeutics(C-CAT) database, which was established along with the initiation of coverage of comprehensive genomic profiling (CGP) tests by the National Health Insurance in June 2019 and consists of 44,466 tests with matched detailed clinical information (as of November 27, 2022). We also explored the molecular profile of patients with advanced/metastatic soft tissue sarcoma who underwent second-line chemotherapy to identify potential predictive biomarkers by comparing molecular profiles and clinical data using C-CAT database. In my presentation, I’ll focus on the evidence generated from the analyses of C-CAT data and examine the role and significance of C-CAT database in understanding biology and generating clinical evidence in sarcomas to improve the sarcoma care.

Koji, Ueda, Japanese Foundation of Cancer Research, Japan

Title: Proteogenomic profiling of myxoid liposarcoma

Summary:

Myxoid liposarcoma (MLS) is the second most common liposarcoma characterized by formation of the FUS-CHOP fusion gene. Although most MLS patients have a good prognosis, metastatic recurrence causes quite poor survival. Predicting the prognosis of MLS at the time of diagnosis is important, but no predictive biomarker has been identified yet. To explore prognostic biomarkers for MLS, as a project of International Cancer Proteogenome Consortium (ICPC), we performed global genomic and proteomic analyses of MLS tissues (n = 79). 

From RNA expression dataset, we identified that the transcription levels of 16 genes defined immunologically hot/cold MLS. Importantly, this classification clearly associated with survival rate and metastasis occurrence rate. 

On the other hand, from the comprehensive proteomic analysis of MLS tumor regions and adjacent normal regions, a specific key factor was determined from 4966 identified proteins, which was considered to induce tumor-associated macrophages (TAMs) within the tumor microenvironment. Multiple color immunohistochemical staining analysis confirmed that the expression level of this protein and also the number of infiltrated TAMs in MLS tissues strongly correlated with survival and recurrence rates after surgery. 

Thus, our proteogenomic datasets highlight an important fact that microenvironmental immune activity has a great impact on the prognosis of MLS, which may lead to development of a new way of immunotherapy for MLS. 

Le Su, Jacksonville State University, USA

Title: Study of SS18-SSX fusion-independent vulnerabilities in synovial sarcoma

Summary:

Missense mutations in the p53 gene are commonly seen in human cancer. Compelling evidence shows that mutant p53 proteins contribute to malignant transformation of many different tissues. Yet, there is a lack of knowledge about mutant p53 oncogenic function in synovial sarcoma, especially because the underlying mechanism of this disease is thought to exclusively rely on the characteristic SS18-SSX fusion oncoprotein. In this lecture, we will discuss an essential role for mutant p53 activity in synovial sarcoma biology. Our work shows that mutant p53 depletion inhibits synovial sarcoma cell growth, even though SS18-SSX expression and function remain largely unaffected. At a molecular level, the c-Met receptor tyrosine kinase acts as a primary effector of mutant p53. Inhibition of c-Met not only has a detrimental effect on p53-mutant synovial sarcoma cell growth, but also exhibits synergy with the compound FK228 which disrupts SS18-SSX function by attenuation of the HDAC activity. These results indicate that the combination of c-Met and HDAC inhibitors might create a new therapeutic strategy to effectively target a subset of synovial sarcoma with oncogenic p53 mutations.

Marc Ladanyi, Memorial Sloan Kettering, USA

Title: Insights from routine, large scale clinical cancer genomic testing of sarcomas

Summary:

In 2014, MSKCC embarked on a massive clinical genomics initiative based on the MSK-IMPACT assay, a hybrid capture based DNA NGS assay initially covering 341 genes, now covering 505 cancer-related genes. This has been complemented by a customized targeted RNAseq assay, MSK-Fusion, based on the Archer anchored PCR technology, designed to detect gene fusions not readily detectable by DNA-based NGS. As of early 2023, MSK-IMPACT has been used in approximately 3500 adult and pediatric patients with advanced bone and soft tissue sarcomas, representing one of the largest sarcoma genomic datasets publicly available. In this presentation, I will highlight some of the intriguing observations emerging from this large dataset.   

