Final Report: The Princeton Physical Sciences Oncology Center

Complex Ecologies for Explosive Evolution of Interacting Cancer Cell Populations

Robert Austin, Princeton University, Princeton, NJ, USA.

1. Summary

The Princeton Physical Sciences Oncology Center failed as a Center but succeeded in developing several new ideas and technologies we think in the coming years will be recognized as fundamental novel and important ideas which were not given a chance to grow. We developed a novel micro-engineered microfluidic cell culture device that resembles the in vivo landscapes of stress heterogeneity in a tumor. With this “evolution accelerator” we were able to recapitulate the adaptive cellular response to a heterogeneous microenvironment as well as investigate interactions amongst prostate cancer of cells in this tumor-like community. The technology not only can be used to investigate various fundamental biological questions from an ecological point of view by analyzing the population dynamics of multiple cell types in the microenvironment, but also has a potential to work as a platform for preclinical drug development and assays of likely performance.

2. Physicists Tilting at Windmills with Expected Results!

The Princeton Physical Sciences Oncology Center (PPSOC) was one of the most heavily “physics centric” of the 12 Physical Sciences Oncology Centers (PSOCs) funded by the National Cancer Institute. Unlike many of the other Centers, it spanned the United States, with members at Princeton University, Johns Hopkins Medical Institute, University of California San Francisco, University of California Santa Cruz, and the Salk Institute. It represented a mixture of physicists, electrical engineers, oncologists and biochemists. It was made de novo, that is, none of the branches had ever worked together, the physicists and engineers knew nothing about cancer except it seemed to be killing members of their families at enormous cost quite relentlessly. So, it represented what was the initial intent of the PSOC effort done with great energy and courage by Ann Barker, and led by Larry Nagahara and Jerry Lee.

The PPSOC from the start was deliberately high risk and high reward: the mission was to rethink cancer as an intrinsically evolutionary phenomena, not a disease in the normal sense of the word as an invasion by some foreign entity to fought and destroyed, but rather primarily a condition of inappropriate growth of the body's own cells, including invasion of tissues with the body remote from the local lesion: metastasis. But not a disease. The approach driven by the physics arm of the PPSOC was to stress four main aspects of this condition: (1) The importance of the local ecology within which the cancer cells are growing; (2) the importance of a stress landscape within the ecology which drives part of the heterogeneity of the cancer tumor; (3) The importance of the role of the number ni of individuals within a local population within the overall ecology; (4) the importance of understanding interactions kijninj between different subpopulations with the overall ecology, which brought in aspects of game theory.

Why did we fail? Probably the hardest part was bridging the enormous cultural, scientific and ideological chasms that separated the different parts of the Center. It is one thing to talk about multidiscipline work, quite another to do it at an equal footing without one discipline becoming subservient to the other. In the case of the PPSOC it was extremely difficult for the physicists to change the path and direction of the already powerful and directed oncology arms of the Center, especially when the oncology arms did not exactly get along with each other, which became very clear at the start. Even between the physicists and engineers there was an inner tension: engineers tend to want to build things using already developed technologies which are understood if not yet mature, while physicists tend to want to be the creators of new ideas and concepts which MIGHT become useful technologies down the road, but not right now.

A consequence of this split was that the Center, while there were at least weekly SKYPE meetings between the branches and at least bi-monthly trips to the various branches, at times resembled a cancer-based United Way, where funds were sent from the Princeton hub to the branches, and the branches used these funds to continue their own previously established work with no real change in their direction. This is not really a criticism but rather a statement of fact about how hard it is to change the direction of an already strong group, especially if the changes come from a branch which has no expertise in the field, but perhaps foolishly thinks it knows a better way to attack the problem of cancer resistance.

Another aspect of the problems the PPSOC faced was the fundamentally different way that physicists and oncologists carry out research. Physicists tend to study the simplest system they can that illustrates a fundamental property they are trying to understand, while oncologists are forced to study an extremely complex system which changes by the minute and varies greatly from sample to sample. Also, there is in oncology enormous financial aspect to the consequences of the research. The combination of these two factors results in the perhaps not well known fact that of 53 "landmark" papers in oncology only 6 were found reproducible even with cooperation of the original authors [1]. In the PPSOC the poor reproducibility of the oncology literature bit us: one of the principles in the Center had to withdraw after the company he founded could not reproduce his own data, and later the seven core papers were retracted. Although this happens in physics too [2], it is much more common in oncology and you have to be prepared for the huge uncertainties, the lack of reproducible data, and the premature drive to the clinic and Big Pharma.

