Home / microfluidic research summaries / Light-field flow cytometry (LFC): high-throughput, high-resolution volumetric imaging for multiparametric 3D single-cell analysis
Microfluidic research summary

Published on 02 September 2024

Light-field flow cytometry (LFC): high-throughput, high-resolution volumetric imaging for multiparametric 3D single-cell analysis

Abstract

Imaging flow cytometry (IFC) combines the strengths of flow cytometry and fluorescence microscopy while facing challenges in higher-resolution and higher-dimensional data acquisition, which often requires sequential acquisition methods or complex instrumentation that limits throughput and accessibility. To conquer these challenges, this study introduces a light-field flow cytometer (LFC) for high-throughput, high-resolution, volumetric, and multiparametric single-cell analysis1. The LFC integrates an advanced light-field optofluidic system, achieving a near-diffraction-limited resolution and multi-color scanless 3D imaging at high speeds. We demonstrate the capabilities of the system by examining a variety of biological samples. We expect the LFC system to be an accessible and compatible imaging platform to significantly advance cell biology and translational research.

Introduction

Imaging Flow Cytometry

Imaging flow cytometry (IFC), integrating two vital tools for biological and medical research: flow cytometry and fluorescence microscopy, allows for both rapid analysis of diverse cellular populations and high-resolution image acquisition of individual cells2-5. Combining its strength of high-throughput, multiparametric single-cell analysis with rich spatial details, high sensitivity, and molecular specificity, the IFC technologies have been applied across various basic and translational fields. Recent years have witnessed significant advancements in the imaging capabilities of IFC, including speed, sensitivity, and resolution. Nevertheless, current IFC systems remain disadvantageous in higher-resolution and higher-dimensional data acquisition compared with other single-cell imaging platforms. Alternatively, 3D IFC developed based on 3D microscopy techniques may suffer from trade-offs between 3D resolution, volumetric coverage, and throughput due to sequential acquisition, as well as increased instrumental complexity and limited accessibility on commonly used microscopic platforms. Therefore, IFC-based platforms for single-cell investigations have yet to find the optimal balance in revealing 3D subcellular details with high resolution, throughput, sensitivity, and straightforward instrumentation.

Light Field Microscopy

Light-field microscopy (LFM) showcases a particularly appealing solution for capturing 3D details of fast-moving single-cell specimens by simultaneously recording the spatial-angular information of the light field. Such a scheme enables computational reconstruction of the sample volume using only a single camera frame6,7. Moreover, recent advancements in Fourier LFM have further improved image quality and computational efficiency8-11, facilitating 3D subcellular, millisecond spatiotemporal investigations across various biological systems ranging from functional brains9,12, organoids13 to single-cell specimens14,15. Compared to other 3D techniques, the light-field approach allows for single-shot 3D image acquisition and simple instrumentation on epi-fluorescence platforms, which are highly desirable features for cytometric imaging.

Aim of the study

The aim of this work is to develop a light-field flow cytometer (LFC) for 3D volumetric, high-throughput, and multiparametric analysis of single-cell populations. By incorporating a high-resolution light-field optofluidic platform, hydrodynamic focusing, and stroboscopic illumination, the LFC system achieves a near-diffraction-limited and multi-color resolution of various 3D subcellular morphologies at high speeds. By examining and quantifying phantoms and biological morphologies, functions, and heterogeneities, this study presents multiparametric singlecell analysis for peroxisomes and mitochondria in cultured cells, isolated cells from mice and humans, apoptotic alterations in staurosporine-treated Jurkat cells, and tdTomato expression following Cre mRNA delivery in mice.

