Over view

Our laboratory investigates normal and cancer epithelial cells within a three-dimensional, organ-like, context to identify new pathways that regulate organ morphogenesis and stem cell differentiation, and to use that knowledge to identify new strategies to control and prevent cancer. We accomplish this goal by developing and using organoid culture systems to grow normal or tumor epithelial cells from organs such as breast, pancreas and prostate as mini-organs. These mini-organs bridge the gap between monolayer cell culture and in vivoconditions and provide a tractable model system to understand dynamics of cell behavior during both tissue morphogenesis and cancer cell growth. 

 Using organoid cultures, we have identified cell polarity proteins such as, SCRIB, PARD3 and PARD6 as key regulators of tumorigenesis regulated by oncogenes such as ERBB2 and MYC. These cell polarity proteins regulate cell death and cell-cell cohesion pathways and polarization changes during initiation of tumor and progression to metastasis. Research projects in our laboratory continues to use 3D organoid cultures to identify how cell polarity proteins regulate signaling pathways involved in cell junctions, cell differentiation and epithelial to mesenchymal transition during cancer, initiation, progression and response to therapeutic drugs.

 More recently, have developed new methods to generate pancreatic epithelial organoids from human pluripotent stem cells. This normal organoid platform creates opportunities for us to investigate signaling and cell polarity mechanisms regulating normal human pancreas development in culture and to model benign human pancreatic diseases such as pancreatitis. 

 Our ability to grow primary pancreatic tumor cells as tumor organoids has empowered us and to identify new therapeutic options for pancreatic cancer. Using tumor organoids, we discovered that drugs against the epigenetic regulator, EZH2, is a potential therapeutic target for pancreatic cancer and efforts are currently underway to initiate a clinical trial to translate this finding towards patient care. Last, but not least, we are expanding the tumor organoid technology to other types of solid tumors such as prostate to broaden the scope of the method. We believe that tumor organoids are likely to serve as a powerful, personalized, platform for understanding drug response and resistance mechanisms in a patient-specific manner and to translate the findings towards caring for cancer patients, one patient at a time.  



We use a combination of 3D cell culture and mouse models to understand the role played by cell polarity pathways during normal epithelial cell morphogenesis and during transformation of polarized epithelial cells.

The PAR Cell Polarity Complex:

The PAR polarity complex cooperates with the receptor tyrosine kinase HER2/ERBB2 during cancer initiation and metastatic progression.

We have discovered that the PARD3/PARD6/APKC polarity protein complex is a downstream effector of the receptor tyrosine kinase HER2/ERBB2 during transformation of polarized 3D breast epithelial structures. In addition, we have demonstrated that loss of PARD3 cooperates with HER2 to promote metastasis without inducing EMT but, by modulating E-Cadherin junction maturation and decreasing cohesion between HER2 expressing tumor cells. Moving forward, we are investigating how other oncogenes interact with cell polarity proteins and we are also collaborating with bioengineers to investigate the molecular mechanisms that regulate cell-cell cohesion and its implications for metastatic behavior transformed epithelial cells.

We demonstrated that the human epidermal growth family receptor 2 (HER2/ErbB2) oncogene interacts with and disrupts the PARD3-PARD6-aPKC cell polarity complex (Nat Cell Biol, 2006, 8: 1235-45). An intact PARD6-aPKC complex was required for the ability of ErbB2 to transform 3D acini, whereas, it was not required for inducing aberrant proliferation. Thus cell polarity and cell proliferation pathways can be uncoupled from each other during initiation of transformation of mammary epithelial cells. More recently, we demonstrated that the cell polarity proteins function as metastasis suppressors (Nat Cell Biol. 2013, 15:189-200).  We showed that loss of Pard3 cell polarity protein induced invasion and metastasis in vivo, by affecting cortical actin dynamics, decreasing E-cadherin junction maturation and cell-cell cohesion without inducing any detectable epithelial to mesenchymal transition. These studies are beginning to define a framework for how polarity proteins/pathways regulate oncogene-induced transformation of polarized breast epithelial cells. 

Scribble, Lethal Giant Larvae Cell Polarity Proteins:

We have demonstrated SCRIB functions as a tumor suppressor in mammals both dues loss of expression and due the mislocalization from the cell membrane. The latter identifies an unexpected biology where the misclocalized results in a gain-of-function phenotype and not simply as another inactivation mechanism. Moving forward, we use a combination of inducible RNAi mouse model, 3D cell culture and proteomics to understand how SCRIB and LGL proteins regulate the cell biology of normal and cancer cells.

In addition to cooperating with oncogenes, cell polarity proteins by themselves functions as modulators of cancer biology. We published the first report (Cell 2008, 135:865-78)that the cell polarity protein SCRIB functions as a tumor suppressor in mammary epithelia by inhibiting cell polarity, blocking three-dimensional morphogenesis, inhibiting apoptosis and inducing dysplasia in vivo. Subsequent studies by others have expanded on this observation and demonstrate that inactivation of SCRIB functions as a tumor suppressor in prostate and lung, implicating SCRIB as regulator of biology of multiple cancers.

However, we realized a paradox, SCRIB is frequently amplified (up to 30%) and mislocalized, (not lost/downregulated), in multiple carcinoma including breast, ovary, prostate, lung and head and neck. We recently reported (Cancer Res. 2014, 74:3180-94)that SCRIB interacts with PTEN and that expression of a mislocalized form SCRIB, but not the wild type protein, decreases membrane localization of PTEN and results in activation of PI3K signaling. Transgenic mice overexpressing a mislocalized form of SCRIB develop tumors with basal characteristics, demonstrating that mislocalized SCRIB can have a positive role during tumorigenesis. These studies provide a new perspective on the role on cell polarity proteins in cancer that they may not only function tumor suppressor due to loss-of-function, but can also have a gain-of-function by acting as a neomorph. 

