110 Researchers led by Xian-Yang Qin at the RIKEN Centre for Integrative Medical Sciences (IMS) in Japan have developed a score that predicts the risk of liver cancer. Published in the scientific journal Proceedings of the National Academy of Sciences, the study establishes that the protein MYCN drives liver tumourigenesis, specifically in the tumours found in the deadliest subtype of liver cancer. The study characterises the microenvironment of genes that permit overexpression of MYCN and describes a machine-learning algorithm that uses these data to predict how likely a tumour-free liver is to develop tumours. Liver cancer, or hepatocellular carcinoma, causes more than 800,000 deaths worldwide every year. The mortality rate is very high because the disease often remains undetected until the late stages and because the recurrence rate is between 70–80%. In hopes of discovering a much-needed method that accurately predicts at-risk livers before tumours develop, Qin and his team have been studying a protein called MYCN. The MYCN gene is recognised as a contributor to liver cancer that develops from damaged livers, but exactly how this occurs has remained unclear. The researchers reasoned that if its overexpression directly leads to liver tumourigenesis, it would be an ideal candidate as a biomarker and for further study. To test their theory, the team used a hydrodynamic tail-vein injection-based transposon system to insert MYCN into the mouse liver genome, creating a mouse liver that overexpressed MYCN. The team found that when they used the system to overexpress MYCN together with always-active AKT, 72% of the mice developed liver tumours within 50 days. A variety of tests showed that these tumours had all the characteristics of human hepatocellular carcinoma. Tumours did not develop when either gene was overexpressed on its own. Understanding how early microenvironmental cues trigger liver tumourigenesis is critical for developing ways to counter it. To characterise the microenvironment, the researchers turned to spatial transcriptomics. This technique shows which genes are activated in a tissue and precisely where that activity is occurring. In a mouse model of metabolic dysfunction-associated liver cancer, the researchers used this method to examine gene expression over time and by location as liver tumours developed, focusing on areas where MYCN levels were increasing. They discovered a cluster of 167 genes that were differentially expressed in tumour-free sections of liver with increased MYCN levels. They named this cluster the “MYCN niche.” Based on the mouse spatial transcriptomics data, the researchers then developed a machine-learning model that can analyse the characteristics of a given gene-expression pattern and generate a score indicating whether it corresponds to a MYCN niche. The model achieves 93% accuracy. The MYCN niche score was subsequently calculated for human hepatocellular carcinoma datasets. Patients with higher MYCN niche scores showed a greater risk of tumour recurrence and poorer clinical outcomes. This relationship was stronger when the score was derived from non-tumour tissue rather than tumour tissue. The score therefore represents a proof-of-concept spatial biomarker that predicts prognosis based on microenvironments that promote tumour formation. “We have developed a clinically actionable strategy to identify high-risk patients by profiling gene expression in non-tumour liver tissue,” explains Qin. “By integrating spatial transcriptomics with machine learning, we have established a MYCN niche score that predicts recurrence risk and detects precancerous microenvironments predisposed to de novo liver tumourigenesis.” “In the future, we aim to further dissect the biological mechanisms captured by machine learning-derived spatial feature scores and determine how cancer-permissive environments are established and maintained.”
Researchers led by Xian-Yang Qin at the RIKEN Centre for Integrative Medical Sciences (IMS) in Japan have developed a score that predicts the risk of liver cancer. Published in the scientific journal Proceedings of the National Academy of Sciences, the study establishes that the protein MYCN drives liver tumourigenesis, specifically in the tumours found in the deadliest subtype of liver cancer. The study characterises the microenvironment of genes that permit overexpression of MYCN and describes a machine-learning algorithm that uses these data to predict how likely a tumour-free liver is to develop tumours. Liver cancer, or hepatocellular carcinoma, causes more than 800,000 deaths worldwide every year. The mortality rate is very high because the disease often remains undetected until the late stages and because the recurrence rate is between 70–80%. In hopes of discovering a much-needed method that accurately predicts at-risk livers before tumours develop, Qin and his team have been studying a protein called MYCN. The MYCN gene is recognised as a contributor to liver cancer that develops from damaged livers, but exactly how this occurs has remained unclear. The researchers reasoned that if its overexpression directly leads to liver tumourigenesis, it would be an ideal candidate as a biomarker and for further study. To test their theory, the team used a hydrodynamic tail-vein injection-based transposon system to insert MYCN into the mouse liver genome, creating a mouse liver that overexpressed MYCN. The team found that when they used the system to overexpress MYCN together with always-active AKT, 72% of the mice developed liver tumours within 50 days. A variety of tests showed that these tumours had all the characteristics of human hepatocellular carcinoma. Tumours did not develop when either gene was overexpressed on its own. Understanding how early microenvironmental cues trigger liver tumourigenesis is critical for developing ways to counter it. To characterise the microenvironment, the researchers turned to spatial transcriptomics. This technique shows which genes are activated in a tissue and precisely where that activity is occurring. In a mouse model of metabolic dysfunction-associated liver cancer, the researchers used this method to examine gene expression over time and by location as liver tumours developed, focusing on areas where MYCN levels were increasing. They discovered a cluster of 167 genes that were differentially expressed in tumour-free sections of liver with increased MYCN levels. They named this cluster the “MYCN niche.” Based on the mouse spatial transcriptomics data, the researchers then developed a machine-learning model that can analyse the characteristics of a given gene-expression pattern and generate a score indicating whether it corresponds to a MYCN niche. The model achieves 93% accuracy. The MYCN niche score was subsequently calculated for human hepatocellular carcinoma datasets. Patients with higher MYCN niche scores showed a greater risk of tumour recurrence and poorer clinical outcomes. This relationship was stronger when the score was derived from non-tumour tissue rather than tumour tissue. The score therefore represents a proof-of-concept spatial biomarker that predicts prognosis based on microenvironments that promote tumour formation. “We have developed a clinically actionable strategy to identify high-risk patients by profiling gene expression in non-tumour liver tissue,” explains Qin. “By integrating spatial transcriptomics with machine learning, we have established a MYCN niche score that predicts recurrence risk and detects precancerous microenvironments predisposed to de novo liver tumourigenesis.” “In the future, we aim to further dissect the biological mechanisms captured by machine learning-derived spatial feature scores and determine how cancer-permissive environments are established and maintained.”