Experiment Setup

Materials

  • Eclipse Ti2-U microscope (Nikon Instruments)Nano-F100S piezo nano-positioner (Mad City Labs)
  • Multicolor laser lines (MPB Communications)
  • MS2000 micro-positioning system (Applied Scientific Instrumentation)
  • Optics and optomechanics (Edmund Optics, Thorlabs)
  • Customized microlens array (RPC Photonics)
  • ORCA-Flash 4.0 V3 sCMOS (Hamamatsu Photonics)
  • OB1 MK3+ microfluidic flow controller (Elveflow)
  • MFS3 microfluidic flow sensor (Elveflow
  • 10001824 microfluidic chips (ChipShop)
  • LVF-KPT-M-2 microfluidic reservoirs (Darwin Microfluidics)
  • BD-PLSTPK-LL-01 syringe (Darwin Microfluidics)
  • Elveflow Software Interface (Elveflow)
  •  

Approach

Light-field imaging system

The high-resolution Fourier light-field microscopy system, as shown in Figure 1, was developed on an epi-fluorescence microscope with an oil-immersion lens (100x, NA=1.45). A piezo nano-positioner was used for precise positioning. Multicolor laser lines (488 nm, 561 nm, 647 nm), modulated by a stroboscopic illumination controlling module (SICM) and reflected by a quadband dichroic mirror (DM), illuminate samples in microfluidic flow. An oil-immersed objective lens (OBJ), emission filter (EF), and tube lens (TL) create wide-field images of the fluorescence at the native image plane (NIP). A Fourier lens (FL, fFL = 275 mm) optically transforms the NIP onto its back focal plane, where a microlens array (MLA, fML = 117 mm) partitions the light field to generate elemental images onto an sCMOS camera (CAM) located at the MLA back focal plane.

Figure 1. (a) A LFC system. (b) Microfluidic setup. The microfluidic chip has a main sample channel with two side channels. The pressure difference is adjusted to create the proper hydrodynamic focusing for the sample solution. (c) Stroboscopic illumination is synchronized within each camera exposure to minimize motion blur (insets i and ii). Multiple illumination cycles can be generated within each exposure. (d) Axial stack projection of the hybrid point-spread function (hPSF) through the MLA within an axial range of 10 µm. (e) Light-field image formation, capturing the spatial-angular information in an uncompromised manner. (f) Image processing pipeline containing image conversion, ACsN denoising, elemental image selection, and deconvolution-based image reconstruction. Scale bars: 100 μm (b), 5 μm (c). 10 μm (d).

Flow cytometer and microfluidic preparation  

The microfluidic setup was constructed with a 3-channel microfluidic flow controller (OB1, Elveflow), a microfluidic flow sensor (MSF, Elveflow), microfluidic chips, microfluidic reservoirs, a syringe, and a waste tank. Prior to the experiments, the reservoirs were filled with deionized water. During the experiments, we first opened the pump side valve (Valve 1) while blocking the syringe side valve (Valve 2) to flush the chip with deionized water from all three channels, effectively cleaning the channels before the measurements. Following the pre-experimental cleaning, we halted the pump and replaced the solution in the two small reservoirs connected to the side channels of the chip with HBSS (saline solution). We then reactivated the pump to establish a stable, focused flow. Upon achieving the flow without bubbles in the channels, we closed Valve 1 and opened Valve 2. The samples were injected into the tubes and the chip by a syringe. Once the samples filled the tubes (1-2 mL, determined empirically), we closed Valve 2 and reopened Valve 1, allowing the samples to be automatically and controllably introduced into the chip.

Figure 2. Schematic of detailed microfluidics illustration. CH1~4: Channel #1~4.

Key Findings

Imaging capability characterization of the LFC system

Fluorescent Microspheres

The LFC system was characterized using a mixture of Tetra-Speck fluorescent microspheres of varying diameters (200 nm, 1 μm, 2 μm, and 4 μm) injected at a flow rate of 0.4-0.6 μL/min. The high SNR (Signal to Noise Ratio) of microspheres enables a spatial resolution measurement between 300-600 nm and a >5× extended depth of focus (~6 μm) compared to conventional epi-fluorescence microscopy. Reconstructed objects revealed four distinct populations, with microsphere diameters and intensities matching expected values, demonstrating reliable identification within the phantom mixture.

