Pictured above, Adolfo Ferrando (left), professor of pediatrics and of pathology and cell biology at Columbia, with Luis Arnes, associate research scientist and first-place winner of the symposium's poster competition; For photos from the symposium, visit the gallery page . Credit: Lydia Lee Photography

A multidisciplinary team of researchers across Columbia University have been busy addressing the complex challenges in basic and translational cancer research. Faculty and investigators are bridging their expertise in fields ranging from mathematics, biology, and engineering to physics, genomics, and chemistry to develop innovative approaches to better understand, for instance, cancer disease progression, drug resistance, and the systems-wide network of tumor evolution.

Central to this ongoing work is research grounded in cancer genomics and mathematical data analysis explored during a two-day conference Feb. 7-8 co-hosted by the National Cancer Institute (NCI) centers at Columbia University Medical Center, Cornell University, and Memorial Sloan Kettering Cancer Center. (Visit the Rabadan Lab YouTube Channel for video of the symposium).

"Genomics is becoming an important tool for the quantitative study of biological systems,” says Raul Rabadan, PhD , professor of systems biology at Columbia and director of the Center for Topology of Cancer Evolution and Heterogeneity and of the Program for Mathematical Genomics . “This meeting organized by four different NCI centers addressed some of the important challenges and new perspectives on the quantitative understanding of cancer using genomics tools.”

Three of the investigators in new Columbia, Broad Institute research collaboration aimed at gastric and esophageal cancer; L to R: Dr. Andrea Califano, Dr. Cory Johannessen, and Dr. Adam Bass (Johannessen image: Martin Adolfsson; Bass image: Sam Ogden/Dana-Farber Cancer Institute)

A research collaboration underway between Columbia’s Department of Systems Biology, the Broad Institute of MIT and Harvard, and Columbia University Medical Center (CUMC) is working to accelerate the discovery of new cancer drug combinations targeted at gastric and esophageal cancer. These tumors have not yet attracted prominent research focus and attention, and yet the general outcome for patients with these diseases is poor. According to the American Cancer Society, survival rates are only 20% at five years after diagnosis.

The newly formed research alliance between research teams at Columbia and at the Broad Institute came about thanks to a four-year gift by the Price Family Foundation, known for its philanthropic support of education, health, and biomedical research.

The Columbia-Broad team includes Dr. Andrea Califano , cofounder and chair of the Department of Systems Biology; Dr. Adam Bass , associate member of the Broad Institute; Dr. Cory Johannessen, senior research scientist at the Broad Institute; Dr. Josh Sonnett , the director of The Price Family Center for Comprehensive Chest Care, Lung and Esophageal Center at Columbia; and Dr. Naiyer A. Rizvi , the Price Chair in Clinical Translational Research at Columbia.

In 2016, the Price Family Foundation suggested that a team of scientists at the Broad Institute meet with researchers from CUMC. At the time, the Foundation was eager to leverage the project at the Broad—where researchers had uncovered an interesting finding for gastric and esophageal cancer—with innovative cancer systems biology work it was supporting at CUMC, focusing on the same diseases.

An example of tumor oncotecture. Transcription factors involved in the activation of mesenchymal glioblastoma subtype are shown in purple. Together, they comprise a tightly knit tumor checkpoint, controlling 74% of the genes in the mesenchymal signature of high-grade glioma. CEBP (both β and δ subunits) and STAT3 regulate the other three transcription factors in the tumour checkpoint, synergistically regulating the state of mesenchymal GBM cells. (Image: Nature Reviews Cancer)

In a detailed Perspective article published in Nature Reviews Cancer, Department of Systems Biology chair Andrea Califano and research scientist Mariano Alvarez (DarwinHealth) summarize more than a decade of work to propose the existence of a universal, tumor independent “oncotecture” that consistently defines cancer at the molecular level. Their findings, they argue, indicate that identifying and targeting highly conserved, essential proteins called master regulators — instead of the widely diverse genetic and epigenetic alterations that initiate cancer and have been the focus of much cancer research — could offer an effective way to classify and treat disease.

As coverage of the paper in The Economist reports:

ONE of the most important medical insights of recent decades is that cancers are triggered by genetic mutations. Cashing that insight in clinically, to improve treatments, has, however, been hard. A recent study of 2,600 patients at the M.D. Anderson Cancer Centre in Houston, Texas, showed that genetic analysis permitted only 6.4% of those suffering to be paired with a drug aimed specifically at the mutation deemed responsible. The reason is that there are only a few common cancer-triggering mutations, and drugs to deal with them. Other triggering mutations are numerous, but rare—so rare that no treatment is known nor, given the economics of drug discovery, is one likely to be sought. 

