Despite the enormous effort by the scientific community to understand the toll SARS-CoV-2 infection takes on the lungs and other organs, little is known about the pathophysiology of the disease. Now, a collaborative team has developed a foundational dataset consisting of tissue atlases that elucidate the biological impact of severe SARS-CoV-2 infection in the body. In this work, scientists use spatial analysis to reveal how infected cells from multiple organs exhibit a range of molecular and genomic changes. They also saw signs of multiple, unsuccessful attempts by the lungs to repair themselves in response to respiratory failure—the leading cause of death in COVID-19 patients. The ability to pinpoint cellular processes, expression pathways, and immune cell profiles in this way offers key insights into systemic COVID-19 disease progression.

This work is published in Nature in the paper, “COVID-19 tissue atlases reveal SARS-CoV-2 pathology and cellular targets.

“You really feel the tragedy of the disease when you see that result,” said Aviv Regev, PhD, co-senior author of the study and executive vice president, Genentech Research and Early Development. “The lung tries everything at its disposal, and it still can’t fix itself. This was a very emotional study. We are grateful to the patients and families who agreed to donate tissue for COVID-19 research to help advance understanding of this devastating disease.”

The researchers studied tissue obtained at autopsies of 17 individuals who succumbed to COVID-19 and were cared for at Beth Israel Deaconess Medical Center, Brigham and Women’s Hospital, and Massachusetts General Hospital.

The team investigated how the SARS-CoV-2 virus interferes with the function of cells and their genetic programs. They used single-cell RNA sequencing data from tissue samples taken from 11 organ systems—including the lungs, heart, liver, and kidneys—to build a comprehensive “cell atlas” of hundreds of thousands of individual cells showing how COVID-19 can lead to organ failure and death.

The researchers used NanoString Technologies’ GeoMx DSP to characterize the infected autopsy lung tissue samples. In conjunction with single-cell RNA sequencing data collected for the project, researchers utilized the Cancer Transcriptome Atlas (CTA) and the Whole Transcriptome Atlas (WTA), with additional custom content specific for SARS-CoV-2 genes, to produce and spatially map a comprehensive cell-type atlas.

The team profiled RNA from the individual cells and developed new methods to analyze and annotate the large amounts of sequence data. They compared gene expression signatures from different cells: COVID-19-damaged cells and uninfected cells from the COVID-19 patients, as well as cells from patients with other diseases and from healthy individuals.

RNAScope (left) was utilized to assess the SARS-CoV-2 viral load of autopsy samples, and inform a strategy for region of interest selection (middle). Regions showing noteworthy biology were selected and segmented on CD45 and CD68 to focus on specific structures within the tissues (right). [Nanostring Technologies]
The team’s cell atlas is freely and openly available. They also created a 420-specimen biobank from the autopsy samples that can be used for other COVID-19 studies. “We created a foundational resource for other researchers to use in the future to ask specific questions,” said Orit Rozenblatt-Rosen, PhD, co-senior author and an institute scientist and the scientific director of the Klarman Cell Observatory at the Broad when the study began.

The most extensive suite of findings were from the lungs. The scientists were astounded by the extent of the changes in genetic programs they found there. “The virus wreaks havoc in the lungs and we see it in the cells,” Regev said.

One main cause of lung damage in COVID-19 is the destruction of AT1 cells, which enable breathing and gas transfer. The scientists found that as AT1 cells died, related lung cells called AT2 attempted to convert themselves into AT1 cells through a process called transdifferentiation. But this attempt halted mid-way through, leaving the cells in an intermediary state that is often seen in patients with other lung diseases such as pulmonary fibrosis.

In a last-ditch attempt at self-repair, the lungs tried to turn cells from higher up in the airways, known as intrapulmonary basal-like progenitor cells, into AT1 cells. This attempt at transdifferentiation had only previously been seen in mouse models.

The findings suggest that the lung failure in patients was caused by the inability of lung cells to outpace the damage caused by the virus as the cells tried to regenerate.

The paper also describes how the virus impacts other tissues outside of the lungs. One surprising finding was that while the heart sustained significant damage and showed evidence of altered genetic programs in many different cell types, there was very little viral RNA in the heart tissue itself. “Whether that means the virus had already been cleared, or that the heart was collateral damage is an area for further research,” said Regev.

The researchers also looked at 27 different genes that previous genome-wide association studies have linked to severe COVID-19. They zeroed in on a handful that were highly expressed in key cell types in the new study, particularly those in infected lungs. This finding helps narrow down the list of potential genetic factors for severe disease and highlights the cell types that may be most relevant in severe COVID-19.

The team now plans to finish analyzing the other autopsied tissues, such as brain, spleen, and trachea, to paint a more complete picture of COVID-19 pathology and provide a resource for future studies.