Capillary-assisted microfluidic biosensing platform captures single cell secretion dynamics in nanoliter compartments

https://doi.org/10.1016/j.bios.2020.112113Get rights and content

Highlights

  • A novel microfluidic biosensor for real-time monitoring of single cell secretion dynamics has been introduced.

  • The microfluidic device is designed based on capillary-assisted formation of nanoliter compartments.

  • One-step FRET-based biosensor has been employed to monitor single cancer cells' protease activity.

  • Using the single cell protease secretion data, we identified three clusters in the breast cancer cell sample.

Abstract

Cancer cells continuously secrete inflammatory biomolecules which play significant roles in disease progression and tumor metastasis toward secondary sites. Despite recent efforts to capture cancer cells' intercellular secretion heterogeneity using microfluidics, the challenges in operation of these systems as well as the complexity of designing a biosensing assay for long-term and real-time measurement of single cell secretions have become grand research barriers. Here, we present a new capillary-based microfluidic biosensing approach to easily and reliably capture ~500 single cells inside isolated dead-end nanoliter compartments using simple pipette injection, and quantify their individual secretion dynamics at the single cell resolution over a long period of culture (~16 h). We first present a detailed investigation of the fluid mechanics underlying the formation of nanoliter compartments in the microfluidic system. Based on the measurement of single cell capture efficiency, we employ a one-step FRET-based biosensor which monitors the single cancer cells' protease activity. The sensor reports the fluorescent signal as a product of amino acid chain cleavage and reduction in its quenching capability. Using the single cell protease secretion data, we identified modes of cell secretion dynamics in our cell sample. While most of the cells had low secretion levels, two other smaller and more aggressive secretion dynamics were cells with secretion modes that include sharp spikes or slow but progressive trend. The method presented here overcomes the difficulties associated with performing single cell secretion assays, enabling a feasible and reliable technique for high throughput measurement of metabolic activities in cancer cells.

Introduction

Cellular heterogeneity is playing a critical role in driving and regulating processes in human organs as in chemotherapeutic drug response (Stevens et al., 2016), neural systems behavior (Stefen et al., 2018), immune response (Satija and Shalek, 2014), organ regeneration (Hassanzadeh-Barforoushi et al., 2016) and embryonic development (Krieger and Simons, 2015). Over the past decade, single cell analysis tools have been introduced to measure a wide range of heterogeneous factors such as genomic (Shendure et al., 2017), transcriptomic (Pickrell et al., 2010), and proteomic (Wu and Singh, 2012) variations. Despite extensive efforts in obtaining data from single cells using technologies such as Fluorescence-activated cell sorting- FACS (Atkin-Smith et al., 2017), SCI-seq (Vitak et al., 2017), and Drop-seq (Macosko et al., 2015), none of those methods are able to quantify critical dynamic phenomena such as single cell protein secretion and metabolic activity. Such information is invaluable for unraveling mechanisms controlling the single cells' sensing, decision making and signaling (Konry et al., 2016).

Real-time monitoring of the single cell dynamics requires identification of the target analyte, design of an appropriate biosensor, and finally employment of a suitable signal read-out method (Jeknić et al., 2019). Single cell intracellular dynamics have been quantified by targeting biomolecules such as transcription factors (Tay et al., 2010), metabolic byproducts (Li et al., 2019) and second messengers (Colombo et al., 2017) using fluorescent fusion proteins, fluorescent bio-probes and Fluorescence Resonance Energy Transfer (FRET) probes, respectively. Measurement of these intracellular signals is relatively easy as the cell membrane keeps proteins inside the cell with the concentration in a well-detectable range (Lo et al., 2015).

Unlike intracellular events, extracellular secretion is diffused into the cell's surrounding environment, diluted by the media and therefore yields significantly lower concentrations. Measuring such events necessitates cell encapsulation to preserve the secreted molecules and requires incorporation of more sensitive biosensors. Sandwich immunoassays with capture antibody coated on microwells (Shirasaki et al., 2014) or co-encapsulated microspheres (Konry et al., 2011; Chokkalingam et al., 2013) and capture technology on cell membrane (Liu et al., 2019) have been introduced to measure pg/ml of cytokine secretion from single cells. In a new effort, the sandwich assay has also been used for multiplexed single cell secretion using antibody barcode patterning in a nanoliter microtrough array (Chen et al., 2019) and measuring secretion frequency and magnitude through quantitative microengraving technique (Han et al., 2010). While the well-known mechanism of sandwich assay makes it a desirable technique, the binding nature of the assay, its incapability to perform continuous measurement and its reliance on the antibody specificity limits the rate of the measurement and its robustness. To overcome these challenges, label-free methods based on mass spectrometry (Amantonico et al., 2010), electrochemistry (Zhou et al., 2013) and optical biosensing (Li et al., 2018) have been developed. However, the level of complexity associated with these assays as well as their low throughput compromises their widespread adoption. Methods based on one step chemical reaction of the target analytes and recognition molecules are simple, eliminate complicated surface functionalization steps, and are capable of continuous measurement of cell secretion dynamics (Cao et al., 2018; Zhang et al., 2019). For instance, protease activity of single circulating tumor cells (CTCs) (Dhar et al., 2018) and human leukemia cell lines (Haaβ et al., 2015) were quantified using peptide cleavage and uptake by the single cells.

