Advancing single-cell proteomics and metabolomics with microfluidic technologies

Yifan Liu , Xuyue Chen , Yiqiu Zhang and Jian Liu *
Jiangsu Key Laboratory for Carbon-based Functional Materials and Devices, Institute of Functional Nano and Soft Materials (FUNSOM), Soochow University, Suzhou, Jiangsu Province 215123, China. E-mail:

Received 5th August 2018 , Accepted 29th September 2018

First published on 1st October 2018

Recent advances in single-cell analysis have unraveled substantial heterogeneity among seemingly identical cells at genomic and transcriptomic levels. These discoveries have urged scientists to develop new tools that are capable of investigating single cells from a broader set of “omics”. Proteomics and metabolomics, for instance, are of particular interest as they are closely correlated with a dynamic picture of cellular behaviors and phenotypic identities. The development of such tools requires highly efficient isolation and processing of a large number of individual cells, where techniques such as microfluidics are extremely useful. Here, we review the recent advances in single-cell proteomics and metabolomics, with a focus on microfluidics-based platforms. We highlight a vast array of emerging microfluidic formats for single-cell isolation and manipulation, and how the state-of-the-art analytical tools are coupled with such platforms for proteomic and metabolomic profiling.


Over the past decade, there have been remarkable advances in the field of single-cell analysis. A diverse set of innovative tools has been developed, enabling the high-throughput profiling of important biomolecules in individual cells, which can either be extracted from complex multicellular systems or represent distinct cellular states. For example, single-cell RNA sequencing techniques permit transcriptomic profiling of up to hundreds of thousands of single cells in a single experiment.1 Functional studies of thousands of single cells at the proteomic level can be achieved on a single microscope slide using the single-cell western approach.2,3 These powerful single-cell analysis tools offer new insights into the phenotypic identity of cells and remarkable opportunities in various fields, such as cancer research and drug discovery.4

It has been increasingly recognized that the genotype of a cell defines its overall potential, but the phenotype describes its functions in an organism.5 In addition to genomic characteristics, the phenotypic identity of cells is affected by various non-genetic factors such as developmental states and environmental influences. Therefore, proteomic and metabolomic studies are critical because they may provide pictures of cellular behaviors and phenotypic identities in a dynamic and closely correlated manner (Fig. 1a). Proteins bridge the genetic information and the cellular biological functions as key executors and signaling regulators. Relative to their cognate mRNAs, proteins typically exhibit a broader dynamic range of expression levels and higher stability and tolerance to transcriptional noise.6 Metabolites can directly reflect, and also modulate, cell responses and phenotypical changes by playing a crucial role in energy balance, intercellular signaling, and many other cellular functions throughout the lifespan of cells.7,8 Large-scale protein or metabolite profiling of single cells has been proposed as a way to reveal the important features in biological events that are of great research interest, including the metastasis of circulating tumor cells and stem cell differentiation9 (Fig. 1b).

image file: c8an01503a-f1.tif
Fig. 1 The significance of single-cell proteomics and metabolomics in the puzzle of “omics”. (a) Schematic drawing with a highlight of the proteome and metabolome which are dynamic and more closely correlated with the cellular phenotypes in various “omics” of a single cell. (b) Conceptual illustration showing that the single-cell methods are capable of identifying cell types, and distinguishing seemingly similar cell groups, compared to the population-based methods.

