Exploring the Tumor Microenvironment

Introduction

Why do some cancers metastasize to infiltrate distant organs and tissues? And why are some metastatic tumors drug resistant? These have been extremely difficult questions for researchers to answer. Many have looked to the tumor cells themselves for the cause. However, it’s become clear that the inner workings of these cells don’t fully account for the behavior. Researchers have expanded the scope of their studies, looking into the role elements of the tumor microenvironment might play. This includes blood vessels, immune cells, fibroblasts,  and the extracellular matrix surrounding tumors. The research is yielding dramatic insight into how and why cancer cells migrate to other parts of the body and become drug resistant.

Associate Professor Alex Swarbrick, PhD is the head of the Tumor Progression Laboratory at the Garvan Institute of Medical Research in Sydney, Australia. His research uses single-cell techniques, such as single-cell RNA sequencing (scRNA-Seq), to investigate tumor microenvironments in breast and prostate cancers. The goal is to elucidate the patterns of gene expression within them and identify potential targets for future cancer treatments.

iCommunity spoke with Dr. Swarbrick about the advantages of single-cell next-generation sequencing (NGS) techniques and how an understanding of the tumor microenvironment could be vital for developing a stronger cancer-fighting therapeutic arsenal.

Alex Swarbrick, PhD is the head of the Tumor Progression Laboratory at the Garvan Institute of Medical Research.

Q: When did you join the Garvan Institute?

Alex Swarbrick (AS): I’ve been at the Garvan Institute for 12 years. I was previously a postdoctoral fellow at the University of California, San Francisco. My work at the Garvan Institute focuses on cellular genomics. I’m using single-cell methods to understand human disease. One of the exciting things about cellular genomics is that we can go straight to humans to make fundamental and translational discoveries. We then use experimental models to work backwards and help us identify mechanisms of disease.

Many labs realize the opportunities that cellular genomics affords in providing new insights into disease, whether its cancer, autoimmune disorders, immunology, or neurobiology. Cellular genomics studies are being conducted in many labs at the Garvan Institute. The Institute also recently established a partnership with the Weizmann Institute of Science in Israel, creating the Garvan-Weizmann Centre for Cellular Genomics, to focus studies on this new field.

Q: How could tumor microenvironments impact disease?

AS: The tumor microenvironment consists of all cells that comprise a tumor, not just malignant cells. There’s an increasing acceptance that the other cells within the tumor are not merely passengers, but are active participants in disease progression. Essentially, cancer operates like an ecosystem. If we understand all the components of that ecosystem, we expect to gain a far greater insight into its behavior. That includes its ability to form new metastatic lesions and its response to therapy.

Q: What other cells in the microenvironment could play a role in cancer onset and progression?

AS: The most obvious are the immune cells in the microenvironment. It’s clear that the immune system plays a role in sculpting cancer progression from the beginning. In our lifetimes, we all will develop abnormal cells through processes of mutation. In many instances, the immune system will recognize those cells and eliminate them. In rare instances, mutated cells develop strategies to evade the immune system and progress into cancer. There are cancer therapies in development and in clinical use to combat those strategies, for instance the 'immune checkpoint' drugs targeting PD1, PDL1, and CTLA4.

There are other cellular components of the microenvironment where the evidence is less clear, but we suspect play important roles. We are interested in stromal cells, including fibroblasts. These cells often make up a substantial proportion of a tumor and play a mechanical role. For example, fibroblasts determine how stiff or fluid a tumor might be. But fibroblasts also communicate with other cells in the tumor, influencing their behavior. They’re communicating with cancer and immune cells.

Q: How might an understanding of the role of fibroblasts in cancer facilitate the development of new therapeutics?

AS: Some beautiful work has been performed in experimental systems over the last two decades to show that fibroblasts influence the behavior of a tumor in response to treatment. In aggressive triple-negative breast cancers, we’ve shown that fibroblasts establish a niche that promotes a malignant, drug-resistant cancer behavior.¹ If we can find a way to target those fibroblasts with a small molecule inhibitor, we could stop them from providing that support and sensitize the cancer cells to respond better to therapy. We succeeded in doing it in experimental models and ran a Phase 1 trial to demonstrate the feasibility of such a strategy. That’s an example of the early evidence that supports targeting the fibroblasts or stromal cells to create a new weapon to treat cancer.

