Susan E Celniker’s research while affiliated with Lawrence Berkeley National Laboratory and other places

PURPOSE Tissue-agnostic biomarkers that capture the commonality in cancer biology, may provide a new avenue for treatment development and optimization across cancer types. Here, we aimed to evaluate and validate the clinical value of a tissue-agnostic cellular morphometrics biomarker (CMB) signature, which was discovered by artificial intelligence (AI) from H&E-stained whole-slide images (WSI) of diagnostic slides of colon cancers, in pan-gastrointestinal (pan-GI) pre-cancer lesions and cancers. METHODS We discovered CMBs from WSI using our well-established CMB-ML pipeline and established a CMB risk score (CMBRS) using multivariate regression models. Based on CMBRS, we assigned individual patients from The Cancer Genome Atlas Colon Adenocarcinoma Cohort (TCGA-COAD) (n=430) to CMB risk groups (CMBRG). We then extensively evaluated tissue-agnostic clinical value of CMB signature, CMBRS and CMBRG in multi-cohorts with different types of GI cancer (n=2,219) and risk assessment of precancerous lesions (n=1,016). We unraveled each CMB-related biological function using bulk RNA-sequencing, single-cell RNA-sequencing (scRNA-seq) and opal multiplex immunohistochemistry (IHC) techniques. RESULTS From the TCGA-COAD cohort, we developed a 13-CMB signature and constructed CMBRS/CMBRG that predict prognosis of colon cancer patients. Importantly, this 13-CMB signature proved prognostic and predictive values for TCGA patients with rectal, gastric and esophageal cancer independent of traditional clinical factors. These findings were independently validated using multiple cohorts from Drum Tower Hospital. Moreover, 13-CMB signature exhibited the power for risk stratification of colon adenoma and early esophageal neoplastic lesion patients for predicting cancer progression. In addition, we demonstrated and validated independent prognostic impacts of gene signatures and CMB signatures and a significant increase in predictive power by integration of CMB signature, gene signature and clinical factors. Correlations between CMBs and gene expression levels revealed the association of each CMB with biological functions including cell proliferation, epithelial-to-mesenchymal transition and immune microenvironment. The association of CMBs with the immune microenvironment was prospectively validated by scRNA-seq and was further confirmed by Opal multiplex IHC staining in colon cancer. CONCLUSION This study demonstrates the clinical value of tissue-agnostic AI-empowered CMB signature from WSI with defined biological functions, which can be used in clinical settings to assess risk, diagnose disease, and guide clinical interventions. Tissue-agnostic CMBs potentially provide a new avenue for a rapid, robust and cost-effective cross-cancer prediction that is essential for developing common treatment strategy for multiple cancers.

A catalog of transcription factor (TF) binding sites in the genome is critical for deciphering regulatory relationships. Here we present the culmination of the efforts of the Model Organism ENCyclopedia Of DNA Elements (modENCODE) and the model organism Encyclopedia of Regulatory Networks (modERN) consortia to systematically assay TF binding events in vivo in two major model organisms, Drosophila melanogaster (fly) and Caenorhabditis elegans (worm). These datasets comprise 605 TFs identifying 3.6M sites in the fly and 356 TFs identifying 0.9 M sites in the worm and represent the majority of the regulatory space in each genome. We demonstrate that TFs associate with chromatin in clusters termed "metapeaks", that larger metapeaks have characteristics of high occupancy target (HOT) regions, and that the importance of consensus sequence motifs bound by TFs depends on metapeak size and complexity. Combining ChIP-seq data with single cell RNA-seq data in a machine learning model identifies TFs with a prominent role in promoting target gene expression in specific cell types, even differentiating between parent-daughter cells during embryogenesis. These data are a rich resource for the community that should fuel and guide future investigations into TF function. To facilitate data accessibility and utility, all strains expressing GFP-tagged TFs are available at the stock centers for each organism. The chromatin immunoprecipitation sequencing data are available through the ENCODE Data Coordinating Center, GEO, and through a direct interface that provides rapid access to processed data sets and summary analyses, as well as widgets to probe the cell type-specific TF-target relationships.