Roger Reddel, ProCan, Children's Medical Research Institute, Australia

Title: The ProCan® pan-cancer knowledge base of proteomic and clinical data

Summary:

The ProCan® program objectives are to develop methods for generating proteomic data from cancer tissues in a clinically-relevant time frame and build a knowledge base of cancer proteomes and associated clinical data so proteomic data can assist with treatment decisions in the cancer clinic. A preparative pipeline for extracting purified peptides from formalin-fixed paraffin-embedded or fresh frozen cancer samples or cancer model systems including cell lines has been developed, and Data-Independent Analysis (DIA) mass spectrometry is being used to generate reproducible proteomic data. The total processing time from receiving a tumor section to the completed proteome is < 9 hours. This methodology was used to analyse 949 cancer cell lines from Wellcome Sanger Institute for which drug-response data had been obtained for >600 drugs, and the proteomic data was demonstrated to be at least as good as the transcriptome for predicting drug response. The same methods have been used to analyse ~17,000 cancer samples representing most types of cancer, including many sarcomas. Machine learning models trained on cell line data can predict with high accuracy the cancer type and, in some cases, sub-type from tumor proteomes. For several cancer types, proteomic data have been used to predict clinical outcome.

Ross Okimoto, University of California San Francisco, USA

Title: Molecular characterization of CIC-DUX4 sarcomas toward novel therapy

Summary:

CIC-DUX4 rearrangements define an aggressive and chemotherapy insensitive subset of undifferentiated sarcomas. The CIC-DUX4 fusion drives oncogenesis through direct transcriptional upregulation of cell-cycle and DNA-replication genes. Notably, CIC-DUX4 mediated CCNE1 upregulation compromises the G1/S transition, conferring a survival dependence on the G2/M cell-cycle checkpoint. Through an integrative transcriptional and kinase activity screen using patient-derived specimens, we now show that CIC-DUX4 sarcomas depend on the G2/M checkpoint regulator, WEE1, as an adaptive survival mechanism. Specifically, CIC-DUX4 sarcomas depend on WEE1 activity to limit DNA-damage and unscheduled mitotic entry. Consequently, genetic or pharmacologic WEE1 inhibition in vitro and in vivo leads to rapid DNA damage associated apoptotic induction of patient-derived CIC-DUX4 sarcomas. Thus, we identify WEE1 is an actionable therapeutic vulnerability in CIC-DUX4 sarcomas.

Rei Noguchi, NCC, Japan

Title:Comprehensive kinase activity assay providing prognostic biomarker candidates for treatments in oncology

Summary:

Protein phosphorylation is a key process regulating various cellular physiological events such as proliferation, migration, cell cycle progression, apoptosis, and differentiation. Dysregulated kinase activity occurs in the process of carcinogenesis and disease progression due to genetic alterations including copy number alterations and point mutations. Moreover, aberrant kinome consisting of the dysregulated self-phosphorylation and kinase-to-kinase regulatory relationship is also responsible for the pathogenesis of cancer. Aberrantly regulated kinases are considered to be biomarkers and therapeutic targets, and several dozens of kinase inhibitors have been approved for the cancer treatments. A three-dimension peptide array has been applied to comprehensive kinase activity assay to investigate 100 tyrosine kinases simultaneously using a faint amount of proteins from biopsy samples. We discuss the perspectives of kinomics in cancer research, and introduce comprehensive kinase activity assay using the three-dimensional peptide array. We performed the comprehensive kinase activity assay to identify prognostic biomarker candidates for advanced non-small cell lung cancer harboring sensitive EGFR mutations for EGFR-tyrosine kinase inhibitors. The comprehensive kinase activity assay generates proteome data that can lead to prognostic biomarker candidates for novel treatments in oncology.

