Label-free identification of non-activated lymphocytes using three-dimensional refractive index tomography and machine learning

Abstract

Identification of lymphocyte cell types is crucial for understanding their pathophysiologic roles in human diseases.Current methods for discriminating lymphocyte cell types primarily relies on labelling techniques with magnetic beads or fluorescence agents, which take time and have costs for sample preparation and may also have a potential risk of altering cellular functions. Here, we present label-free identification of non-activated lymphocyte subtypes using refractive index tomography. From the measurements of three-dimensional refractive index maps of individual lymphocytes, the morphological and biochemical properties of the lymphocytes are quantitatively retrieved. Machine learning methods establish an optimized classification model using the retrieved quantitative characteristics of the lymphocytes to identify lymphocyte subtypes at the individual cell level. We show that our approach enables label-free identification of three lymphocyte cell types (B, CD4+ T, and CD8+ T lymphocytes) with high specificity and sensitivity. The present method will be a versatile tool for investigating the pathophysiological roles of lymphocytes in various diseases including cancers, autoimmune diseases, and virus infections.Lymphocytes consisting of various cell types including B, helper (CD4+) T, cytotoxic (CD8+) T, and regulatory T lymphocytes, and play crucial roles in the adaptive immune system 1 . Each lymphocyte cell type has different functions: B lymphocytes produce antibodies, and T lymphocytes recognize a specific antigen and execute effector functions. The lymphocyte population and function are tightly regulated to defend the host against harmful invaders or abnormal conditions 1,2 . Disturbances in lymphocyte function and regulation are related to various diseases including cancers 3-5 , autoimmune diseases 6,7 , and virus infections 8,9 .. CC-BY 4.0 International license not peer-reviewed) is the author/funder. It is made available under a The copyright holder for this preprint (which was . http://dx.doi.org/10.1101/107805 doi: bioRxiv preprint first posted online Feb. 11, 2017; To understand the roles of different types of lymphocytes, several methods based on labelling techniques have been developed to identify and discriminate lymphocyte cell types. Because different kinds of lymphocytes have very similar cellular morphology such as a large nucleus with small cytosolic regions and round shapes, conventional optical methods such as bright-field microscopy or phase contrast microscopy have limited ability in classifying lymphocyte cell types 10 .To overcome this, a specific surface membrane proteins, known as surface markers, are recognized and tagged with magnetic beads or fluorescence molecules via antigen-antibody binding. And then, specific types of lymphocytes are identified and separated by magnetic forces or fluorescence signals 11 . Targeting surface markers is a precise and efficient manner to determine the cell types; however, labelling methods have potential risks of altering cellular functions b...

Measurements of morphological and biophysical alterations in individual neuron cells associated with early neurotoxic effects in Parkinson's disease

“…We exploit the fact that RI is an intrinsic optical property of materials and linearly proportional to material concentration (26). Recently, measuring 3D RI distributions has been widely applied to study the pathophysiology of various biological samples, such as red blood cells (28)(29)(30)(31)(32), white blood cells (33)(34)(35), cancer cells (36)(37)(38)(39)(40)(41)(42), phytoplankton (43), and bacteria (44)(45)(46). Thus, dry mass of non-aqueous molecules making up the cell can be measured from RI information without any invasive labeling process.…”

Holotomography: refractive index as an intrinsic imaging contrast for 3-D label-free live cell imaging

“…For example, from the 2-D quantitative phase images of individual bacteria, the genus of various bacteria was distinguished using a machine learning algorithm (Jo et al 2015;Jo et al 2014). In addition, non-activated lymphocytes were recently identified and classified from their 3-D RI tomograms (Yoon et al 2017). Recently, the weapons-grade anthrax spores are optically detected using 2-D QPI techniques and deep learning .…”

“…In addition, it can save time and cost for sample preparation. This label-free feature might become powerful for some applications where cells are to be reinjected to human bodies, for example, as in immune therapy (Yoon et al 2017;Yoon et al 2015) or stem cell therapy (Braydich-Stolle et al 2005).…”

“…Certain molecules such as lipid (Jung et al 2018;Kim et al 2016b) or metal nanoparticles (Kim et al 2018a;Turko et al 2013) have distinctly high RI values, which can be addressed by measuring 3-D RI maps. RI values of a solution are linearly proportional to its concentration (Yoon et al 2017). Local RI values in cells can be converted into cytoplasmic protein concentration (Barer 1952).…”

“…Furthermore, these RI values are intrinsic quantitative parameters of live cells, and thus can be utilized for a biophysical marker. For example, 2-D and 3-D RI distributional maps can be utilized in label-free cellular identification (Chalut et al 2012;Jo et al 2015;Jo et al 2014;Jo et al 2017;Yoon et al 2017) and long-term growth monitoring (Bettenworth et al 2014;Chalut et al 2012). In addition, RI itself can be used as a new marker which represents cellular states or abnormalities (Lenz et al 2013;Schürmann et al 2016), and the factors which change the intracellular RI are being actively studied (Ekpenyong et al 2013;Wang et al 2011b).…”