Ionizing radiation induces complex changes in cells and tissues. The conventional approach to biological dosimetry has been to integrate physical and clinical measurements to optimize dose assessment. Molecular biodosimetry is an effective strategy to monitor radiation exposure and hematologic, cytogenetic, protein and transcript-based approaches have been developed to increase dose estimation accuracy. However, these approaches are invasive, time-consuming, have limited effectiveness over time, and importantly do not accurately inform on low dose radiation exposures. Therefore, novel, non-invasive, biomarkers are required that can overcome these limitations. We developed a pipeline that employs Fourier transform infrared (FTIR) spectroscopy in the mid-infrared spectrum to identify a signature of low dose ionizing radiation exposure in mouse ear pinnae over time. Two cohorts of C57BL/6J and one cohort of BALB/c mice were followed for ninety days after total body X-ray exposures of 10, 50, 100 or 200 cGy. The stratum corneum layer of the ear pinnae were repeatedly measured with attenuated total reflectance - focal plane array (ATR-FPA)-FTIR at 5, 14, 21, 49 and 90 days after radiation exposure. We found statistically significant discriminative power for all doses and all tested time-points out to 90 days after exposure. Classification accuracy was maximized when testing 14 days after exposure (specificity > 0.9 with a sensitivity threshold of 0.9) and dropped by roughly 30% sensitivity at 90 days. Since only hundreds of samples were required to learn highly discriminative signatures, developing human-relevant diagnostic capabilities is likely feasible and this non-invasive procedure points toward future non-invasive biodosimetry applications at population scales.

The insights into interactions between host genetics and gut microbiome (GM) in colorectal tumor susceptibility (CTS) remains lacking. We used Collaborative Cross mouse population model to identify genetic and microbial determinants of Azoxymethane-induced CTS. We identified 4417 CTS-associated single nucleotide polymorphisms (SNPs) containing 334 genes that were transcriptionally altered in human colorectal cancers (CRCs) and consistently clustered independent human CRC cohorts into two subgroups with different prognosis. We discovered a set of genera in early-life associated with CTS and defined a 16-genus signature that accurately predicted CTS, the majority of which were correlated with human CRCs. We identified 547 SNPs associated with abundances of these genera. Mediation analysis revealed GM as mediators partially exerting the effect of SNP UNC3869242 within Duox2 on CTS. Intestine cell-specific depletion of Duox2 altered GM composition and contribution of Duox2 depletion to CTS was significantly influenced by GM. Our findings provide potential novel targets for personalized CRC prevention and treatment.

Non-invasive methods of detecting radiation exposure show promise to improve upon current approaches to biological dosimetry in ease, speed, and accuracy. Here we developed a pipeline that employs Fourier transform infrared (FTIR) spectroscopy in the mid-infrared spectrum to identify a signature of low dose ionizing radiation exposure in mouse ear pinnae over time. Mice exposed to 0.1 to 2 Gy total body irradiation were repeatedly measured by FTIR at the stratum corneum of the ear pinnae. We found significant discriminative power for all doses and time-points out to 90 days after exposure. Classification accuracy was maximized when testing 14 days after exposure (specificity > 0.9 with a sensitivity threshold of 0.9) and dropped by roughly 30% sensitivity at 90 days. Infrared frequencies point towards biological changes in DNA conformation, lipid oxidation and accumulation and shifts in protein secondary structure. Since only hundreds of samples were used to learn the highly discriminative signature, developing human-relevant diagnostic capabilities is likely feasible and this non-invasive procedure points toward rapid, non-invasive, and reagent-free biodosimetry applications at population scales.

A catalog of transcription factor (TF) binding sites in the genome is critical for deciphering regulatory relationships. Here we present the culmination of the modERN (model organism Encyclopedia of Regulatory Networks) consortium that systematically assayed TF binding events in vivo in two major model organisms, Drosophila melanogaster (fly) and Caenorhabditis elegans (worm). We describe key features of these datasets, comprising 604 TFs identifying 3.6M sites in the fly and 350 TFs identifying 0.9 M sites in the worm. Applying a machine learning model to these data identifies sets of TFs with a prominent role in promoting 10 target gene expression in specific cell types. TF binding data are available through the ENCODE Data Coordinating Center and at https://epic.gs.washington.edu/modERNresource, which provides access to processed and summary data, as well as widgets to probe cell type-specific TF-target relationships. These data are a rich resource that should fuel investigations into TF function during development.

To gain insight into the transcription programs activated during the formation of Drosophila larval structures, we carried out single cell RNA sequencing during two periods of Drosophila embryogenesis: stages 10 – 12, when most organs are first specified and initiate morphological and physiological specialization, and stages 13 – 16, when organs achieve their final mature architectures and begin to function. Our data confirm previous findings with regards to functional specialization of some organs – the salivary gland and trachea – and clarify the embryonic functions of another – the plasmatocytes. We also identify two early developmental trajectories in germ cells and uncover a potential role for proteolysis during germline stem cell specialization. We identify the likely cell type of origin for key components of the Drosophila matrisome and several commonly used Drosophila embryonic cell culture lines. Finally, we compare our findings to other recent related studies and to other modalities for identifying tissue-specific gene expression patterns. These data provide a useful community resource for identifying many new players in tissue-specific morphogenesis and functional specialization of developing organs.