DSpace Community:

Proteomic Profiling of Human Extracellular Vesicles Reveals Diagnostic Biomarkers for Colon Adenocarcinoma

2026-04-01T00:00:00Z

Title: Proteomic Profiling of Human Extracellular Vesicles Reveals Diagnostic Biomarkers for Colon Adenocarcinoma Authors: Seo, Yura; Han, Yoon Dae; Bojmar, Linda; Kim, Kyung‐A; Seo, Yurin; Kim, Taeyul K.; Lee, Suho; Kim, Yeleem; Choi, Hye Bin; Lim, Yujin H.; Kim, Chae Hyun; Sandberg, Alexander; Fan, Chuanwen; Lauritzen, Pernille; Molina, Henrik; Peralta, Christopher; Geri, Jacob B.; Burdette, Colin; Han, Dai Hoon; Gee, Heon Yung; Lee, Insuk; Shin, Jeon Soo; Kim, Hyunwook; Li, Leon; Tobias, Gabriel C.; Wortzel, Inbal; Shin, Sang Joon; Jung, Hyo‐Il; Lee, Min Goo; Paik, Soonmyung; Schwartz, Robert E.; Ahn, Joong Bae; Lyden, David; Kim, Han Sang; 김현욱 Abstract: Early detection of colon adenocarcinoma (COAD) remains suboptimal. Fecal tests fail to diagnose 30% of stage I cancer, and serum CEA lacks sensitivity (< 40%). Extracellular vesicles (EVs) circulate systemically and package tumor-related cargo, making them attractive non-invasive biomarkers for cancer diagnosis. We profiled the EV proteome from 233 human patients using LC-MS/MS, including stage I-IV tumors with matched non-tumor colon tissues (n = 50 each; n = 100), paired pre-/post-operative plasma (n = 90) and healthy plasma (n = 43). Circulating EVs contained both tumor-specific and stromal/immune cell-derived proteins, reflecting the systemic nature of EV biology in the cancer setting. Proteomic analysis identified 745 proteins enriched in tumor-derived EVs (e.g., SRPK1, THBS2) and 127 proteins enriched in adjacent tissues. Plasma EVs revealed 166 proteins enriched in COAD (e.g., UBA1, FCN1) and 233 enriched in healthy controls. Pathway analysis linked tumor EV cargo to angiogenesis, mRNA splicing, TGF-beta signalling and RNA translation. Notably, a cross-cancer comparison (pancreatic = 10, lung = 14 cases) revealed that 76% of tumor EV proteins were COAD-specific, highlighting tissue of origin specificity. We further developed a 10-protein EV panel comprising seven tumor-associated and three healthy-enriched EV proteins, which effectively distinguished COAD patients from healthy controls in the two validation cohorts (n = 104 and n = 215), achieving > 90% sensitivity for differentiating COAD from healthy and non-COAD colorectal conditions. Six weeks after curative resection, tumor-associated EV proteins decreased by > 70%, whereas healthy-associated proteins rebounded to baseline, indicating surgical responsiveness. Collectively, EV protein signatures provide a sensitive and tissue-specific window into tumor-host communication, further supporting blood-based early detection of COAD.

Proteogenomic decoding of chemotherapy resistance in patients with triple-negative breast cancer

