by Eric P. Grewal
figures by Abby Burrus
The human body is made of thousands of types of cells, from neurons to blood cells and skin cells to kidney cells. While these cells differ vastly in shape and purpose, they all share one thing in common—their DNA, the set of “master instructions” that is carried in every cell in an individual. But if all cells possess the same set of instructions, how can they perform such different tasks? What really makes a muscle cell twitch and a neuron transmit signals to the brain?
What is transcription?
In many ways, DNA is a like a giant book with thousands of chapters, which we call genes. Although every cell has a copy of the book, different cells may read different chapters (i.e., genes) to carry out their unique functions. Cells “read” genes by copying them into molecules called messenger RNA. As the name suggests, these messages are what inspire a cell to act by providing the “recipe” for a cell to build proteins, grow, and carry out its function.
The entire process of copying a DNA-encoded gene into messenger RNA is a process called transcription, and scientists have been interested in understanding transcription for decades. Since the messenger RNA contains all of the instructions a cell is reading, it’s a great source to tap into to understand the activity of a cell. In recent years, researchers have been able to analyze transcription in many cell types as well as how it can go wrong in diseases such as cancer and viral infection. However, an emerging technology is about to revolutionize our understanding of transcription.
Old Methods: Bulk RNA-sequencing
To study transcription, you traditionally begin by choosing a cell type to investigate—perhaps retinal cells if you want to learn about vision, or pancreatic cells if you’re curious about diabetes. Cells of this type must then be physically isolated from the sample, such as a whole eye or an organ extract from a diseased patient. Luckily, different cell types have different markers on their surface, so we can use microscopic labels that uniquely tag these surface markers to sort out the cell type we’re interested in.
Once the right cells have been collected and pooled together in a test tube, it’s a straightforward process to “break open” the cells to extract the RNA. The information in the RNA then undergoes sequencing, a process that can convert RNA messages into gene names that a scientist can easily understand. From this information, researchers learn exactly what the cells are doing and what genes are being transcribed (Figure 1).
While this method of “bulk” RNA-sequencing has been widely adopted, there are multiple limitations to this technology. The main problem is that bulk analysis requires you to sort out a single population of cells, using surface markers as mentioned above. But what if you want to study multiple different cells types, or even a new cell type whose surface markers are unknown? Another limitation arises from the bulk nature of the process. Since all of the RNA is pooled together, any information about the individual cells, rather than the group as a whole, is lost. This can be a limitation since cells of the same type may still have different transcription. For example, if a virus infects one lung cell while another is uninfected, you can bet there will be a difference in transcription between the otherwise identical cells, but it won’t be detected since all of the cells are pooled together in a group.
Single-cell RNA-sequencing: A ground-breaking technology
Today, a new paradigm is changing the way we analyze transcription: single-cell analysis. Pioneered in 2009, single-cell RNA-sequencing (scRNA-seq) overcomes many of the limitations of bulk analysis by looking at cells one at a time to see which genes each individual cell is transcribing. With this approach, we are able to detect thousands of genes per cell, unlocking the possibility of seeing differences in gene transcription in small groups of cells or even individual cells that would have otherwise been masked by the information from the rest of the cell population. With single-cell analysis, the sorting process becomes optional, so any mix of cells can be put into the group and the data that comes out still tells you about each one individually.
Despite the vast amount of data generated by scRNA-seq, the procedure is relatively simple. In fact, one of the most widely adopted methods for single-cell analysis was developed separately by researchers in the McCarroll and Kirschner Labs at Harvard Medical School. Scientists begin by choosing a sample for analysis, such as blood from a patient with cancer. They then prepare the cells for sequencing by running them through a small, chip-like device, which has microscopic tubes that only let one cell through at a time. Once the cells are separated into individual droplets in the machine, chemicals that aide in the sequencing process can be added to every droplet (Figure 1). Notably, the chemicals in each droplet are tagged with a unique code that can be transferred onto the RNA. Once the sequencing is done, each cell’s unique RNA is uniquely identifiable by that tag and the genes from each cell are listed together.
scRNA-seq in Medicine
Although scRNA-seq has existed for a number of years, the revolution for human health is just beginning, as the technology has finally become standardized and affordable enough for labs across the globe to begin performing single-cell analysis. One recent study applied scRNA-seq in a study of chronic myelogenous leukemia (CML), a cancer in which white blood cells grow uncontrollably in the blood and bone marrow. CML is often caused by a mutation in a bone marrow stem cell, which can be thought of as the “mother” for all other blood cells. Since these stem cells will divide and seed all of the other cells in the bone marrow, any mutations from the stem cell will also be carried into the “daughter” cells it divides into, allowing the cancer to spread down the line. Understanding which genes these cancerous cells are transcribing could hold the key to better treatment strategies.
Using scRNA-seq, researchers analyzed thousands of CML stem cells from human patient volunteers. They found that even in the same patient, different cancer cells could transcribe different genes, meaning that the cancer had a lot of variation within the body. This finding could be crucial for physicians who design therapies for CML. Often, drugs are made to target specific genes that are only transcribed in one kind of cancer cell. However, if there are many kinds of cancer cells within a patient, some of which have the gene while others don’t, you can imagine how the therapy may only wipe out some of the cancer cells and leave the patient with residual disease. Previously, these different kinds of cancer cells were undetectable using bulk RNA-sequencing, since all the cancer cells were pooled together and the unique gene signals were lost. Thanks to scRNA-seq, doctors may soon be able to detect all the different kinds of cancer cells within a patient and prescribe combination therapies that ensure they are all killed. In this way, single-cell transcriptomics has the potential to transform modern medicine.
What’s Next for scRNA-seq?
Although scRNA-seq is an exciting technology, it doesn’t come without a few downsides. One of the biggest limitations of scRNA-seq is its cost, which must account for the samples, single-cell device, chemicals, sequencing, and labor. With experiments that incorporate a few thousand cells, and costs that average about $1 per cell, the price of this technique adds up quickly. Furthermore, transcription isn’t everything; although RNA-sequencing tells us which genes a cell may be reading, it doesn’t prove that those genes are causing a cell to promote disease or another process. We still need studies that test the function of genes that are identified as interesting in scRNA-seq, such as introducing those genes into mouse models in the lab or studying how they affect drug response in human clinical trials.
Nevertheless, the dawn of the single cell era is a welcome advance for the field of biology. As sequencing costs fall and data analysis techniques improve, perhaps we’ll all benefit from the simple approach of zooming in. Novel methods mean novel conclusions, and single cell RNA-sequencing is a significant step towards understanding and treating disease
Eric P. Grewal is a graduate student in the Immunology Program at Harvard Medical School and a researcher at Dana-Farber Cancer Institute.