Figure 3. Characterization of LFC using fluorescent microspheres. (a) A mixture of deep red TetraSpeck fluorescent microspheres with diameters of 200 nm, 1 μm, 2 μm, and 4 μm. (b-e) 3D reconstructed images of the microspheres with intensity profiles in X, Y, Z, respectively. (f) Histogram counts of the microsphere diameters rendered based on the measured 3D volumes for multi-color excitations at 647 nm. (g) Corresponding scatter plots of the fluorescence intensity as a function of the microsphere volumes in (f). The color gradient serves to visualize the density distribution of the beads.

Mitochondria and Peroxisomes

Two-color imaging of mitochondria and peroxisomes in HeLa cells was conducted using MitoTracker and peroxisome-GFP at the same flow rate (0.4-0.6 μL/min). Reconstructed images depicted the 3D spatial relationship between peroxisomes and mitochondria across ~6 μm cellular thickness, with structural variations resolved as close as 400-600 nm. The results, enhanced by effective denoising, showed high resolution and volumetric capabilities, providing accurate reconstructions compared to other modalities like epi-fluorescence and 3D structured-illumination microscopy (SIM).

Figure 4. Denoised two-color LFC images of mitochondria (a) and peroxisomes (b), their corresponding reconstructed 3D image (c) of HeLa cells. Insets in (a,b) show the corresponding zoomed-in elemental images. (d,e) Zoomed-in images in X-Y (d) and Z (e) of the regions of interest. Scale bars: 10 μm (a, b), 5 μm (a insets, b insets), 500 nm (d,e).

Morphological features of isolated cells from mouse and human

Mouse Blood Cells

The LFC system enabled 3D cytometric imaging of membrane-labeled blood cells extracted from adult mice at a rate of approximately 600 cells/sec. The imagery displayed high specificity and sensitivity for differentiating various 3D morphological features on a cell-by-cell basis with high throughput (~2,300 cells/sec). Detailed 3D morphological features of the cells were accurately captured and analyzed.

Human T lymphocytes

Membrane and nucleus of human activated T cells were labeled and imaged at a rate of ~300 cells/sec. The 3D hollow structures of the cell membrane, enclosing the nucleus, were quantified, revealing two distinct sizes for the cell membrane (7.99 µm) and nucleus (6.57 µm).

Figure 5. Comparative analysis of cell morphologies in isolated mouse and human cells. (a) 3D reconstructed images displaying a variety of shapes in membrane-labeled mouse blood cells. (b,c) Ellipsoid-fitted radii of cells Ra, Rb, Rc (Ra>Rb>Rc) with k-means clustering, categorizing cells based on their morphologies. (d) 3D reconstructed volumes of the membrane (magenta)- and nucleus (blue)-labeled human activated T cells in flow. (e,f) Axial stacks (e) and two-color overlay (f) of a human activated T cell in (d) across a depth range of 6 μm. (g) Intensity-to-volume plots for ~300 human activated T cells. (h) The N:C ratio of human activated T cells as a function of cell volume. Scale bars: 1 μm (e).

Morphological changes in apoptotic Human T lymphocytes

LFC was used to investigate 3D subcellular morphological changes in human T lymphocyte (Jurkat) cells due to STS-induced apoptosis. Jurkat cells were treated with 1-µM STS for 30, 60, 120, and 300 minutes. Significant morphological changes were observed in the nuclei and mitochondria of treated Jurkat cells. Over time, nuclei displayed reduced volumes and fragmented micronuclei, with over 53% of cells showing apoptotic nuclei after 5 hours of treatment. During nuclear fragmentation, mitochondria increasingly enclosed the interstitial spaces of the micronuclei. The study highlights the effectiveness of LFC in elucidating subcellular morphological alterations associated with apoptosis and other cellular functions and dysfunctions.

Figure 6. (a,b) Denoised light-field images of mitochondria (a) and nucleus (b) in a live Jurkat cell without STS treatment. Corresponding insets show the zoomed-in elemental images. (c,d) 3D reconstructed image (c) of the cell in (a, b). (d-g) 3D visualization of Jurkat cells treated with STS for 30 (d), 60 (e), 120 (f), 300 (g) min, respectively, exhibiting fragmented and condensed nuclear dispersion throughout the cells. (h) Percentage of the cells (n > 100) showing apoptotic cell morphology for each STS treatment period. (i) Average volumes of the micronuclei in cells for each STS treatment period. (j) Box plots illustrating the distribution of the volume of mitochondria enclosed within micronuclei, relative to the total volume of micronuclei and mitochondria for individual cells across various STS treatment durations. Scale bars: 10 μm (a, b).