In order to maintain single cells' chemical signature and achieve the level of sensitivity required for low concentration of extracellular secretion, microfluidic methods based on cell encapsulation in microchambers (Eyer et al., 2012; Shen et al., 2015) and pico/nanoliter droplets (J. Collins et al., 2015; Köster et al., 2008) have been proposed. However, both methods entail complex fluidic control such as multilayer valving system (Dang et al., 2019), challenging pressure adjustment for tuning droplet size and droplet immobilization (Courtney et al., 2017). To address these issues, methods for simultaneous generation and immobilization of single cell compartments in the form of semi-droplets have been introduced (Cohen et al., 2010; Hassanzadeh-Barforoushi et al., 2018; Sposito and DeVoe, 2017). Sample digitization into an array of dead-end traps was achieved using geometrical pinning of the liquid sample (Cohen et al., 2010; Sposito and DeVoe, 2017). While providing a self-filling capability, the digitization mechanism works only within a narrow range of geometrical and fluidic properties. Another concern about digitization systems is the formation of trapped air bubbles. This issue has been addressed by designing air vents (Shemesh et al., 2014) or air plugs (Hassanzadeh-Barforoushi et al., 2018) at the end of each trap. Despite operational simplicity offered by such strategy the underlying mechanism governing the air venting and the fluid mechanics of the simultaneous generation and splitting of droplet compartments in those systems remains unanswered.

In this paper, we present a detailed analysis of the engineering of a microfluidic device for the formation of an array of hundreds of bubble-free nanoliter liquid compartments. This analysis provides a deep insight for designing microfluidic systems for long-term single cell encapsulation and real-time secretion profiling. In addition, by incorporating capillary forces, the presented approach offers simplicity and reliability of performing single cell secretion assays. Once single cell encapsulation is achieved, we explore opportunities for performing single cell secretion profiling. We report the rate at which single cells of the same phenotype produce proteases and describe different modes of cell secretion dynamics.

Section snippets

Design and fabrication procedure

Channel designs were drawn in AutoCAD 2015 (Autodesk®) and printed on a high resolution glass photomask. Photolithography using a Karl Suss MA6 Mask Aligner (SUSS MicroTec, Germany) was then used to reflect the mask pattern on an nLOF2020 coated silicon on insulator (SOI) wafer (100 mm wafer diameter, 80 ± 1 μm top layer, 2 μm buried oxide layer, and 500 ± 15 μm base silicon layer). Deep Reactive Ion Etching (DRIE) using STS system was employed to make the final high aspect ratio (1:8) device.

Results and discussion

As shown in Fig. 1A, cancer cell populations inside a tumor or in the circulation system as circulating tumor cells (CTCs) show distinct stages of epithelial-mesenchymal transition (EMT) which corresponds to different levels of invasiveness for these cells. In order to decipher this invasiveness capacity, it is crucial to identify and categorize the cell population based on their heterogeneity in extracellular secretions (Gangoda et al., 2017; Ji et al., 2019); one of which is ECM-degrading

Conclusions

The ability to capture the heterogeneous MMP secretion dynamics among single cancer cells provides valuable information regarding their interaction with their surrounding tissue and the role of individual cells in tumor metastasis (Cathcart et al., 2015). Such information can be very useful in the development of therapeutic drugs targeting these biomolecules. To enable single cell level secretion measurements, we report a new capillary-assisted microfluidic biosensing approach which is based on

Author contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

CRediT authorship contribution statement

Amin Hassanzadeh-Barforoushi: Conceptualization, Investigation, Methodology, Validation, Funding acquisition, Visualization, Data curation, Writing - original draft, Writing - review & editing. Majid Ebrahimi Warkiani: Conceptualization, Resources, Methodology, Writing - review & editing. David Gallego-Ortega: Supervision, Resources, Writing - review & editing. Guozhen Liu: Resources, Writing - review & editing. Tracie Barber: Supervision, Resources, Writing - review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

A. H–B. would like to acknowledge the funding through the postdoctoral writing fellowship scheme provided by the faculty of Engineering, University of New South Wales (UNSW). G. Liu acknowledges the funding support of the ARC Future Fellowship (FT160100039). We would like to thank Garvan Institute of Medical Research/the Kinghorn Cancer Centre for kindly providing the cancer cell lines for this research.

