Treatments targeting the androgen receptor have extended the lives of people with advanced prostate cancer. At first, prostate cancers need androgens like testosterone to survive, and blocking these hormones sends tumors into a tailspin. These treatments push prostate tumors to the brink — but they’re not cures. Often, after developing androgen-independent ways to survive, prostate tumors recur, often with deadly results.  

To fight these tumors, we need to understand their survival strategies. New work, published today in eLife, gives scientists an unprecedentedly detailed look at individual cells as prostate cancer develops and turns treatment-resistant in mouse models of the disease.

“The idea was to create a cell-by-cell tapestry of prostate cancer progression,” said medical oncologist and prostate cancer researcher Andrew Hsieh, MD, who headed the project that was spearheaded by Hsieh Lab graduate student Alexandre Germanos. The scientists aimed to better understand the molecular underpinnings of cancer progression and resistance to androgen-deprivation therapy.

The single-cell studies identified a group of cancer cells in mice that share a gene-expression signature with prostate tumor cells from patients. These cells occur more often in tumors that develop resistance to androgen deprivation and correlate with reduced survival. Prostate tumors that recur after androgen removal are less uniform and have a wider variety of molecular characteristics. Germanos’ approach detailed the range of molecular characteristics of recurrent, castration-resistant cells.

“Alex’s work is really important in showing that that complexity exists and actually figures out some of the things that are driving this complexity,” Hsieh said.

Germanos’ data suggested that one way castration-resistant prostate tumors maintain this new complexity is through higher rates of protein synthesis. They found that blocking protein synthesis killed off the cell type associated with more aggressive, treatment-resistant disease. The findings highlight protein synthesis as a potential treatment target and suggest that oncologists hoping to prevent or counter treatment-resistant advanced prostate cancer explore therapies that reduce tumor complexity.

The team has also made their data publicly available to ensure it’s widely accessible — and easily analyzed — through an interactive data visualization tool.

Bypassing the androgen receptor: An emerging concern

Nearly everyone diagnosed with early-stage prostate cancer will survive it. But more advanced prostate tumors, particularly tumors that have spread through the body, are killers. While more than 99% of men with localized prostate cancer will be alive five years after diagnosis, only 31% of people diagnosed with prostate cancer that has spread will survive that long.

For decades, patients with advanced disease have received androgen-blocking therapies, which dramatically extend their lifespans. But advanced tumors eventually develop resistance to this strategy and develop new ways to get the androgen they need to grow. In the last decade, more targeted treatments beat back these recurrent tumors and offered patients more survival benefits.

Even with these better treatments, prostate tumors recur. In 2017, a Fred Hutch team led by prostate cancer doctor and researcher Pete Nelson, MD, who holds the Endowed Chair for Prostate Cancer Research and also contributed to the current study, reported that more and more of these recurrent tumors are dispensing with androgens and the androgen receptor. Instead, they’re activating new, androgen-independent survival strategies. In 2019, Hsieh’s group described one strategy that some prostate tumor cells use to bypass their need for androgens.

“One of the hypotheses is that, as we’ve gotten better at targeting the androgen-signaling pathway, the cancers have had to adapt and a higher proportion of advanced prostate cancers are just not using androgen signaling at all,” Germanos said. “Androgen receptor-low, castration-resistant prostate cancer is one of the subtypes of advanced prostate cancer that we’re trying to find out more about so that we can find ways and means of treating it.”

Much remains to be discovered about androgen receptor-low, or AR-low, prostate cancer, Hsieh said.

“The clinical problem we’re finding is that AR-low prostate cancer is not one disease entity,” he said. “That’s a huge problem for patients. If the cells aren’t all alike, how do you treat them?”

On top of this, prostate cancer cells can toggle back and forth between different cell states that change their behavior and response to various treatments. They become a collection of disparate, moving targets.

“We really have to answer the fundamental question of how and why (these cells are so different and flexible),” Hsieh said.