Lipid Nanoparticle mRNA delivery to mouse cells

Lipid nanoparticles (LNPs) carrying mRNA have shown promising results in COVID vaccines and clinical trials, while visualizing and quantifying functional mRNA delivery and subsequent protein expression remain challenging. Two-color LFC provided high sensitivity for 3D visualization of individual cells and gene expression. Liver cells showed a high percentage of tdTomato+ expression (79.41%) compared to the spleen (13.45%) and lung (11.39%). The study demonstrates that 3D IFC can effectively visualize and quantify functional mRNA delivery and subsequent protein expression, highlighting the liver’s higher targeting efficiency by LNPs.

Conclusion

In conclusion, the Light-Field Flow Cytometry (LFC) system significantly enhances cell analysis by enabling high-sensitivity, 3D volumetric, and multiparametric data acquisition. This allows for the comprehensive examination of subcellular morphology, behavior, and interactions within their native 3D contexts. This system is designed with low instrumental complexity, allowing it to be compatible with commonly used epi-fluorescence microscopes and microfluidic devices. The Fourier light-field approach is particularly noteworthy for its high scalability to meet various acquisition requirements while retaining its 3D and single-shot capabilities. The functionality of the LFC system, such as the depth of focus and 3D resolution, can be further extended with various optical and computational frameworks. Particularly, deep learning offers a viable alternative to traditional deconvolution algorithms, significantly accelerating the generation of high-quality 3D reconstructions and benefiting LFC analysis in large cellular populations. The approach holds great potential for extensive application in both fundamental and translational research, with the possibility of seamless integration with single-cell genomics, microscopy-based screening and diagnosis, and image-enabled sorting. We expect the LFC system to be a promising approach for various cytometric imaging applications across biology, pharmacology, and medical diagnostics.

Authors Information

  • Keyi Han: Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, GA, USA
  • Shu Jia: Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, GA, USA; Parker H. Petit Institute for Bioengineering and Biosciences, Georgia Institute of Technology, Atlanta, GA, USA
  • Xuanwen Hua: Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, GA, USA

Dr. Xuanwen Hua earned his Bachelor of Science degree in applied physics from the University of Science and Technology of China in 2016. In the same year, he joined Dr. Shu Jia’s lab as a PhD student. Xuanwen completed his Ph.D. in biomedical engineering from the Georgia Institute of Technology and Emory University in May 2024 and began his postdoctoral position immediately after. With more than six years of research experience in optical microscopy, his expertise includes optical wavefront modulation, super-resolution microscopy, light-field microscopy, imaging flow cytometry, and deep learning.

Keyi Han

Dr. Shu Jia is an associate professor in the Coulter Department of Biomedical Engineering at Georgia Institute of Technology and Emory University. He received his Bachelor’s degree in Electronic Engineering from Tsinghua University and his Master’s and Ph.D. degrees in Electronic Engineering from Princeton University, where he studied under Dr. Jason Fleischer. He completed his postdoctoral training in Xiaowei Zhuang’s lab at Harvard University. His research interests include systems biophotonics, single-molecule biophotonics, super-resolution, and advanced optical microscopy, imaging instrumentation and devices.

Xuanwen Hua

Keyi Han earned his Bachelor of Science degree in Biomedical Engineering from Boston University in 2021. He later joined the Coulter Department of Biomedical Engineering at Georgia Institute of Technology and Emory University as a PhD student, mentored by Dr. Shu Jia. His research interests span optical microscopy, computational imaging, super-resolution microscopy, and imaging flow cytometry.