However, profiling the proteome and metabolome of individual cells at the single-cell level remains a serious challenge due to the high diversity and large dynamic range of the cellular proteome and metabolome, the difficulty of amplifying them, and the lack of powerful tools compared with the number of tools available for nucleic acid analysis (for example, next-generation sequencing for single-cell genomics and transcriptomics).5,10,11 Conventional techniques, such as flow or mass cytometry,12 and fluorescence-based detection are not yet ready to be applied to the “omics” level of single cells. Mass spectrometry (MS),13 with demonstrated success in proteomic and metabolomic profiling at the population level and the recently improved detection limit to the range of sub-attomole to zeptomole,14–16 is considered a potential candidate for studying single-cell proteomics and metabolomics. However, there are still practical issues to be overcome, such as the difficulty in interfacing single-cell samples with the MS instrument.5 In single-cell measurements, sample handling in a highly precise format is critical; otherwise, any small losses of analytes that are negligible in cell population-based methods may lead to dramatic fluctuations at the single-cell level. Single-cell data also require simultaneous profiling of a large number of individual cells to overcome the intrinsic noise and to improve the data quality.6 These demands cannot be met by simple modifications of conventional cell population-based methods and regular sample-handling techniques such as micropipetting.

Recently, the landscape of the single-cell research field has been changed rapidly with the aid of microfluidic techniques. Although not an analysis method per se, microfluidics is capable of providing isolation and highly efficient processing (e.g., isolation, adding reagents, lysing, culturing and transporting) of large numbers of single cells in a miniaturized device. It brings unique advantages, including a significant enhancement in the throughput performance and cost-effectiveness, simplification of the workflow, and improvement of the assay consistency. With respect to microfluidics for single-cell analysis, a few comprehensive reviews can be found in the literature; however, they focus on either fundamental microfluidic techniques9,17 or general “omics”.6 Here, a review is provided on how the latest advances in microfluidic technologies have been helpful in addressing the challenges involved in single-cell analysis, with a focus on proteomics and metabolomics. We first highlight the cutting-edge microfluidic platforms used for isolating and manipulating single cells. We then review how the critical information of single-cell proteomics and metabolomics can be obtained by the microfluidic technologies. Finally, we discuss the usefulness of these methods and new insights that have been gained in the study of specific cell types of great interest.

Single cell isolation and manipulation

Single cell isolation is required in most of the current single-cell omics approaches. Unless working with circulating cells (e.g., immune cells in the blood), single-cell isolation begins with the step of dissociating a solid block of tissue into a clean, single-cell suspension. This is accomplished enzymatically by digesting the binding proteins in the extracellular matrix. However, separating individual cells can be problematic for some tissues (e.g., brain tissue, where cells are highly intertwined), which is a particularly pressing challenge for the profiling of human tissues, where samples are often obtained post-mortem. In contrast, the suspension of individual nuclei can be obtained easily, even from preserved tissue. Recently, several reports have demonstrated the use of isolated nuclei in epigenetic and transcriptomic profiling of single brain cells.18,19 For proteomics and metabolomics, the idea can be potentially applied to specific endonuclear targets, such as nuclear proteins.

Once a clean, single-cell suspension is obtained, many single-cell approaches require the isolation of individual cells in separate containers for subsequent processing. Microfluidics is particularly useful for this purpose, as it offers not only the ease of isolating large numbers of single cells but also the ability to perform manipulations. In microfluidics, single-cell isolation is achieved by physically confining the cells within microfluidic structures, among which nano-liter wells, droplets and valve-controlled channels are most extensively used in single-cell analyses.6 Nano-liter well devices20 comprise high-density arrays of wells (or compartments) that are lithographically fabricated onto polydimethylsiloxane (PDMS), glass or silicon. Such wells serve as miniature containers for single cells that are loaded (at a low density so that each well houses one cell at most), sorted, or printed onto the device. The wells can be sealed with capping structures (e.g., glass slides) or liquids immiscible with water (e.g., oil), providing an isolated environment for reactions, incubation and observation. If required, the caps can be functionalized or made selectively permeable, allowing cell secretion products (e.g., cytokines) to be profiled. Droplet-based systems21 feature micrometer-scaled, surfactant-stabilized aqueous droplets in an inert carrier oil. The surfactants also help prevent cross-contamination by limiting the diffusion of analytes (e.g., DNA) into oil and neighboring droplets. With single cells encapsulated, the droplets can be subsequently merged, incubated, reinjected, and sorted with a set of microfluidic modules to perform the desired assays. The droplets can also be broken to extract the cellular components for downstream analysis such as sequencing. Valve-based devices22 feature parallel circuits of microchannels coupled to pressure-controlled valves. With precise control of the valve combinations, a series of operations can be conducted on-chip, such as single-cell isolation, reagent addition, lysis and lysate retrieval, thereby enabling automation and parallelization of complex biological assays. Such well-established platforms are now being broadly used in research laboratories, and several of them have been commercialized. Comprehensive reviews of these single-cell platforms can be found elsewhere.6,9,17 We thus aim to highlight a few emerging and interesting microfluidic platforms for single-cell isolation and manipulation that may surpass the limitations in existing platforms.

