Q+A: Michael Shen, PhD, and the Evolution of Prostate Cancer Research

Michael Shen’s interest in prostate cancer research dates back to the late 1990s when, as a postdoctoral fellow in the late Philip Leder’s laboratory at Harvard Medical School, there was often discussion about how little progress had been made in the field at the time.

Now, prostate cancer research has exploded thanks to a massive increase in funding and a societal shift—it has become more and more acceptable for patients to be vocal about their diagnosis rather than keeping it secret. While 1 in 9 men are diagnosed with prostate cancer in their lifetime, improved treatment options and early prevention strategies for the disease have helped increase survival rates, and the funding vehicles dedicated to prostate cancer research has been instrumental in bettering our understanding behind the biology of the disease.

As a developmental biologist, Dr. Shen’s focus in elucidating the biological mechanisms of prostate cancer continue to be central to his research. In a new paper published in the journal eLife, Dr. Shen and collaborators at Weill Cornell identified distinct cell populations that are conserved between the mouse and human prostate, and advanced our understanding of the heterogeneity of cell populations responsible for prostate growth and are potentially susceptible to oncogenic transformation. His laboratory’s work with three-dimensional cultures to study prostate and bladder cancer could contribute to the development of new therapies and uncover key new insights to the molecular mechanisms responsible for tumor evolution and drug response.

Dr. Shen is co-leader of the Tumor Biology and Microenvironment research program at the Herbert Irving Comprehensive Cancer Center (HICCC) and professor of medical sciences, of genetics and development, and of systems biology at Columbia University Vagelos College of Physicians & Surgeons.

What have been the biggest moments for the prostate cancer field?

What has changed in prostate cancer research is a confluence of a number of factors, including being able to attract leading investigators from other fields, a cultural and societal shift which led to more advocacy and less people feeling shameful about discussing their diagnosis, and a large increase in research funding.

If you look back 25 years there was only one treatment available for prostate cancer patients, and that was androgen deprivation therapy. This was often successful but that was largely it. It had been difficult to make advances in studying prostate cancer.

Why has it been difficult?

There were very few mouse models, and there were very few cell lines. As it turns out, it’s very difficult to grow prostate cancer in a dish. Everybody studied the same five cell lines, and this remains largely true today.  And there was very little interest around the activity in this field. It was perceived to be a dead-end field, that there was not going to be a lot of progress made, so it did not attract interest among top cancer researchers.

All of that has changed. Now it is one of the most active fields of cancer research, whereas it used to be sort of a “follow-the-leader” field. Advances were made in other areas of cancer biology and then researchers in the prostate cancer field would get around to making similar findings. But, now it is frequently the other way around, where the groundbreaking discoveries are often made in studying prostate cancer.

Prostate cancer began attracting top investigators such as Charles Sawyers who is at Memorial Sloan Kettering. He developed a second-generation anti-androgen therapy which is tremendously successful in the clinic. There was an influx of new researchers to the field, and I would include Cory Abate-Shen and myself among them. There was an expansion of grant mechanisms to support these researchers and that was in part due to the DOD Prostate Cancer Research Program and the establishment of the Prostate Cancer Foundation (PCF). The explosion of research in prostate cancer could be attributed in part to PCF. PCF has supported a lot of investigators over the years, and has been instrumental in the development of many therapeutics along the way.

You also mentioned that there was a societal and cultural shift in prostate cancer that helped advance the field.

Prior to about 25 years ago prostate cancer was a “hidden” disease – it was not talked about. If you look at the transformation of public attitudes towards breast cancer, I would say 30-40 years ago, women did not talk about breast cancer but that changed. Now you have Angelina Jolie publicizing how she’s having prophylactic double mastectomy. Years ago prostate cancer was in the same category – it was shameful, and you didn’t talk about it. That changed in part because of Norman Schwarzkopf. General Schwarzkopf led Desert Storm and he came out and said he had prostate cancer. This was an amazingly important thing because it led the way for men to say they have prostate cancer. It’s acceptable now to talk about prostate cancer and that makes a huge difference with respect to fundraising and advocacy.

