Our research program centers on interrogating the interplay of dying tumor cells with immune cells, such as macrophages and NK cells, in the tumor microenvironment (TME). Our approaches include genome-wide CRISPR screening and gene editing, single cell RNAseq, spatial transcriptomics (Nanostring GeoMX and CosMX), spectral cytometry (Cytek), artificial intelligence/deep learning model and innovative animal models as well as in vitro assays. 

Project 1 

Plasma membrane dynamics and signaling

Given that the cell membranes are the first defenders against external (sometimes even internal) stress, in theory, perturbations that may disrupt any type of cell membrane function should activate a multifactorial defense mechanism. We are studying how plasma membrane (PM) damage/repair and lipid mobilization happen and further affect the immune system. We also explore how this signaling evolves during system development. 

For PM repair, now we know ESCRT- III complex can repair the damaged PM by membrane remodeling, in many physiological settings. The broken PM fractions can be shed off. Therefore, cells can tolerate a limited level of PM damage and survive for a long time if ESCRT-III repairing capacity is not overwhelmed. As a result, ESCRT-mediated plasma membrane repair can bring cell back from the edge of death or preserve cell survival when pore-forming activity is sufficiently low (“sub-lethal”). 

The sub-lethal PM damage itself can be recognized and generate signals that result in the elaboration of secreted mediators, such as chemokines and cytokines. S660 p-PKCs sense the pore-forming damage by detecting the local Ca2+ influx, functioning as receptors to the transient plasma membrane integrity loss, leading to the downstream innate immune response (referred as, plasma membrane integrity signaling, PMI). The gene expression signature upon sub-lethal plasma membrane damage (that includes these chemokines) depends on the antagonization/delay in cell lysis affected by ESCRT-III. The roles of this cellular response for peritonitis, solid tumors, cross-priming, and kidney transplantations are also implicated. 

Moving forward, we will firstly hunt for more innate immune-sensing mechanisms for PM integrity loss. We proposed to test a promising ion channel candidate that is very likely to be one of the sensors for PM damage. We have also planned an elegant cell culture system to perform both siRNA and sgRNA screens to search for the PM damage sensors genome-wide. Next, we will use both animal models and human specimens to test the roles of the chemokines/cytokines secreted by these “survivors” in tumorigenesis, metastasis, and beyond. Third, we will try to eliminate the “survivors” directly by suppressing the ESCRT- III repair action

Project 2

Tumor cell death re-programmer

To achieve a robust anti-tumor response, we will find novel ways to “kill” those tumors in an immune system favorable way. To reprogram the cell death, we will use multiple CRIPSR technologies, including genome-wide CRISPR screening and CROP-seq (CRISPR then scRNAseq). Our ultimate goal is to treat cancer by modulating cell death program. 

One ongoing project regarding tumor cell death reprogramming is Phosphatidylserine (PS) externalization blockade. A “healthy” tumor cell has an asymmetric distribution of PS across the plasma membrane bilayer. Meanwhile, PS loses its asymmetry and redistributes to the outer leaflet during cell death. When PS is out, the interaction between PS-exhibiting tumor cells and immune cells elicits profound immune suppression. This response is important for tissue repair and inflammation resolution but needs to be reversed for anti-tumor immunity. 

Thus, our project will study the immune-suppressive roles of PS in tumor microenvironments and how we can OVERCOME that. To achieve this end, we have built up unique models to mobilize PS only on one side of the PM leaflets. We made tumor cells are alive and proliferating but with constant PS externalization (PSout); and tumor models, in which cells would not scramble PS during apoptosis (and the cell death is unaffected, PSin). PSout tumors grew bigger than WT while PSin tumors were more immunosuppressed. 

Project 3

Glucose-6-phosphate dehydrogenases (G6PD) for overcoming chemo drug-induced cell death resistance

By two rounds of genome-wide CRISPR based screen, we identified G6PD as crucial regulators in DNA damage-induced cell death. Our preliminary data suggested an exciting hypothesis that G6PD can “hijack” the Glucose-6-phosphate to pentose phosphate pathway thus limiting aerobic respiration, which is critical for tumor cell death induction. Our work will position G6PD as a novel target for overcoming the tumor cell death resistance in tumors with frequently identified p53, caspase-8, and/or Keap1 mutations, which were observed in lung cancers and many other types.

Project 4

Tumor cell death probing

Collaborating with physicians, our lab have collect probes to detect several major types of cell death in responding to all kinds of treatments, including those are still on the clinic trails. We can also set up real-time recording system for various tumor cell death. We provide insights into analyzing or predicting the therapeutic outcomes from a “cell death” angle.

Project 5

Macrophage responses to the innate immune activation pattern

With multiple pattern recognition receptors, macrophages can sense a variety of innate activation patterns. These patterns can come from bacteria, fungi, and virus, they can also come from tumors, especially dying tumors. We study how macrophages respond to various innate immune stimuli. We focus on two questions, how do the macrophage respond (molecular mechanism) and why do the macrophage respond so (pathological significance). 

Project 6

Machine learning-based AI technology

We are embracing the most updated Convolutional Neural Network technology to train AI models for imaging processing, sequence analysis and target discovery, and cellular function predictions. We have already successfully trained models named LANCE and D-MAINS for cell death, proliferation, and senescence predictions, using grayscale and label-free approaches. More exciting models tailored for other biomedical science are always on the way!  

More to explore...