Hybrid platforms

Each of the aforementioned microfluidic platforms has its own advantages and drawbacks. Droplet-based approaches, for example, feature a high throughput of up to tens of thousands of droplets generated per second; yet on-demand control and manipulation of one or a few droplets of interest are challenging. A promising route to overcome this limitation is to combine the core components in different microfluidic platforms to form a new hybrid platform. Nano-liter wells can be integrated with droplet-based approaches, which further offers static positioning of each droplet/cell, simplifying the manipulation and ensuring compatibility with many measurement modalities. Shemesh and coworkers developed an innovative microfluidic device where stationary nano-liter droplets can be directly generated in a nano-liter well array when the dispersed medium is being replaced by a continuous oil phase.23 This device was used for real-time monitoring of the metabolic activity of single adherent cells. Using a similar concept, Amselem and coworkers developed a droplet array platform, in which anchored droplets or hydrogel beads can be formed in the wells (Fig. 2a).24 The encapsulation and growth of single bacterial cells in the anchored droplets were performed, as well as an on-chip antibiogram by generating a concentration gradient of gentamycin across the nano-liter well array. They also demonstrated that desired droplets can be readily extracted using an infrared laser beam for downstream analysis. In several other reports,25 single cells were pre-encapsulated in the droplets, which were subsequently loaded onto the nano-liter well array. The loading efficiency can be improved by proper design of the nano-liter well geometry or with guiding modules such as a mesh grid.26 A more robust and deterministic route of droplet loading is to use droplet printing techniques. Leung and coworkers developed an approach in which a commercially available liquid dispenser was used to deposit droplets of single cells or reagents onto a planar substrate (Fig. 2b).27 Using this method, they performed single-cell DNA amplification and single nucleotide variation (SNV) detection. Zhu and coworkers reported a semi-contact droplet dispensing approach for more reliable and continuous printing of nano-liter droplets and reagent addition in the oil phase on a nano-liter well array substrate.28 Other than single-cell analysis, droplet-nano-liter well hybrid platforms have also been used in applications where stationary positioning and measurement of droplets are desired, such as drug screening,29 protein crystallization28 and digital bioassays (e.g., digital PCR).
image file: c8an01503a-f2.tif
Fig. 2 Emerging single-cell isolation and manipulation techniques. (a–c) Hybrid platforms and (d) virtual microfluidics. (a) Bright field (top) and fluorescence (bottom) snapshot of the anchored droplet chip, which displays 1495 droplets containing colonies of fluorescent bacteria.19 (b) Illustration of the nano-liter-volume droplet dispensing for single-cell analysis (left) and a planar substrate with printed dye droplets.22 (c) Schematic illustration of the printed droplet microfluidics (top) and micrographs showing various combinations of the dye or single cells printed in individual nano-liter wells (bottom).26 (d) A comparison between conventional single-cell isolation formats and virtual microfluidics (top) and fluorescence micrographs showing isolated single bacteria (green) and the amplified clusters (red).35

