Complex Cellular Phenotype Analysis

Our group studies the spatial architectures and phenotypes of cells and tissues under different human diseases and environmental influences. We develop and use quantitative imaging assays and machine learning models to predict the biological effects of genetic mutations, drugs, and/or environmental agents. Our current research areas include spatial profiling of cells and tissues, bioimage databases and portals, digital medicine for cancer, and toxicodynamics of drugs/chemicals (Fig. 1).

Figure 1. Our current research areas

Our members come from diverse scientific backgrounds, including chemistry, cell biology, immunology, computer science, and bioinformatics. We collaborate with different academic, clinical, industrial, and governmental research groups, including Institute of Molecular and Cell Biology (IMCB), Singapore Institute of Food and Biotechnology Innovation (SIFBI), Singapore General Hospital (SGH), National Cancer Centre Singapore (NCCS), Lee Kong Chain (LKC) School of Medicine, and Harvard Beth Israel Deaconess Medical Center (BIDMC).

SPATIAL PROFILING OF CELLS AND TISSUES

Recent advances in multiplex immunohistochemistry/immunofluorescence (mIHC/IF) technologies have enabled simultaneous measurements of large numbers of markers on the same tissue sections, and more comprehensive views of the cellular compositions and immune responses at the tumor microenvironment (TME). We have developed computational methods and software tools to construct quantitative and compact representations of cellular or tissue phenotypes based on these multiplexed cellular or tissue images (Fig. 2). Our methods and tools can handle terabytes of image data collected under large numbers of experimental conditions or patients. Phenotypic profiles constructed using these methods have been used to classify the effects of small molecules and assess potentially harmful effects of chemicals and environmental agents (Miller et al., 2020; Hussain et al., 2020; Friedman et al., 2019; Lee et al., 2018; and Su et al., 2016).

Figure 2. CellXpress 2.0 can handle and quantify highly-multiplexed and large human tissue images.

BIOIMAGE DATABASES AND PORTALS

The reproducibility and interpretation of the complex staining patterns and analysis results obtained from mIHC/IF technologies are vital to their general adoptions. We develop and maintain an online platform for managing, visualizing, and sharing large tissue images called the HistoPathology Analytics (HPA) Platform (Fig. 3). The platform can help researchers and clinicians to more rapidly and accurately quantify the effects of cancer therapeutic agents (Yeong et al., 2022, Leong et al., 2021). An online public portal for mIHC/IF images and results for immuno-oncology called ImmunoAtlas (https://ImmunoAtlas.org) has also been built based on the HPA Platform (Lee et al., 2021).

Figure 3. HistoPath Analytics (HPA) is a cloud-based digital histopathology platform for organizing, sharing, visualizing, and analyzing large histological images.

DIGITAL MEDICINE FOR CANCER

Recent advances in genomic, transcriptomic, phenotypic, and histopathological profiling technologies have enabled the generation of large amounts of molecular and phenotypic information about the physiopathology of individual cancer patients. We are developing machine learning and data analytics methods to integrate data from these diverse technologies to stratify patients and select optimum targeted interventions for hepatocellular carcinoma, breast and other cancers.  We are also developing imaging assays that can predict the pathogenicity of genetic variants of key cancer-associated genes.

TOXICODYNAMICS OF XENOBIOTICS

Many xenobiotics have unknown and/or non-specific intracellular targets. To study the toxicodynamics of these chemicals, unbiased approaches that do not require prior information about the targets or mechanisms of the chemicals are required. Our focus is to study chemical analogs with related structures but differential cellular effects (Goh et al., 2021; Jaladanki et al., 2021), and develop fit-for-purpose assays that will be used by regulatory agencies and industrial research laboratories to assess chemical safety. We have developed a high-throughput and predictive in vitro pulmonary toxicity assay based on a human bronchial epithelial cell line, BEAS-2B (Fig. 4; Lee et al., 2018). The assay can accurately classify 33 reference chemicals with human pulmonotoxicity information (88.8% balance accuracy, 84.6% sensitivity, and 93.0% specificity). We also participated in an international case study that demonstrates the utility of in vitro bioactivity as a lower bound estimate of in vivo adverse effect levels in risk-based prioritization (Friedman et al., 2019).