Surprising Insights Into Epithelial Morphogenesis:

 We use multidimensional live-cell imaging analysis to track the morphogenetic process starting from a single cell to the development of a multicellular, spherical structure composed of polarized epithelial cells surrounding a hollow lumen (Proc Natl Acad Sci, 110(1):163-8).In this study we discovered that in addition to actively maintaining apico-basal polarity, the structures rotated 360 degrees every four hours (See videos below). This rotational motion was required for assembly of endogenous basement membrane matrix around the 3D structures suggesting laminin matrix assemble as a weaving process. Neither cells lacking cell polarity proteins (SCRIB or PARD3) nor cancer-derived cells rotate in 3D culture. Thus laminin matrix assembly in vivo is likely to involve a mechanophysical step that is lost in cancer-derived epithelial cells.



In 2001, we reported the establishment and use of a three-dimensional (3D) culture model for breast aciniar organogenesis using a human mammary epithelial cell line MCF-10A (Nat Cell Biol. 2001, 9:785-92). In 2015, we reported the established of 3D models for human exocrine pancreas organogenesis and to grow patient-derived primary tumor cells as tumor organoids (Nat. Med. 2015, 11:1364-71). We use these organoid models to understand new mechanisms regulating normal morphogenesis, tumor cell biology and discover new therapeutic strategies.

Normal Organoids

 Three-dimensional (3D) cell culture models of human epithelial organogenesis are powerful platforms for understanding mechanisms of normal organogenesis and for modeling organ-specific disease in a culture setting.

We generate breast epithelial acini-like structures using MCF-10A, an immortalized human mammary epithelia cell line. This has been a robust experimental platform to understand how epithelial cells grow into proliferation-arrested 3D epithelial structures with a hollow central lumen; how cell polarity is established and maintained during epithelial morphogenesis; how extracellular matrix is assembled during morphogenesis and understand mechanisms of tumor-initiation and progression. 

 More recently we developed a new method to induce formation of pancreatic ductal and acinar epithelial structures from human pluripotent stem cells. These exocrine cells generate three-dimensional dutal and acinar structures with cell-type specific biological traits both in culture and when transplanted in to mice. These models are being used in our laboratory for understanding the role played by cell polarity pathways during exocrine organogenesis and for modeling pancreatic diseases such as pancreatitis and tumor-initiation.

  We use both mouse models and 3D cell culture to understand mechanisms by which cell polarity and hormone signaling cooperate during epithelial morphogenesis. We discovered that overexpression of PARD6b increased the size of the 3D mammary acini by activating ERK signaling demonstrating that cell polarity proteins can regulate mechanisms that govern organ size (Cancer Res. 2008, 68:8201-9).To better understand the morphogenesis, we performed multidimensional live-cell imaging analysis to track the morphogenetic process starting from a single cell to the development of a multicellular, spherical structure composed of polarized epithelial cells surrounding a hollow lumen (Proc Natl Acad Sci, 110(1):163-8).We find that in addition to actively maintaining apico-basal polarity, the structures underwent rotational motions at rates of 15 - 20 microns per hour and the structures rotated 360 degrees every four hours during the early phase of morphogenesis (See videos below). This rotational motion was required for assembly of endogenous basement membrane matrix around the 3D structures, and that structures that failed to rotate were defective in weaving exogenous laminin matrix. Neither cells lacking cell polarity proteins (SCRIB or PARD3) nor cancer-derived cells rotate in 3D culture. These studies identify a novel mechanophysical process observed during normal 3D morphogenesis that regulates laminin matrix assembly and is lost in cancer-derived epithelial cells.

 Moving forward we are using inducible RNAi mouse models of SCRIB to understand how cell polarity pathways regulate normal developmental processes.

 Tumor OrganoidIn 2015 we developed a powerful methodology for growing primary patient tumor tissue in a three-dimensional culture condition as tumor organoids (Nat. Med. 2015, 11:1364-71). These tumor organoids – “mini tumors” – recreate histoarchitecture, differentiation status and epigenetic status of the matched patient tumor. Importantly, these patient-derived tumor organoids maintain both inter-tumor and intra-tumor heterogeneity for all the above traits. We can established tumor organoids with >80% take-rate. They can be serially passaged, genetically modified, test for drug response and used for establishing xenograft models. We believe that these tumor organoids are a robust platform for target validation and target discovery efforts. Using this technology, we have discovered that inhibition of a chromatin regulating enzyme, EZH2, blocks growth of some, but not all pancreatic tumor-derived organoid culture. This finding raises the possibility that EZH2 inhibitors (EZH2i) are likely to be a therapeutic drug for treating patients with pancreatic cancer, a possibility that we are pursuing for clinical translation.

Every patient’s tumor is unique for its molecular traits, hence, matching patients to therapies is likely to improve our chances to control cancer. The ability to grow a patient’s tumor ex vivofor pre-clinical or co-clinical (analysis performed in conjunction with clinical care) is emerging to be a powerful strategy to match the patient tumor genotype to the clinical response phenotype. The tumor organoid technology has created an unprecedented opportunity to keep primary patient tumors cells alive for extended periods of time and use them to design treatment plans, one patient at a time.