"Essentially, cancer operates like an ecosystem. If we understand all the components of that ecosystem, we expect to gain a far greater insight into its behavior."

Q: How did you initially investigate cancer microenvironments?

AS: We used methods such as immunohistochemistry, as well a flow cytometry to separate the cells out of the tumors so that we could sort them into different tubes. We then used RNA-Seq on the cell pools to understand their biology. That work brought us a long way, but it was laborious, and we were still left with heterogenous cell populations in the tubes.

Q: How does single-cell sequencing enable you to tease apart the different cell types within a tumor?

AS: The cell is the fundamental unit of life and disease. Just as epidemiological studies of millions of humans enable us to understand the behavior of a population, it’s also vital to study biological systems at the cellular level to gain new insights.

By using single-cell sequencing, we no longer start with an  a priori assumption about what we will find in a cancer microenvironment. We can simply observe and highlight what we did find. That has led to striking new insights that we wouldn’t have gained otherwise. For example, in our breast cancer studies we have found a population of stromal cells, called smooth muscle cells, that we didn’t realize existed in this state within breast cancer (unpublished). For the last 20 years, these cells were thought to be fibroblasts because they express some of the same markers. We never realized that they were a different cell type.

Because we are performing clinical studies, we’re working with a limited amount of tissue. Often, we have enough sample to perform only one assay. So, we’ve spent a lot of time learning as much as we can from a single sample. We’ve invested in multiomic methods that provide multiple dimensions of data from a single sample. For instance, we perform scRNA-Seq to obtain full transcriptomes and targeted cell surface proteomes for thousands of cells per sample on the NovaSeq 6000 System. We’ve developed a method called repertoire and gene expression by sequencing (RAGE-Seq) that enables us to perform full-length, single-molecule sequencing simultaneously for molecules where that information is critical, such as lymphocyte receptors.² We’ve also adopted methods developed by other labs, such as Cellular Indexing of Transcriptomes and Epitopes by Sequencing (CITE-Seq), that enable us to perform protein measurements at the same time as scRNA-Seq. Using CITE-Seq and the NextSeq 500 System, we can obtain data on 150 different proteins on the surface of thousands of single cells.

"We’ve invested in multiomic methods that provide multiple dimensions of data from a single sample. For instance, we perform scRNA-Seq to obtain full transcriptomes and targeted cell surface proteomes for thousands of cells per sample on the NovaSeq 6000 System."

Q: How did you identify different types of stromal cells and their role in intercellular signaling?

AS: We’ve used single-cell NGS techniques to identify the different stromal states that are present within a cancer. We identified four different stromal populations, including so-called myofibroblasts, inflammatory fibroblasts, and two forms of smooth muscle cells (unpublished). Based on their gene expression features, we can identify their cell type, predict the types of molecules they are expressing, and identify the molecules with which they signal to adjacent cells.

A hypothesis we are currently pursuing is that the nature of the stromal microenvironment in a breast cancer patient might, in part, dictate whether the immune system is competent to eliminate tumor cells. We had previously found evidence of fibroblasts signaling to cancer cells and making them more aggressive. We have also found evidence of fibroblasts signaling to CD8+ T-cells, the killer cells of the immune system, within the tumor microenvironment and secreting molecules to make those T-cells less effective in killing cancer cells (unpublished). This signaling could partly explain why the current generation of breast cancer immunotherapies are often ineffective.

We are now using experimental systems to test whether blocking this signaling can enhance antitumor immunity in breast cancer. I can imagine a scenario where patients might one day be given immunotherapy in addition to an agent that blocks the suppressive message from the stromal microenvironment.

Q: How do transcriptional properties reveal intratumoral heterogeneity?

