Zelei Yang
Zelei Yang has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Sarah J. Hill
Dana-Farber Cancer Institute 1 Department of Medical Oncology Boston, Massachusetts
67 shared publications - Hyeji Jun
Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts
66 shared publications - Shannan J. Ho Sui
Harvard Chan Bioinformatics Core, Harvard School of Public Health, Harvard University, Boston, USA
58 shared publications - Upendra Raj Bhattarai
Harvard Chan Bioinformatics Core, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
58 shared publications - Ursula A. Matulonis
Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, USA
58 shared publications - Amy Cameron
Dana-Farber Cancer Institute Lurie Family Imaging Center, Center for Biomedical Imaging in Oncology Boston, Massachusetts
57 shared publications
Latest Publications
- Open Access
- 30 October 2026
- Published by American Association for Cancer Research (AACR) in Clinical Cancer Research
- p. OF1-OF12
Transcriptional Signatures Indicating RB Function Correlate with Response to CDK4/6 Inhibition in Endometrial Cancer
Purpose: CDK4/6 inhibitors (CDK4/6i) are being investigated in endometrial cancer clinical trials, with some patients showing a significant response. However, there are few biomarkers to predict which endometrial cancer patients will respond to CDK4/6is. Current inclusion criteria include endometrial cancer histology and/or lack of certain genetic alterations. We hypothesize that instead, endometrial cancer patients should be included or excluded from CDK4/6i clinical trials based on the retinoblastoma protein (RB)-functional status of the tumor, defined by the activity of RB function–predictive transcriptional signatures. Experimental Design: We analyzed RB1 and TP53 mutation status, RB protein expression, baseline transcriptional profiles, and RB function-predictive transcriptional signatures in endometrial cancer samples from 39 patients who participated in two combination clinical trials of the CDK4/6i abemaciclib, one testing letrozole/abemaciclib and the other letrozole/metformin/abemaciclib (NCT03675893). All patients had estrogen receptor–positive endometrial cancers of varying molecular subtypes, with the majority of cases being of endometrioid histology. We statistically assessed for correlation of the results of these analyses with objective response to the CDK4/6i in the clinical trials. Results: We show that RB protein expression levels and RB1 mutation status do not correlate with CDK4/6i response. Rather, we show that decreased activity of multiple RB function–predictive transcriptional signatures, which suggests that RB is functional, corresponds to response to the CDK4/6i. We show that reduced expression of specific E2F family members correlates with CDK4/6i response. Conclusions: These results suggest that RB function–predictive transcriptional signatures may be a relevant biomarker for CDK4/6i response in endometrial cancer and merit further refinement and validation in the clinical setting.
- 30 October 2026
- Published by American Association for Cancer Research (AACR) in Clinical Cancer Research
- p. OF1-OF12
Open AccessTranscriptional Signatures Indicating RB Function Correlate with Response to CDK4/6 Inhibition in Endometrial Cancer
Purpose: CDK4/6 inhibitors (CDK4/6i) are being investigated in endometrial cancer clinical trials, with some patients showing a significant response. However, there are few biomarkers to predict which endometrial cancer patients will respond to CDK4/6is. Current inclusion criteria include endometrial cancer histology and/or lack of certain genetic alterations. We hypothesize that instead, endometrial cancer patients should be included or excluded from CDK4/6i clinical trials based on the retinoblastoma protein (RB)-functional status of the tumor, defined by the activity of RB function–predictive transcriptional signatures. Experimental Design: We analyzed RB1 and TP53 mutation status, RB protein expression, baseline transcriptional profiles, and RB function-predictive transcriptional signatures in endometrial cancer samples from 39 patients who participated in two combination clinical trials of the CDK4/6i abemaciclib, one testing letrozole/abemaciclib and the other letrozole/metformin/abemaciclib (NCT03675893). All patients had estrogen receptor–positive endometrial cancers of varying molecular subtypes, with the majority of cases being of endometrioid histology. We statistically assessed for correlation of the results of these analyses with objective response to the CDK4/6i in the clinical trials. Results: We show that RB protein expression levels and RB1 mutation status do not correlate with CDK4/6i response. Rather, we show that decreased activity of multiple RB function–predictive transcriptional signatures, which suggests that RB is functional, corresponds to response to the CDK4/6i. We show that reduced expression of specific E2F family members correlates with CDK4/6i response. Conclusions: These results suggest that RB function–predictive transcriptional signatures may be a relevant biomarker for CDK4/6i response in endometrial cancer and merit further refinement and validation in the clinical setting.