2026-04-01T00:00:00Z

Title: Proteogenomic decoding of chemotherapy resistance in patients with triple-negative breast cancer Authors: Lee, Dong Ki; Kim, Min Hwan; Hwang, Yumi; Kim, Seul-Gi; Ryu, Won-Ji; Kim, Geon-Uk; Yun, Hyun Myoung; Park, Shinyoung; Lee, Jeong Dong; Han, Hyun Ju; Kim, Gun Min; Kim, Kyung-Hee; Park, Jong Bae; Kim, Min Jung; Koo, Ja Seung; Kim, Jee Ye; Park, Hyung Seok; Kim, Seung Il; Gee, Heon Yung; Park, Seho; Sohn, Joohyuk; 유원지; 김지예 Abstract: BackgroundThe clinical utility of integrated proteogenomic biomarkers for predicting chemotherapy response in triple-negative breast cancer remains underexplored. We prospectively analyzed paired baseline and post-treatment tumor samples from 50 patients with stage II-III TNBC treated with anthracycline- and taxane-based neoadjuvant chemotherapy, integrating whole-exome sequencing, RNA sequencing, global proteomics, and phosphoproteomics.ResultsNon-negative matrix factorization clustering identifies five proteogenomic subtypes. The immune-enriched subtype demonstrates the highest pathologic complete response rate (55.6%), whereas no pathologic complete response was observed in the xenobiotic metabolism or epithelial-mesenchymal transition subtypes. Immune-related pathways are enriched in tumors with pathologic complete response, while epithelial-mesenchymal transition pathways are enriched in non-pathologic complete response tumors. The estrogen response pathway is selectively enriched in non-pathologic complete response tumors at the proteomic level and inversely correlated with immune activation. Post-translational modification and in vitro analyses suggest estrogen-linked GRK2 activation contributes to chemotherapy resistance. ITGB8 copy number loss is associated with higher pathologic complete response rates and immune activation, while non-pathologic complete response tumors of the immunomodulatory subtype show increased expression of AKR1C2 and ABCA13. Comparison of baseline and post-treatment tumors reveals AURKB pathway activation in residual disease, with Aurora B kinase inhibition synergizing with paclitaxel. A predictive model incorporating these biomarkers outperforms RNA-based models in predicting response.ConclusionIntegrative proteogenomic profiling enables robust prediction of chemotherapy resistance in triple-negative breast cancer and identifies actionable biomarkers providing a framework for advancing personalized therapeutic strategies.

High-resolution functional mapping of androgen receptor variants

2026-04-01T00:00:00Z

Title: High-resolution functional mapping of androgen receptor variants Authors: Oh, Hyeong-Cheol; Chang, Yoojin; Park, Jihye; Cheong, Yumin; Lee, Kwang Seob; Han, Hyunho; Kim, Hyongbum Henry Abstract: The androgen receptor (AR) plays a central role in the progression and therapy resistance of prostate cancer, the most common cancer in men. While AR signalling inhibitors such as enzalutamide are key treatments, their efficacy is often compromised by drug-resistant AR variants. In addition, mutations leading to germline loss of AR function cause androgen insensitivity syndrome, yet most AR variants are of uncertain significance. Here we use advanced prime editing to generate and assess 2,765 AR variants, covering 99.95% of all possible single amino acid variants encoded by single nucleotide variants in the ligand-binding domain. This mapping identified 755 new non-functional AR variants and revealed 225 and 40 new variants resistant to enzalutamide and bavdegalutamide, an AR degrader, respectively. Our findings also enabled prognosis prediction for patients with prostate cancer based on AR mutation profiles. The broader implications of the study include improved androgen insensitivity syndrome diagnosis, better prostate cancer prognosis prediction and precision treatments for patients with prostate cancer.

A regulatory axis for tonotopic MYO7A expression in cochlear hair cells

2026-04-01T00:00:00Z

Title: A regulatory axis for tonotopic MYO7A expression in cochlear hair cells Authors: Li, Sihan; Jun, Sujin; Kim, Ye-Ri; Gogineni, Ayathi; Lee, Franklin; Kim, Chul Hoon; Kim, Un-Kyung; Peng, Anthony W.; Shin, Jung-Bum Abstract: Myo7a, a gene mutated in Usher syndrome and nonsyndromic deafness, encodes an unconventional myosin essential for hair cell function. Our previous work revealed that cochlear hair cells express distinct Myo7a isoforms with unique spatial and cell type-specific patterns. The canonical isoform (Myo7a-C) and an additional isoform (Myo7a-N) are co-expressed in outer hair cells (OHCs) but exhibit opposing tonotopic gradients, while inner hair cells primarily express Myo7a-C. These isoforms arise from distinct transcriptional start sites, indicating separate regulatory inputs. Here, we identify an intronic cis-regulatory element, EnhancerA, essential for tonotopically graded Myo7a expression. EnhancerA deletion reduces MYO7A protein levels in a tonotopically varied manner, disrupts hair bundle morphogenesis, alters OHC mechanotransduction, and leads to hair cell degeneration and hearing loss. We further identify SIX2, a tonotopically expressed transcription factor that may interact with EnhancerA to regulate Myo7a-N in OHCs. These findings define a cis-trans regulatory axis critical for isoform-specific Myo7a expression and cochlear function.