AS: There is diversity within all the cell types within a tumor, including in the cancer cells. However, the drivers of that heterogeneity have remained elusive. We’ve started investigating that heterogeneity using scRNA-Seq.Endocrine therapies, such as tamoxifen and aromatase inhibitors, have been profoundly successful and saved hundreds of thousands of lives. However, many women still relapse following treatment with these drugs. We don’t understand the nature of the relapse in every case. Sometimes, it’s driven by mutations in the estrogen receptor and there are lots of researchers studying this area. In other cases, it could be driven by heterogeneity within the cellular population being treated. There might be populations of cells that are not endocrine dependent and that might explain the failure of endocrine therapies in treating these women effectively.

"In some ways, our Illumina systems have become invisible to us. We get our data, we know it’s good, and we move on. Illumina systems have performed very well."

Since it’s a transcription factor, the activity of the estrogen receptor can be readily measured with single-cell genomic techniques, such as scRNA-Seq, to assess diverse populations of cells and identify endocrine-resistant cell populations. The hope is that we will someday identify a vulnerability to eliminate these populations. This research could eventually drive the development of combination therapies for women who are at greater risk of relapsing.

Q: How have Illumina NGS systems performed in your research?

AS: We use Illumina NGS systems for all our high-throughput sequencing needs. We’re performing translational studies, working with precious clinical samples, so we need technology platforms where the reagents are reliable, consistent, and of high quality. In some ways, our Illumina systems have become invisible to us. We get our data, we know it’s good, and we move on. Illumina systems have performed very well.In the early days, we used the NextSeq 500 System. Our single-cell sequencing partners, the Ramaciotti Centre for Genomics at the University of New South Wales, acquired a NovaSeq 6000 System because of its high-capacity sequencing and we began using it in early 2019. We work with our colleagues to batch our samples to fill sequencing runs, enabling us to benefit from economies of scale. It has driven our sequencing costs down and our data are as good as ever.

Q: What are the next steps in your research?

AS: We will be performing more cellular genomics to fill in the gaps in our understanding of some common cancers. In prostate cancer, we have studies underway in early localized disease and metastatic disease. Prostate cancer has some wild heterogeneity. Pathologists who study prostate cancer often need to specialize in that disease because of its complex histopathology. There is often more than one cancer focus preexisting within a prostate, with those cancers surrounded by dynamic stromal remodeling. It’s an exciting research area where we’re going to learn a lot by shifting from pure morphological studies to much deeper genetic insights.We’re investing a lot of effort in finding the right way to filter NGS hits and drive them back into experimental systems to test their relevance. Using this approach, we could go from an observation of a phenomenon to something that’s mechanistically or therapeutically important.

Q: What upcoming trends do you see in the use of NGS sequencing in cancer research?

AS: One trend is spatial RNA profiling and related techniques. Currently, most single-cell genomics studies start with making a cell suspension. Therefore, we lose the spatial, morphological, and locational information in those cells.

New technologies are enabling us to perform cell resolution sequencing studies in tissue. In that way, we can maintain the morphology of the cells and the information about their spatial relationship to other cells that is critical to identifying and demonstrating signaling between cells. It’s exciting and I’m looking forward to what we learn from these new technologies.

Read more about sequencing at the Garvan Institute:

Garvan Institute Case Study

Learn more about:

Tumor Microenvironment Analysis

NovaSeq 6000 System

NextSeq 500 System: This system has been discontinued. We suggest the NextSeq 1000 and 2000 Systems as alternatives.

References

1. Cazet AS, Hui MN, Elsworth BL, et al. Targeting stromal remodeling and cancer stem cell plasticity overcomes chemoresistance in triple negative breast cancer. Nat Commun. 2018;9:2897. doi: 1038/s41467-05220-6.
2. Singh M, Al-Eryani G, Carswell S, et al. High-throughput targeted long-read single cell sequencing reveals the clonal and transcriptional landscape of lymphocytes. Nat Commun. 2019;10(1):3120. doi: 10.1038/s41467-019-11049-4.