Integrated bulk and single-cell RNA-seq analysis reveals psychological stress-related prognostic genes for risk stratification in lung adenocarcinoma

Lung adenocarcinoma is the most common form of lung cancer, a disease that remains one of the leading causes of cancer death worldwide. While modern medicine has made strides with targeted therapies and immunotherapies, the outcome for patients varies wildly. Some people respond well to treatment and live for years, while others face a much shorter journey, often because the cancer behaves differently in every individual. For decades, doctors have looked at the physical tumor and the patient's overall health to guess how the disease will progress. However, a growing body of research suggests that the mind and the body are deeply connected, and that a patient's emotional state might actually influence how their cancer grows. Psychological stress, the body's reaction to difficult life events, triggers a cascade of chemical signals that can alter how cells behave. Scientists have long suspected that this internal pressure could make tumors more aggressive or harder to treat, but pinning down exactly which genes are responsible for this link has been difficult.

A team of researchers set out to bridge this gap by looking at the genetic blueprint of lung cancer tumors through a new lens. They wanted to find specific genes that not only drive the cancer but are also tied to the body's response to stress. To do this, they combined two powerful types of data. First, they examined bulk RNA sequencing, which acts like a smoothie of genetic information, blending the signals from millions of cells in a tumor sample to see the average activity. Second, they used single-cell RNA sequencing, a much sharper tool that reads the genetic code of individual cells one by one, allowing them to see exactly which cell types are active and how they talk to one another. By cross-referencing these genetic maps with a curated list of genes known to react to psychological stress, they aimed to build a model that could predict a patient's survival based on the unique genetic signature of their tumor.

The researchers began by comparing the genes in lung cancer tumors against those in healthy lung tissue. They identified thousands of genes that were turned on or off in the cancer. From this massive list, they filtered for the specific genes that are known to react to stress. This process narrowed their focus to a small group of candidates that seemed to sit at the intersection of cancer growth and stress response. Using advanced statistical methods, they tested which of these genes were most strongly linked to how long patients survived. After rigorous screening, they settled on a signature of just four genes: CCNB1, TLR4, CCNA2, and EXO1. These four genes acted as a powerful predictor. When the researchers calculated a risk score based on the activity levels of these genes, they could clearly divide patients into two groups: a high-risk group and a low-risk group.

The difference between these two groups was stark. Patients in the high-risk group, whose tumors showed high activity of the stress-related genes, had significantly shorter survival times compared to those in the low-risk group. This pattern held true not only in the initial group of patients they studied but also when they tested their model on a separate group of patients from a different database, proving that the finding was reliable and not just a fluke of one specific dataset. The researchers also built a tool called a nomogram, which combines this genetic risk score with standard clinical factors like the size of the tumor and whether it has spread to lymph nodes. This tool allows doctors to estimate a patient's chance of survival over one, two, or three years with greater accuracy than using clinical factors alone.

Digging deeper, the team explored what these genes were actually doing inside the body. They found that the high-risk group had a distinct biological landscape. Their tumors were more active in the cell cycle, the process by which cells divide and multiply, and they showed signs of greater genomic instability, meaning the DNA inside the cancer cells was more prone to errors and mutations. The high-risk tumors also had a different mix of immune cells surrounding them. While the low-risk group had a more balanced immune environment, the high-risk group showed signs of an inflammatory state that, paradoxically, seemed to support the cancer's growth rather than fight it. Furthermore, the study suggested that patients in the high-risk group might respond differently to chemotherapy. The model predicted that these patients would be more sensitive to certain drugs like cisplatin and docetaxel, while the low-risk group might respond better to other agents. This suggests that knowing a patient's stress-related genetic profile could help doctors choose the right chemotherapy from the start.

To understand which cells were driving this behavior, the researchers turned to their single-cell data. They discovered that the four key genes were most active in "cycling cells," a broad category of cells that are currently in the process of dividing. These cycling cells were not just a single type; the researchers found they could be broken down into subgroups, including cycling epithelial cells, cycling immune cells, and cycling macrophages. In the high-risk tumors, these cycling cells were communicating with each other more intensely than in normal tissue or low-risk tumors. They were sending stronger signals to one another, creating a network that seemed to fuel the tumor's progression. The study also traced the life cycle of these cells, showing how their genetic activity changed as they moved from early stages of development to later stages, with the stress-related genes shifting their expression patterns at critical moments.

The researchers did not stop at computer models. They took samples of lung cancer tissue and cell lines from a hospital in Xinjiang and used a technique called RT-qPCR to measure the actual levels of these genes in the lab. They confirmed that three of the four genes—CCNB1, CCNA2, and EXO1—were indeed present at higher levels in the cancer tissues compared to healthy tissue, matching the predictions from their large-scale data analysis. One gene, TLR4, did not show a significant difference in this specific lab test, which the authors noted could be due to differences in the patient populations or the specific methods used. This experimental step was crucial because it moved the findings from theoretical data to physical reality, confirming that these genes are indeed overactive in human lung cancer.

The study concludes that psychological stress is not just a feeling that happens to a patient; it leaves a molecular fingerprint on the tumor itself. The four genes identified in this work serve as a biological bridge, connecting the external pressure of stress to the internal machinery of cancer growth. While the researchers acknowledge that their work is based largely on analyzing existing data and that further experiments are needed to fully understand the mechanisms, the results offer a new way to think about lung cancer risk. By identifying patients whose tumors are driven by these stress-related pathways, doctors may soon be able to stratify patients more accurately, predicting who is at higher risk of poor outcomes and tailoring treatments to target the specific vulnerabilities of their cancer. This approach moves beyond simply treating the tumor and begins to treat the unique biological context in which that tumor exists.