You mentioned it is difficult to grow prostate cancer in a dish, but in recent years, your laboratory has been experimenting with organoid systems to study drug response and cancer biology. What are organoids?

Organoids are simply clusters of cells that are grown in three-dimensional culture conditions, and with appropriate media conditions, these clusters of cells can recapitulate many of the features of a tissue. Organoids can have many differentiated cell types. They typically are propagated by cells with stem cell-like properties and in many cases these cultures can be propagated indefinitely. There is an immediate appeal to this type of system because it overcomes limitations of two-dimensional cell lines that can do a limited number of things but cannot recapitulate tissue architecture or the complexity of having different cell types interacting with each other. Organoids also can circumvent some of the problems in working with mouse models, which of course require mice, and consequently, are more long-term experiments.

How did you start working with organoids and how are they unique?

Our interest in organoids grew out of our longstanding focus in developmental biology. When organoid approaches started becoming popular again – actually organoids have been around for decades but regained popularity about ten years ago—we realized immediately the potential significance that these sorts of approaches might have for studying not just development but also cancer.

Even if we’re using genetically-engineered mouse systems, you can do things a lot faster in organoid culture. But as we’ve seen in our work and that of others, you can also recapitulate a lot of heterogeneity in organoid culture. That’s a property inherent in organoids. Organoid systems allow you to recapitulate, or mimic, the heterogeneity in a tumor. That’s particularly advantageous when you’re studying human tumors. This allows you to address problems that are otherwise difficult to access and the ability to grow human tumors in a dish means, again, you have the ability to do all kinds of experiments that you otherwise could not do. For example if you’re interested in testing a large number of drugs using traditional xenograft approaches, you need a large number of mice. With organoids, you can screen a lot more drugs more readily because this is done in cell culture.

How are you using organoids in prostate cancer research?

With our prostate cancer research we have a different type of problem. To date, it’s been very difficult to generate organoid lines from human prostate tumors. It has been done by several groups, but the efficiency is low. This is something that presents a technical hurdle that the field has not yet overcome. So it’s quite different from what many groups are doing with other cancer types, or what we’ve been able to accomplish in bladder cancer.

With respect to prostate cancer, we are developing organoids from genetically engineered mouse models. We started with a genetically engineered mouse model that was developed originally by our collaborator, Dr. Abate-Shen, that we published together a few years ago, which we showed was a model of neuroendocrine differentiation in prostate cancer. In particular, we showed that neuroendocrine cells arose by transdifferentiation from luminal cells. Most likely, in these advanced prostate tumors the luminal adenocarcinoma cells are changing their cell type to become neuroendocrine. This form of plasticity is what our lab is most interested in, and we’re using organoids to study plasticity in prostate cancer, and also in bladder cancer.

In the case of prostate cancer this is linked to a problem we know is of clinical significance—the plasticity that is involved in castration-resistant prostate cancer leads to neuroendocrine differentiation, which is a major topic in the prostate cancer field right now. Many groups are trying to study this. We think we have a different approach, utilizing our organoid systems, that can potentially provide key insights into the molecular mechanisms.

How will this ultimately impact cancer research?

Organoids have the potential to be revolutionary in cancer research, but it remains to be demonstrated that drug response in organoid culture can mimic what happens in human patients. This is something that a lot of laboratories are interested in investigating with their own systems.

There’s now a fair amount of evidence from different cancer types, most notably colorectal and pancreatic, that drug response seen in organoids can correlate with responses seen in patients. But these are retrospective studies. They lay the foundation for prospective studies, which are much more challenging—organoids are made, their drug response is studied, and then the patient is treated based on the results of those studies. That kind of experiment is much more challenging and requires a clinical trial with a lot more ground work and approvals.

We think that the organoids actually model key features of what’s going in the human tumors. Obviously they aren’t a perfect model, but we think that the key features of neuroendocrine differentiation in prostate cancer, for example, are similar in organoids, and the important point is that in this system, we can study what’s going on in great detail. We can perform biochemical experiments and molecular experiments to alter gene function to pursue functional studies, and then of course for the key findings, we can go back to human tissue samples and try to validate them.