Figure 4: Immunofluorescence microscopy images of human lung cells showing different phenotypic responses to non-toxic (blue) and toxic (red) chemicals.

• Yeong J, Lum HYJ, Teo CB, Tan BKJ, Chan YH, Tay RYK, Choo JRE, Jeyasekharan AD, Miow QH, Loo LH, Yong WP, Sundar R. “Choice of PD-L1 immunohistochemistry assay influences clinical eligibility for gastric cancer immunotherapy” Gastric Cancer, 25:741–750 (2022).


• Lee JYJ, Yeong J, Lee LWJN, Loo LH, Dong J. “ImmunoAtlas: an online public portal for sharing, visualizing, and referencing multiplex immunohistochemistry/ immunofluorescence (mIHC/IF) images and results for immuno-oncology”. Journal for ImmunoTherapy of Cancer, 9 (Suppl 2):A657 (2021).


• Leong TKM, Lo WS, Lee WEZ, Tan B, Lee XZ, Lee LWJN, Lee JYJ, Suresh N, Loo LH, Szu E, Yeong J. “Leveraging advances in immunopathology and artificial intelligence to analyze in vitro tumor models in composition and space”. Advanced Drug Delivery Reviews, 177:113959 (2021).


• Goh JJN, Behn J, Chong CS, Zhong G, Maurer-Stroh S, Fan H, Loo LH. “Structure-based virtual screening of CYP1A1 inhibitors: towards rapid tier-one assessment of potential developmental toxicants”. Archives of Toxicology, 95(9):3031-3048 (2021).


• Jaladanki CK, He Y, Zhao LN, Maurer-Stroh S, Loo LH, Song H, Fan H. “Virtual screening of potentially endocrine-disrupting chemicals against nuclear receptors and its application to identify PPARγ-bound fatty acids”. Archives of Toxicology, 95(1):355-374 (2021).


• Miller J, and Loo LH. “Optimum concentration–response curve metrics for supervised selection of discriminative cellular phenotypic endpoints for chemical hazard assessment”. Archives of Toxicology, 94:2951-2964 (2020).


• Hussain F, Basu S, Heng JJH, Loo LH, and Zink D. “Predicting direct hepatocyte toxicity in humans by combining high-throughput imaging of HepaRG cells and machine learning-based phenotypic profiling”. Archives of Toxicology, 94:2749-2767(2020).


• Van Der Ven L, Rorije E, Sprong C, Zink D, Derr R, Hendriks G, Loo LH, and Luijten M. “A case study with triazole fungicides to explore practical application of next generation hazard assessment methods for human health”. Chemical Research in Toxicology, 33(3):834-848 (2020).


• Paul Friedman K, Gagne M, Loo LH, Karamertzanis P, Netzeva T, Sobanski T, Franzosa JA, Richard AM, Lougee RR, Gissi A, Lee JJ, Angrish M, Dorne JL, Foster S, Raffaele K, Bahadori T, Gwinn MR, Lambert J, Whelan M, Rasenberg M4, Barton-Maclaren T, Thomas R. “Utility of in vitro bioactivity as a lower bound estimate of in vivo adverse effect levels and in risk-based prioritization”. Toxicological Sciences, 173(1):202-225 (2020).


• Lee JYJ, Miller JA, Basu S, Kee TZV, and Loo LH. “Building predictive in vitro pulmonary toxicity assays using high-throughput imaging and artificial intelligence”. Archives of Toxicology, 92(6):2055-2075 (2018).


• Loo LH, Zink D. “High-throughput prediction of nephrotoxicity in humans”. Alternatives to Laboratory Animals, 45:241-252 (2017a).


• Loo LH, Bougen-Zhukov N, Tan WLC. “Early spatiotemporal-specific changes in intermediate signals are predictive of cytotoxic sensitivity to TNF and co-treatments”. Scientific Reports, 7:43541 (2017b).


• Laksameethanasan D, Tan RZ, Toh GWL, and Loo LH. “cellXpress: a fast and user-friendly software platform for profiling cellular phenotypes”. BMC Bioinformatics, 14(Suppl 16):S4 (2013).


• View full list of publications