Most prior work on cancer essentially mashed all the molecular information from all the cells in a tumor together. But that’s like a nearsighted person trying to learn about a tree by peering at it without glasses. Eyeglasses help bring individual leaves and branches into focus. To do the same for prostate cancer, Germanos turned to single-cell RNA sequencing, which gives snapshots of which genes are turned on and off in individual cells.

Complexity: prostate cancer’s secret weapon

Germanos used single-cell RNA-seq to compare cells from normal prostates, prostate cancer, and castration-resistant prostate cancer. He turned to mouse models that lack a gene in prostate cells that usually acts to check excessive cell growth. Without this gene, eventually some prostate cells will turn cancerous.

“The goal is to see how all of these cell populations in the prostate evolve and progress as the cancer itself progresses and see if there are any hints as to what AR-low prostate cancer needs to survive,” Germanos said.

By comparing cells at different stages of cancer progression, Germanos was able to identify a distinct cell population that arose in prostate cancer. After androgen removal pushed prostate tumors to develop castration resistance, the group of cells matching this type expanded. This group of cells also grew less similar. Germanos saw a wider range in the level of androgen receptor RNA, and a wider range in the levels of RNAs of genes involved in cell proliferation. But some characteristics appeared to align: a group of cells that relied less on androgens also proliferated more.

He also found links to human prostate cancer. Germanos compared key genes turned on in the cells he’d characterized to genes expressed in human prostate tumors. He saw that tumors that shared this gene signature were more resistant to androgen deprivation and progressed more quickly than tumors without the signature.

Germanos’ single-cell approach allowed him to look at non-cancerous cells as well. Prostate tumors generally don’t respond well to immunotherapy. Germanos’ findings suggest this is likely because prostate tumors attract types of immune cells that shield them from further attack. He also saw that tumors ramped up proteins that dampen immune activity, suggesting that strategies that reverse this immunosuppression or remove the tumor-shielding immune cells could improve prostate tumors’ responses to immunotherapy.

Reining in protein synthesis to reduce complexity

Germanos noticed that, after androgen deprivation, prostate tumor cells that cranked up proliferation genes also cranked up genes involved in protein synthesis, also known as RNA translation. This suggested to the scientists that castration-resistant prostate tumor cells could be relying on higher rates of protein synthesis to maintain their diversity.

When the team used a genetically modified mouse model to block protein synthesis, they saw that castration-resistant prostate cancer cells withered and died.

“This suggests complexity as target,” Hsieh said.

A uniform set of cancer cells will be vulnerable to the same therapy, making them easier to treat.

Further exploring strategies to rein in complexity

Whether the cell population that Germanos identified in mice and people arises from a group of cells that multiplies, or from other cells taking on these characteristics, remains to be determined, the researchers said. They currently speculate that cells that come from different types of prostate tissue may be adopting similar characteristics that make them more aggressive and able to thrive without androgens.

“One of the potential future uses (of detecting these cells in patients) would be as a tool for prognosis and diagnosis,” Germanos said. “Perhaps if you can detect that cell state in a patient, maybe you know he’ll be less likely to respond to androgen deprivation — but that idea still needs to be tested in patients.”

He also noted that there’s also more to learn about the relationship between RNA translation, cell proliferation and cancer initiation. Hsieh’s group is also working to better understand the molecular and genetic underpinnings of difference in protein synthesis across individual cells.

The idea of blocking protein synthesis is not unprecedented: several protein synthesis inhibitors are currently being tested in clinical trials and one, omacetaxine mepesuccinate, or Synribo, was approved by the Food and Drug Administration for treatment of chronic myeloid leukemia.  

But protein synthesis is likely just one method that castration-resistant prostate cancer cells have to stay flexible, diverse and androgen-independent, Hsieh said. And rather than singling out a single molecular target, Germanos’ findings highlight the potential power of any therapy that could corral prostate cancer-cell chaos.

“It really points to the idea of, can we decrease the complexity?” Hsieh said. “It’s like, there are 31 flavors. Can we get that down to three? That would be easier to treat.”

This story was published by Fred Hutch on December 13, 2022. It is republished with permission.