- Open Access
- 13 May 2025
- Published by American Association for Cancer Research (AACR)
Figure S4 from Functional Profiling of p53 and RB Cell Cycle Regulatory Proficiency Suggests Mechanism-Driven Molecular Stratification in Endometrial Carcinoma
Figure S4 C B A Figure S4. Gating strategies for CDK4/6 inhibitor cell cycle flow cytometry corresponding to Figures 1G and 1I. Cell or organoid lines were treated with vehicle (DMSO) or 0.25µM of the CDK4/6 inhibitor (CDK4/6i) Abemaciclib for 24 hours. The cell lines were pulsed with bromodeoxyuridine (BrdU) or the organoids were pulsed with 5-Ethynyl-2'-deoxyuridine (EdU). Cells were harvested, fixed, and stained with appropriate antibodies or chemicals as well as propidium iodide (PI) and then analyzed on a flow cytometer. Gating strategies from one replicate of some of the data in Figures 1G and 1I are shown here as described below. A) Shown here is a general gating strategy for one replicate of the DMSO treated HEC1B cell line on the left or the EMCA-A organoid line on the right. In each case, cells were first gated on the side scatter (SSC)/forward scatter (FSC) plot. From cells, singlets representing cells with 2N and 4N DNA content were gated on the PI-Height (H)/PI-Area (A) plot. Singlets were then plotted with PI on the X axis and either BrdU-FITC for cell lines or EdU for organoids on the Y axis. Cells which had a 2N DNA content and were negative for BrdU or EdU were gated as G1 phase. Cells which had a 4N DNA content and were negative for BrdU or EdU were gated as G2/M phase. Cells which were EdU or BrdU positive were gated and quantified as being in S phase. B) BrdU (Y-axis)/PI (X-axis) gating strategies are shown for all cell lines treated with vehicle (top) or CDK4/6i Abemaciclib (bottom) corresponding to Figure 1G. C) EdU (Y-axis)/PI (X-axis) gating strategies are shown for all organoid lines treated with vehicle (top) or CDK4/6i Abemaciclib (bottom) corresponding to Figure 1I. HEC1B DMSO DMSO CDK4/6i (Abemaciclib) EMCA-A UPSC-A CS-A CS-B EMCA-A DMSO EdU-A EdU-A EdU-A
- 13 May 2025
- Published by American Association for Cancer Research (AACR)
Open AccessFigure S4 from Functional Profiling of p53 and RB Cell Cycle Regulatory Proficiency Suggests Mechanism-Driven Molecular Stratification in Endometrial Carcinoma
Figure S4 C B A Figure S4. Gating strategies for CDK4/6 inhibitor cell cycle flow cytometry corresponding to Figures 1G and 1I. Cell or organoid lines were treated with vehicle (DMSO) or 0.25µM of the CDK4/6 inhibitor (CDK4/6i) Abemaciclib for 24 hours. The cell lines were pulsed with bromodeoxyuridine (BrdU) or the organoids were pulsed with 5-Ethynyl-2'-deoxyuridine (EdU). Cells were harvested, fixed, and stained with appropriate antibodies or chemicals as well as propidium iodide (PI) and then analyzed on a flow cytometer. Gating strategies from one replicate of some of the data in Figures 1G and 1I are shown here as described below. A) Shown here is a general gating strategy for one replicate of the DMSO treated HEC1B cell line on the left or the EMCA-A organoid line on the right. In each case, cells were first gated on the side scatter (SSC)/forward scatter (FSC) plot. From cells, singlets representing cells with 2N and 4N DNA content were gated on the PI-Height (H)/PI-Area (A) plot. Singlets were then plotted with PI on the X axis and either BrdU-FITC for cell lines or EdU for organoids on the Y axis. Cells which had a 2N DNA content and were negative for BrdU or EdU were gated as G1 phase. Cells which had a 4N DNA content and were negative for BrdU or EdU were gated as G2/M phase. Cells which were EdU or BrdU positive were gated and quantified as being in S phase. B) BrdU (Y-axis)/PI (X-axis) gating strategies are shown for all cell lines treated with vehicle (top) or CDK4/6i Abemaciclib (bottom) corresponding to Figure 1G. C) EdU (Y-axis)/PI (X-axis) gating strategies are shown for all organoid lines treated with vehicle (top) or CDK4/6i Abemaciclib (bottom) corresponding to Figure 1I. HEC1B DMSO DMSO CDK4/6i (Abemaciclib) EMCA-A UPSC-A CS-A CS-B EMCA-A DMSO EdU-A EdU-A EdU-A