Satoshi Nishizuka, Iwate Medical School, Japan

Title: Integrative proteogenomic analysis for cancer therapeutic biomarker discovery

Summary:

The administration of molecular targeting drugs requires genomic companion diagnosis, but most effector molecules are proteins. Therefore, drug efficacy should be evaluated by an integrative approach at the genomic and proteomic levels for cancer therapy with molecular targeting drugs. To clarify the association between genomic and proteomic statuses, we used next-generation sequencing and reverse-phase protein array technology at the genomic and proteomic levels, respectively. Potential driver and actionable mutations, including those in TP53, CTNNB1, EGFR, BRAF, PIK3CA, VEGFR, and ERBB2, were identified in gastrointestinal cancers, whereas no significant differences at protein levels were found based on mutational status, except for TP53 mutations. We next investigated how the proteomic profile is affected by genomic and transcriptomic statuses in the context of liver cancer therapy. The proteomic profile revealed three subclasses that may guide suitable therapeutic strategies categorized by immunologic and genomic statuses. Finally, proteogenomic profiling focused on the response to DNA-damaging agents of gastric cancer cells demonstrated that proteins regulated by immune-response gene expression induced by the NF-kB/IkBa feedback mechanism are potential prognostic biomarkers for 5-FU sensitivity. Collectively, the powerful utility of an integrative proteogenomic approach for cancer biomarker discovery will be presented.

Tadashi Kondo, NCC, Japan

Title:Patient-derived cancer models for pharmaco-proteogenomics in rare cancers

Summary:

Rare cancers are defined as malignancies with extremely low incidence such as 1 per 100,000 people per year. More than 200 malignancies are included in the category of rare cancers, and the total number of patients with rare cancers accounts for approximately 20 % of all cancer patients. Thus, the rare cancers are not rare, and it is important to improve the clinical outcome of the patients with rare cancers. Because the number of patients with rare cancers is extremely small, the effective treatments are limited, and the fundamental research resources are not available well. The limited availability of patient-derived cancer models is one of the major factors which hinder the development of novel treatments in rare cancers. Although the mutation-to-drug response relationship has been used in the clinical practice, it depends on the cancer types, and we need to confirm it using cancer models. In addition, as the molecular backgrounds of disease progression are common among the different cancers, the drugs effective for rare cancers should exist among the present anti-cancer drugs and we can find them by drug screening in cancer models. With these notions, we started to establish sarcoma cell lines and xenografts in 2012. Using the established cell lines, we screened the anti-cancer drugs and examined the proteogenomics backgrounds of response and resistance. The utility and perspective of cancer models will be discussed. 

Yasushi Ishihama, Kyoto University, Japan

Title: Large-scale proteoform analysis in proteogenomics: current status and challenges

Summary:

Fundamental questions that proteogenomics must address include the extent to which various perturbations in the genome, transcriptome and translatome are reflected in the proteome, and the extent to which protein lifetimes are regulated by perturbations in the genome and transcriptome. To solve this problem, protein fragment measurement by conventional shotgun proteomics is not sufficient and requires the analysis of proteoforms, the various molecular forms found in the translated product of a single gene, including gene mutation products, spliced RNA transcripts, single amino acid polymorphisms, post-translational modifications, and proteolytic processing products.

Our laboratory has recently developed the CHAMP method, a chromatographic approach for separating N- and C-terminal peptides of proteins that recognizes the charged position of the target peptide by ion-exchange chromatography or ligand-exchange chromatography. Using this CHAMP method, the isolation of protein terminal peptides can be achieved with >95% selectivity in only two steps: protease digestion and chromatographic separation. By matching a customized transcriptome database with terminal proteomic LC/MS/MS data, large-scale proteoform analyses of different protein termini are possible. In this talk, I will report on the current status and challenges of proteogenomic terminome analysis combining genome/transcriptome data and protein terminome analysis.

Hiroyuki Mano, NCC, Japan

Tadashi Kondo, NCC, Japan

Rei Noguchi, NCC, Japan

Mari Masuda, NCC, Japan

Koichi Ogura, NCC, Japan