In the approaches discussed above, however, single-cell encapsulation in the droplets was achieved by limited dilution, and the actual portion of droplets containing single cells was less than 5%. This low proportion leads to poor analysis efficiency, as the vast majority of anchored droplets are empty. One approach to address this issue is to improve the encapsulation efficiency with techniques such as inertial ordering;30 another approach is to pre-screen the droplets with FACS or a droplet sorter. Notably, Cole and coworkers developed an innovative technique, printed droplet microfluidics (PDM), which combines droplet sorting, module printing and the droplet-nano-liter well platform (Fig. 2c).31 PDM features a fluorescence-activated droplet sorter coupled to a motorized stage that screens the droplets cycling through, sorts the desired droplets containing the cell or reagent, and dispenses them to any given spots in the nano-liter well array. This enables on-demand printing of single cells and the addition of reagents in a fast and highly efficient manner. Using PDM, the authors demonstrated the printing of single PC3 prostate cancer cells and complex combinations of the cells at every position over 100 nano-liter wells and confirmed that 98 contained the correct contents (i.e., printing accuracy: 98%). Potential applications of this platform that the authors mentioned are found in single-cell multiomics and combinatorial biology, where combinations of cells that form functional tissue structures can be studied.

Nano-liter well- or droplet-based approaches can be coupled with valve-based systems for enhanced controllability and programmability. Leung and coworkers developed a microfluidic device in which integrated microvalve circuits are combined with microdroplets to perform single-cell manipulation and analysis.32 While retaining the well-compartmentalized nature of droplet-based systems, the device allows the execution of a series of complex assays, as in the valve-based systems. Demonstrated applications of the device include phenotypic sorting of single bacteria, taxonomic identification and genomic analysis of single cells from diverse environmental samples. Yang et al. reported a valve–nano-liter well hybrid platform for the isolation and analysis of single cancer cells that features a self-seeding nano-liter well design for single-cell capture and a microvalve circuit for peristaltic pumping of reagents.33 Furthermore, the dominant microfluidic platforms are often coupled with advanced manipulation techniques, such as using optical34 or magnetic tweezers35 and surface acoustic waves36 (SAWs) and dielectrophoresis,37 for improved performance. Representative examples include a SAW-integrated nano-liter well device for highly effective single-cell capturing developed by Collins et al.36 and a droplet microfluidic device in which magnetic tweezers are applied for programmable and high-throughput multistep assays, developed by Ali-Cherif and coworkers.35

Virtual microfluidics

The term “virtual microfluidics” was initially used in the literature to refer to microfluidic channels and wells that do not have actual sidewalls.38 In such constructions, liquid confinement is achieved by pre-patterned surface wettability, and often, the liquid can be actively manipulated with integrated electrowetting modules.39 Recently, Xu and coworkers reported a new type of virtual microfluidics (Fig. 2d), a bulk polyethylene glycol (PEG) hydrogel platform for single-cell isolation without the need for discrete partitions.40 Featuring a mesh size of 25 nm, the hydrogel immobilizes and isolates single cells and their DNA in a diluted suspension, while allowing limited diffusion of smaller molecules (e.g., enzymes). Using this approach, the authors performed high-quality whole-genome amplification and sequencing of single bacteria. Importantly, this report demonstrated that microfluidic assays can be performed in a user-friendly “non-microfluidic” manner, without the need for microfabricated devices and complex equipment, possibly empowered by this new kind of virtual microfluidics.

