Discrete single-cell microRNA analysis for phenotyping the heterogeneity of acute myeloid leukemia

Abstract

<p>Acute myeloid leukemia (AML) is a highly heterogenous cancer in <a href="https://www.sciencedirect.com/topics/biochemistry-genetics-and-molecular-biology/hematopoiesis" title="Learn more about hematopoiesis from ScienceDirect's AI-generated Topic Pages">hematopoiesis</a>, and its subtype specification is greatly important in the clinical practice for AML diagnosis and prognosis. Increasing evidence has shown the association between <a href="https://www.sciencedirect.com/topics/biochemistry-genetics-and-molecular-biology/microrna" title="Learn more about microRNA from ScienceDirect's AI-generated Topic Pages">microRNA</a> (miRNA) phenotype and AML therapeutic outcomes, emphasizing the need for novel techniques for convenient, sensitive, and efficient miRNA profiling in clinical practices. Here, we describe a nanoneedle-based discrete single-cell <a href="https://www.sciencedirect.com/topics/biochemistry-genetics-and-molecular-biology/microrna" title="Learn more about microRNA from ScienceDirect's AI-generated Topic Pages">microRNA</a> profiling technique for multiplexed phenotyping of AML heterogeneity without the requirement of sequencing or <a href="https://www.sciencedirect.com/topics/materials-science/polymerase-chain-reaction" title="Learn more about polymerase chain reaction from ScienceDirect's AI-generated Topic Pages">polymerase chain reaction</a> (PCR). In virtue of a biochip-based and non-destructive nature of the assay, the expression of nine miRNAs in large number of living AML cells can be simultaneously analyzed with discrete single-cell level information, thus providing a proof-of-concept demonstration of an AML subtype classifier based on the multidimensional miRNA data. We showed successful analysis of subtype-specific cellular composition with over 90% accuracy and identified drug-responsive leukemia subpopulations among a mixed suspension of cells modeling different AML subtypes. The adoption of <a href="https://www.sciencedirect.com/topics/engineering/machine-learning-algorithm" title="Learn more about machine learning algorithms from ScienceDirect's AI-generated Topic Pages">machine learning algorithms</a> for processing the large-scale nanoneedle-based miRNA data shows the potential for powerful prediction capability in clinical applications to assist therapeutic decisions. We believe that this platform provides an efficient and cost-effective solution to move forward the translational prognostic usage of miRNAs in AML treatment and can be readily and advantageously applied in analyzing rare patient-derived clinical samples.</p>