Defining transcriptomic niches in human fibrotic lung with multi-sample spatial transcriptomics analysis using the MAPLE algorithm

The human lung is a vast, intricate landscape where air meets blood, a delicate exchange that keeps us alive. In a disease called idiopathic pulmonary fibrosis, this landscape is slowly transformed. Healthy, spongy tissue is replaced by thick, stiff scar tissue, much like a lush garden being overtaken by concrete. This process is not uniform; it happens in patches, leaving some areas of the lung relatively untouched while others are ravaged. For decades, scientists have understood that this disease involves a chaotic mix of different cell types, but they have struggled to see how these cells are arranged in space. Traditional methods of studying cells often involve grinding up tissue samples, which destroys the map of where each cell was sitting. Without that map, it is impossible to understand how neighboring cells talk to one another or how the disease spreads from one patch to another.

To solve this, a team of researchers developed a new way to look at the lung that preserves its geography. They used a technique called spatial transcriptomics, which allows scientists to read the genetic instructions inside cells while keeping them exactly where they were in the tissue. However, analyzing this kind of data is difficult because it involves many different samples from different patients, each with unique variations. The researchers created a new computer method, which they named MAPLE, to stitch these separate maps together into a single, coherent picture. By applying this method to lung tissue from healthy donors and patients with severe fibrosis, they uncovered a hidden order in the chaos. They found that the disease does not just happen everywhere at once; instead, it follows specific regional rules. In the upper parts of the lung, the damage is driven by a heavy buildup of structural proteins, while in the lower parts, the disease is fueled by a distinct type of immune cell that gathers in specific neighborhoods.

The researchers began by collecting lung tissue from four individuals: two healthy people and two patients with advanced fibrosis. They took samples from both the upper and lower sections of the lungs, preserving the tissue in a way that kept the cells intact. Using a high-tech imaging system, they captured the genetic activity of thousands of tiny spots across these tissue sections. Each spot contained a mix of cells, and the goal was to figure out what kind of cells were there and how they were organized. The new MAPLE algorithm acted as a sophisticated guide, grouping these spots into distinct regions based on their genetic signatures and their location. It did this by learning from all the samples at once, recognizing patterns that were common across patients while also accounting for the differences between the upper and lower lobes.

Once the regions were mapped out, the team could see a clear story of disease progression. In the healthy lungs, the tissue was dominated by cells that line the air sacs and the blood vessels, the essential workers of gas exchange. As the disease moved from healthy tissue to the upper lobes of the fibrotic lungs, and finally to the lower lobes, these essential cells disappeared. In their place, a different type of cell took over: the mesenchymal cells. These are the cells responsible for building the structural framework of the body, but in fibrosis, they become overactive, laying down too much scar tissue. The researchers observed a steady shift where the healthy, air-exchanging cells were replaced by these scar-building cells, a transition that became most extreme in the lower parts of the diseased lungs.

The study also revealed that the upper and lower parts of the lung are not just damaged in the same way; they are damaged by different mechanisms. In the upper lobes, the fibrotic areas were characterized by a massive remodeling of the extracellular matrix, the scaffolding that holds cells together. This area was rich in genes that build collagen and other structural proteins. In contrast, the lower lobes told a different story. Here, the fibrotic regions were surrounded by a dense gathering of B-cells, a specific type of immune cell. The researchers confirmed this finding using a second, higher-resolution imaging technique that could see individual cells. They saw that these B-cells were not just floating randomly; they were clustered tightly around the fibrotic scars, suggesting a close and active relationship between the immune system and the scarring process in that specific part of the lung.

Another critical discovery involved the presence of senescent cells, often called "zombie" cells. These are cells that have stopped dividing but refuse to die, instead releasing signals that can harm their neighbors. The researchers found that these senescent cells were not scattered evenly throughout the lung. Instead, they were concentrated in specific spots that were right next to a strange, abnormal type of cell called an aberrant basaloid cell. These basaloid cells are a sign of severe lung injury and repair gone wrong. The data showed a strong spatial link: where these abnormal basaloid cells were found, the senescent cells were also present in high numbers. This suggests that the accumulation of these harmful, non-dividing cells is closely tied to the expansion of these abnormal repair cells, creating a local environment that may drive the disease forward.

The work provides a new framework for understanding how complex diseases like fibrosis organize themselves in the body. By combining data from multiple patients and accounting for the specific geography of the lung, the researchers were able to see patterns that would have been missed if each sample had been studied alone. They demonstrated that the disease is not a single, uniform process but a collection of regional programs, each with its own mix of immune cells, structural changes, and cellular behaviors. The new method they developed, MAPLE, offers a way to explore these spatial relationships in any tissue, helping scientists move beyond simple lists of genes to a true understanding of how cells interact in their natural environment. This approach does not just describe what is wrong with the lung; it shows exactly where and how the breakdown is happening, offering a clearer path for future research into how to stop or reverse the scarring process.