It was tough sledding, but I don't want to project the impression that all was darkness. All the branches of the PPSOC worked very hard to overcome the chasms that separated us, and while ultimately there was no real cohesion there were some real achievements that I believe in about ten years will lead to major new ways we view the origins and progression of cancer. It simply takes a long time to change the course of experimental science, particularly in biomedicine. Too bad so much is actually at stake as the ship plows on ahead full speed into the icebergs.

3. Progress After the Fall

If you believe in what you are doing you don't quit. “Illegitimi non carborundum” Since we have a belief that what this Center started was too good to be to “carborundum'd” by peer review, which is of course basically inevitable, we have gone underground to keep alive the Dream, but we now carefully avoiding spinning windmills in the dark, and avoiding the grinding teeth of peer review as much as possible.

We had a strong feeling about physics, ecology and cancer when we began our effort in 2010. We believed that there is a deep connection between ecology and the physics of living systems. John Dunne wrote evocatively in 1623 in his Devotions Upon Emergent Occasions: “No man is an Iland, intire of it selfe; every man is a peece of the Continent, a part of the maine". While Dunne probably had a different use of the word “Emergent” than modern physicists such as Phil Anderson use, we take those words seriously:“emergence" for us means the revelation of unexpected and not reductively predictable collective behaviors of biological agents as they exist as communities in complex ecologies. We also take seriously the later words of Charles Darwin, who famously noted the complexity of life in the “tangled bank, clothed with many plants of many kinds, with birds singing on the bushes, with various insects flitting about, and with worms crawling through the damp earth.”Life exists in this tangled bank and in an entangled state, the question is can we deal with that complexity quantitatively?

We expected that if we can build sufficiently complex ecologies even for supposedly “simple" organisms such as bacteria, or certainly as complex as cells in human tissue, emergent and quantitative principles of complexity will be made clear. Rather than passively observing the “tangled bank" of the classical ecologist we construct using the tools of the semiconductor industry micro-fabricated environments where we can control the fitness landscape for communities of biological agents, while still allowing for rich complexity and self-sculpting of the landscape by multi-component organisms at high densities.

Cancer develops within a large ecology, the human body. Within that large ecology, tissue based cancers progress in a complex local ecology which recapitulates Darwinian natural selection amongst different cell types, both cancer cells and non-cancer cells, in a confined volume. The complex stress landscape within which a confined cancer population develops applies spatial and time dependent selective pressures which present spatially varying opportunities for genetic and epigenetic transmission to daughter cells. If the daughter cells are cancer cells and have higher fitness under stress than the mother, the cancer progresses. While animal models can to a certain degree reproduce this ecology that drives cancer progression, better ex vivo ecology models are needed for quantitative studies of evolution under high-stress conditions both to predict the progression of cancer and test the efficacy of drugs under high stress and cellular heterogeneity. On a time scale of weeks we can now demonstrate what normally takes months or years of evolution under metabolic stress. Figure 1 shows the competing growth of prostate cancer cells (red: PC3 epithelial and green:PC3 EMT) cells over a period of two weeks in a chemotherapy gradient (docetaxel). By tracking the behaviors of multiple cell types in the device, including the proliferation rate, population dynamics, cell motility, biosensor activity and the composition of metabolic waste, the technology we have developed could potentially work as a tool to investigate various fundamental biological questions from an ecological point of view, as well as a platform for pre-clinical drug development, or even pre-clinical invitro experiments that allow personalized therapy selection in cancers.

cancer cells

Figure 1: The growth of competing cancer cells in a complex ecology in adrug gradient.

4. Conclusion

Our attempt to renew our Center was met with extreme vetting: our accomplishments were deemed not worthy of being discussed amongst the wise and learned panel cognoscenti. I guess I should not have been surprised. Sitting in on another NIH Panel after the fall, it was pointed out by a reviewer that a proposal which posited that evolution can predict the course of a cancer must surely be wrong. Why? Because suppose that patient steps into traffic and is killed by a car: well, evolution didn't predict that did it? This line of reasoning seemed to meet with sage nods of approval. You can see how hopeless it can be. There is a Chinese saying: Don't go to the front gate of the Kung Fu Master Bodhidharma to show off your Kung Fu skills. Perhaps we did that and were punished, but the oncologists did not all seem like Kung Fu masters. We really did try to change things in the cancer community using ideas from physics. I think we did come up with new ideas, but the results of that work lie in the future, and I am trying to make that day happen. One lesson, however has been that there is some truth to the adage that NIH stands for Not Invented Here.


[1] Berry, C. Reproducibility in experimentation - the implications for regulatory toxicology. Toxicology Research 3, 411{417 (2014).

[2] Reich, E. Plastic Fantastic (Macmillian, 2009).

These contributions have not been peer-refereed. They represent solely the view(s) of the author(s) and not necessarily the view of APS.