References (50)

  • R. Satija et al.

    Heterogeneity in immune responses: from populations to single cells

    Trends. Immunol., Gene Expr. Immune Syst.

    (2014)
  • M. Wu et al.

    Single-cell protein analysis

    Curr. Opin. Biotechnol., Anal. Biotechnol.

    (2012)
  • A. Amantonico et al.

    Single-cell MALDI-MS as an analytical tool for studying intrapopulation metabolic heterogeneity of unicellular organisms

    Anal. Chem.

    (2010)
  • G.K. Atkin-Smith et al.

    Isolation of cell type-specific apoptotic bodies by fluorescence-activated cell sorting

    Sci. Rep.

    (2017)
  • Z. Chen et al.

    Multiplexed, sequential secretion analysis of the same single cells reveals distinct effector response dynamics dependent on the initial basal state

    Adv. Sci.

    (2019)
  • V. Chokkalingam et al.

    Probing cellular heterogeneity in cytokine-secreting immune cells using droplet-based microfluidics

    Lab Chip

    (2013)
  • D.E. Cohen et al.

    Self-digitization of sample volumes

    Anal. Chem.

    (2010)
  • D.J. Collins et al.

    The Poisson distribution and beyond: methods for microfluidic droplet production and single cell encapsulation

    Lab Chip

    (2015)
  • M. Courtney et al.

    Droplet microfluidic system with on-demand trapping and releasing of droplet for drug screening applications

    Anal. Chem.

    (2017)
  • B.V. Dang et al.

    Microfluidic actuation via 3D-printed molds towards multiplex biosensing of cell apoptosis

    ACS Sens.

    (2019)
  • M. Dhar et al.

    Functional profiling of circulating tumor cells with an integrated vortex capture and single-cell protease activity assay

    Proc. Natl. Acad. Sci. Unit. States Am.

    (2018)
  • K. Eyer et al.

    A microchamber array for single cell isolation and analysis of intracellular biomolecules

    Lab Chip

    (2012)
  • L. Gangoda et al.

    Proteomic profiling of exosomes secreted by breast cancer cells with varying metastatic potential

    Proteomics

    (2017)
  • W. Haaß et al.

    Measurement of separase proteolytic activity in single living cells by a fluorogenic flow cytometry assay

    PloS One

    (2015)
  • Q. Han et al.

    Multidimensional analysis of the frequencies and rates of cytokine secretion from single cells by quantitative microengraving

    Lab Chip

    (2010)
  • Cited by (21)

    • Rapid metabolomic screening of cancer cells via high-throughput static droplet microfluidics

      2023, Biosensors and Bioelectronics
      Citation Excerpt :

      But technical complications, including reliance on skilled operators, specialised equipment and costly reagents, have limited the clinical applicability of this method (Rezaei et al., 2021). As an alternative approach, static droplets have been developed in recent years to overcome the abovementioned limitations (Hassanzadeh-Barforoushi et al., 2018, 2020; Shemesh et al., 2014; Avesar et al., 2018; Boukellal et al., 2009). However, low-throughput and low-flexibility in down-sizing droplets (i.e., inability to generate sub-picolitre droplets) have remained a major barrier for these devices, only allowing them to be suitable for a limited number of applications that deals with small samples (Rezaei et al., 2021).

    • High-resolution imaging of protein secretion at the single-cell level using plasmon-enhanced FluoroDOT assay

      2022, Cell Reports Methods
      Citation Excerpt :

      Microfluidics-based single-cell secretion assays can achieve high-throughput measurements; however, they are prone to clogging by micron-size fragments and dust, affecting the reliability of the results (Jammes and Maerkl, 2020). Capillary-assisted microfluidics compartments can provide spatiotemporal dynamics of the single-cell secretion but fail to describe the secretion pattern around the cell (Hassanzadeh-Barforoushi et al., 2020). Some of these systems also incur high background due to the absence of washing steps (Choi, 2020).

    View all citing articles on Scopus
    View full text