 

Shu Jia
References
  1.     Hua X, Han K, Mandracchia B, et al. Light-field flow cytometry for high-resolution, volumetric and multiparametric 3D single-cell analysis. Nat Commun. Mar 4 2024;15(1):1975. doi:10.1038/s41467-024-46250-7
  2.     Han Y, Gu Y, Zhang AC, Lo YH. Review: imaging technologies for flow cytometry. Lab Chip. Nov 29 2016;16(24):4639-4647. doi:10.1039/c6lc01063f
  3.     Stavrakis S, Holzner G, Choo J, deMello A. High-throughput microfluidic imaging flow cytometry. Curr Opin Biotechnol. Feb 2019;55:36-43. doi:10.1016/j.copbio.2018.08.002
  4.     Rees P, Summers HD, Filby A, Carpenter AE, Doan M. Imaging flow cytometry. Nature Reviews Methods Primers. 2022/11/03 2022;2(1):86. doi:10.1038/s43586-022-00167-x
  5.     Doan M, Vorobjev I, Rees P, et al. Diagnostic Potential of Imaging Flow Cytometry. Trends Biotechnol. Jul 2018;36(7):649-652. doi:10.1016/j.tibtech.2017.12.008
  6.     Levoy M, Ng R, Adams A, Footer M, Horowitz M. Light field microscopy. 2006:924-934.
  7.     Levoy M, Zhang Z, McDowall I. Recording and controlling the 4D light field in a microscope using microlens arrays. Journal of microscopy. 2009;235(2):144-162.
  8.     Llavador A, Sola-Pikabea J, Saavedra G, Javidi B, Martinez-Corral M. Resolution improvements in integral microscopy with Fourier plane recording. Opt Express. Sep 5 2016;24(18):20792-8. doi:10.1364/OE.24.020792
  9.     Cong L, Wang Z, Chai Y, et al. Rapid whole brain imaging of neural activity in freely behaving larval zebrafish (Danio rerio). Elife. Sep 20 2017;6:e28158. doi:10.7554/eLife.28158
  10.   Scrofani G, Sola-Pikabea J, Llavador A, et al. FIMic: design for ultimate 3D-integral microscopy of in-vivo biological samples. Biomed Opt Express. Jan 1 2018;9(1):335-346. doi:10.1364/BOE.9.000335
  11.   Guo C, Liu W, Hua X, Li H, Jia S. Fourier light-field microscopy. Opt Express. Sep 2 2019;27(18):25573-25594. doi:10.1364/OE.27.025573
  12.   Zhang Z, Bai L, Cong L, et al. Imaging volumetric dynamics at high speed in mouse and zebrafish brain with confocal light field microscopy. Nat Biotechnol. Jan 2021;39(1):74-83. doi:10.1038/s41587-020-0628-7
  13.   Liu W, Kim GR, Takayama S, Jia S. Fourier light-field imaging of human organoids with a hybrid point-spread function. Biosens Bioelectron. Jul 15 2022;208:114201. doi:10.1016/j.bios.2022.114201
  14.   Hua X, Liu W, Jia S. High-resolution Fourier light-field microscopy for volumetric multi-color live-cell imaging. Optica. May 20 2021;8(5):614-620. doi:10.1364/optica.419236
  15.   Sims RR, Rehman SA, Lenz MO, et al. Single molecule light field microscopy. Optica. Sep 20 2020;7(9):1065-1072. doi:10.1364/Optica.397172
Want to run a similar experiment? Feel free to contact us at: contact@elveflow.com
Elveflow team at work

    How can we help you?




    I hereby agree that Elveflow uses my personal data

    We will answer within 24 hours

    Contact
    How can we help you?
    Quoteor technical request Job application Job
    application
    Collaboration or partnerships Collaborations
    or partnerships
    Customer support Customer
    support
    Others questions Other

      Get a quote




      We will answer within 24 hours

      By filling in your info you accept that we use your data.

      Contacting for
      a job application?
      We are happy that you are interested in Elveflow. You can apply to our open jobs or send us your open application on WelcomeToTheJungle. Over here!

        Collaborations




        We will answer within 24 hours

        By filling in your info you accept that we use your data.

          Need customer support?







          I hereby agree that Elveflow uses my personal data

          We will answer within 24 hours

            How can we help you?




            We will answer within 24 hours

            By filling in your info you accept that we use your data.