Facts such as these have led many cancer biologists to question how useful the gene-led approach to understanding and treating cancer actually is. And some have gone further than mere questioning. One such is Andrea Califano of Columbia University, in New York. He observes that, regardless of the triggering mutation, the pattern of gene expression—and associated protein activity—that sustains a tumour is, for a given type of cancer, almost identical from patient to patient. That insight provides the starting-point for a different approach to looking for targets for drug development. In principle, it should be simpler to interfere with the small number of proteins that direct a cancer cell’s behaviour than with the myriad ways in which that cancer can be triggered in the first place. (Read full article.)

PrePPI predicts the likelihood that two proteins A and B are capable of interacting based on their similarities to other proteins that are known to interact. This requires integrating structural data (green) as well as other kinds of information (blue), such as evidence of protein co-activity in other species as well as involvement in similar cellular functions. PrePPI now offers a searchable database of unprecedented scope, constituting a virtual interactome of all proteins in human cells. (Image courtesy of eLife.) 

The molecular machinery within every living cell includes enormous numbers of components functioning at many different levels. Features like genome sequence, gene expression, proteomic profiles, and chromatin state are all critical in this complex system, but studying a single level is often not enough to explain why cells behave the way they do. For this reason, systems biology strives to integrate different types of data, developing holistic models that more comprehensively describe networks of interactions that give rise to biological traits. 

Although the concept of an interaction network can seem abstract, at its foundation each interaction is a physical event that takes place when two proteins encounter one another, bind, and cause a change that affects a cell’s activity. In order for this to take place, however, they need to have compatible shapes and physical properties. Being able to predict the entire universe of possible pairwise protein-protein interactions could therefore be immensely valuable to systems biology, as it could both offer a framework for interpreting the feasibility of interactions proposed by other methods and potentially reveal unique features of networks that other approaches might miss. 

In a 2012 paper in Nature, scientists in the laboratory of Barry Honig first presented a landmark algorithm and database they call PrePPI (Predicting Protein-Protein Interactions). At the time, PrePPI used a novel computational strategy that deploys concepts from structural biology to predict approximately 300,000 protein-protein interactions, a dramatic increase in the number of available interactions when compared with experimentally generated resources.

Since then, the Honig Lab has been working hard to improve PrePPI’s scope and usefulness. In a paper recently published in eLife they now report on some impressive developments. With enhancements to their algorithm and the incorporation several new types of data into its analysis, the PrePPI database now contains more than 1.35 million predictions of protein-protein interactions, covering about 85% of the entire human proteome. This makes it the largest resource of its kind. In parallel with these improvements, the investigators have also begun to apply PrePPI in new ways, using the information it contains to provide new kinds of insights into the organization and function of protein interaction networks.

Assistant Professor Peter Sims and postdoctoral research scientist Jinzhou Yuan displaying their platform for automated single cell RNA-Seq. Photo: Lynn Saville.

RNA sequencing (RNA-Seq) has become a workhorse technology for research in systems biology. Unlike genome sequencing, which reveals a sample’s DNA blueprint, RNA-Seq catalogs the constantly changing transcriptome; that is, it itemizes and quantifies the complete set of messenger RNA transcripts that are present in cells at a specific time and under specific conditions. In this way, RNA-Seq makes it possible to investigate how the information encoded in the genome is functionally transformed into observable traits, and provides valuable data for defining and comparing different biological states.

Conventional RNA-Seq generates an average summary of mRNA abundance across all of the cells in a sample. Recent research, however, has created a demand for higher resolution technologies capable of generating mRNA profiles at the level of single cells. In cancer biology, for example, there is an increasingly acute awareness that gene expression in the cells that make up malignant tumors is highly heterogeneous. This suggests that in order to understand how the cells work together to drive a tumor’s cancerous behavior, scientists need better methods for characterizing the entire ecology of cells of which it is made. Being able to quantify differences in gene expression cell by cell could be one valuable way to explore such complex environments and understand how they sustain malignancy.

Although several single cell RNA-seq technologies have been unveiled in the past two years, they are expensive to operate and are not optimized to produce data on the scale that is required for systems biology research, particularly in tissue specimens with limited numbers of cells. In a new paper just published in the journal Scientific Reports, however, researchers in the laboratory of Department of Systems Biology Assistant Professor Peter Sims describe a novel approach that offers several important advantages over other existing methods.

The new, automated platform builds on previous innovations in the Sims Lab to offer a cheap, efficient, and reliable way to simultaneously measure gene expression in thousands of individual cells from a single tissue sample. Using custom designed microwell plates, microfluidics, temperature control systems, and software, the technology captures, tags, and generates a readout of the complete transcriptome in each cell, providing robust data that can then be analyzed to distinguish functional diversity among the cells in the sample. Already, the technology is playing a key role in several research projects being conducted in the Department of Systems Biology and promises to become even more powerful as the field of single cell genomics continues to evolve.