Skip to main content
Premium Trial:

Request an Annual Quote

Normal Breast Cell Types, States Characterized by Single-Cell Transcriptome Analysis

NEW YORK – A team led by investigators at the MD Anderson Cancer Center has profiled the expression features and spatial organization of individual cells from normal breast tissue, providing a baseline look at cell types and expression states that is expected to inform future cancer studies.

"It's essential to chart out what a normal cell reference would look like," explained Tapsi Kumar, a graduate student working in genetics with Nicholas Navin and Andrew Futreal at MD Anderson, during an online presentation done for the American Association for Cancer Research Virtual Annual Meeting on Tuesday.

Along with collaborators at MD Anderson and UC Irvine, the researchers focused on 35 breast samples from 19 women undergoing mastectomy or reduction mammoplasty surgery. Following a rapid dissociation and analysis protocol, Kumar explained, they did microdroplet-based 10x Genomics single-cell RNA sequencing on around 160,000 individual breast stroma cells, along with single-nucleus sequencing on more than 89,300 cells.

"Single-cell genomics allows us to study the genes that are active in each and every cell, giving us an idea about what cells are present in an organ, in what state they are in, and what is their function," Kumar said, noting that breast tissue can change with an individual's age, during cycles of menstruation or pregnancy, and with the presence of diseases such as cancer.

In the past, she noted, histology-based analyses highlighted 10 main cell types in breast tissue, though those classifications do not encompass the full suite of cells and cell states in the breast.

With the single-cell transcriptomes, the team saw expression clusters corresponding to around a dozen major and minor non-adipose breast cell types. These included fibroblast and epithelial cells — the most abundant cell types in the pathologically normal breast tissues being considered — as well as multiple immune cell types such as T cells, B cells, macrophages, and natural killer cells.

The individual cell transcriptomes also pointed to the presence of previously unappreciated cell type markers and expression-based cell states not described in the past, Kumar noted. For example, the available single-cell expression data revealed gene signatures corresponding to several populations of fibroblasts — a type of collagen-expressing, connective cell that can comingle with burgeoning tumors when breast cancer occurs.

Such cell states could stem from cells from different stages of differentiation, she explained, or from cells within the same cell type that have distinct behaviors or roles in the breast depending on their broader location and interactions.

For their adipocyte-focused analysis, the researchers focused on samples from four women who had reduction mammoplasty and half a dozen participants undergoing mastectomy surgery, identifying at least two distinct adipocyte states and additional mast cells that had not been picked up in the original single-cell RNA-seq experiments.

Owing to technical problems caused by the large adipocyte fat cells, those cells were profiled separately by single nuclei profiling, Kumar explained, which produces lower overall coverage of the genes considered.

"When we get the tissues and we dissociate them, adipocytes are like a layer of fat that we need to remove eventually," she noted during her AACR presentation. "Since these are big cells, they clog our machines."

The team was also able to look at subtle expression differences across the vascular, immune, and other cell types depending on their location within the breast using cell type-specific markers, barcode-based spatial transcriptomic mapping, and other approaches.

Though the current analysis focused on pathologically normal samples processed within a couple hours of surgery, Kumar noted, the cellular dynamics identified may inform future studies of breast tissue across a range of normal to cancerous conditions.

"We're trying to learn how these cells are and how they behave in their normal ecosystem when there is no presence of any tumor cells," she said. "This helps to study how these cells are transformed in the presence of a tumor cell, and how the entire microenvironment helps in the progression of the tumor."

The work was presented as part of a session focused on cancer genomics through the lens of evolution, spanning normal tissue features to tumors and metastasis, and is intended to guide future studies of normal breast biology and the processes that become altered when breast cancer develops.

It complements single-cell sequencing strategies that Navin and his colleagues have used in the past to look at everything from breast cancer evolution, mutation rates, or subtype-related copy number profiles to potential treatment resistance mechanisms.

"The breast cell atlas data provides an invaluable normal reference for the research community to understand how normal cell types are reprogrammed in diseases such as breast cancer," Kumar, Navin, and their co-authors wrote in an abstract accompanying the online conference presentation.