With hydrogel-based approaches included, a broader set of studies can fit into the category of virtual microfluidics. In fact, there has recently been increasing attention paid to the use of hydrogels for single-cell analysis. A major advantage of hydrogels, as briefly mentioned above, is that they provide a mildly confined environment where cells and large molecules (genomic DNA) can be trapped, but small molecules are rather free to migrate, which facilitates serial reagent exchange. In a recent study, Lan et al. reported single-cell isolation, lysis, genome fragmentation and barcoding in agarose microgel beads.41 The use of agarose beads allows direct washing and processing of encapsulated cells, which are impractical to perform with conventional droplet-based microfluidics. If desired, hydrogels can be modified to be photodegradable. Son and coworkers developed a reconfigurable microfluidic platform in which a photodegradable hydrogel array was a pattern for single CD4 cell capturing and secretion analysis.42 Upon UV exposure, the gel islands can be readily degraded, allowing the retrieval of selective cells. Given their unique properties, hydrogels are an ideal template for single-cell patterning and the study of single-cell growth and differentiation. Moffitt and coworkers reported an agarose-based microfluidic device for the analysis of single bacteria.43 The device features linear tracks of submicron agarose pads for the controlled growth of single E. coli cells. Khetan et al. studied the differentiation behavior of human mesenchymal stem cells in three-dimensional hydrogels and found that degradation-mediated traction can affect the fate of the cells.94 One potential question is whether hydrogels, or virtual microfluidics, can serve as an ideal platform for single-cell proteomics and metabolomics. Indeed, given their porous nature, hydrogels alone cannot confine proteins and metabolites. However, they can be easily functionalized to immobilize specific low-molecular-weight species of interest. The limited diffusion of these molecules in the gel matrix may further enhance the capture efficiency. Given this and the numerous advantages of hydrogels, we expect to see an increased number of virtual microfluidic approaches for single-cell proteomics and metabolomics.

Single-cell proteomics

The increasing need for characterizing protein expression and modification at the single-cell level calls for an approach that is robust, scalable and multiplexed. Achievement of this goal requires innovations in both handling techniques, i.e., microfluidics, for improved reliability and scalability, and analytical tools for enhanced sensitivity and multiplexing capability. We review how microfluidics has been coupled with a wealth of analytical methods, including DNA barcoding and next-generation sequencing (NGS), fluorescence immunoassays, electrophoresis and mass spectrometry, to yield novel and functional approaches for single-cell proteomics.

DNA barcoding and sequencing

Nucleic acid barcodes have been broadly applied in single-cell transcriptomics for multiplexing. Similarly, antibodies can be labeled with DNA barcode tags, allowing proteins to be profiled with DNA sequencing. Shahi and coworkers developed a droplet-based approach, Abseq, for characterizing proteins of >10[thin space (1/6-em)]000 cells at single-cell resolution (Fig. 3a).44 The method involves two types of barcodes, an antibody barcode that is a short oligonucleotide sequence conjugated to the antibody via a bifunctional crosslinker, and a cell barcode, which is subsequently attached to the antibody barcode through a splicing-by-overlap PCR. This leads to a chimeric DNA sequence in which one portion encodes the antibody identity and the other encodes the cell identity, which can be read by the sequencer. The sensitivity of this strategy is significantly better than that of fluorescence or mass cytometry, as the barcode sequences can be amplified from very low levels. Moreover, due to the use of unique molecular identifiers (UMI), the method is capable of generating quantitative results. However, the workflow involves multiple steps of droplet manipulations using custom microfluidic devices and is thus less friendly to non-expert users. Leveraging a similar barcoding strategy, Stoeckius and coworkers developed a simpler yet more powerful approach for the simultaneous measurement of the epitope and transcriptome at the single-cell level, CITE-seq.45 In this approach, the antibody barcode sequence is modified to have a 3′ polyA tail (Fig. 3a). Therefore, in a standard single-cell RNA-seq workflow (Drop-seq), barcode oligos of antibodies that bind to the cell and cellular mRNAs both anneal to the Drop-seq beads, which contain oligo-dT. After sequencing, the barcode groups and mRNAs can be separated based on the read length. Using a set of 13 antibodies that target surface protein markers for immune cell classification, the authors performed the CITE-seq workflow on 8005 cells from a complex immune cell population and achieved a characterization of immune cell phenotypes superior to that obtained with single-cell RNA-seq alone. Peterson and coworkers reported a similar approach, REAP-seq, in which they quantified 82 barcoded antibodies and >20[thin space (1/6-em)]000 genes of immune cells.46 Compared with Ab-seq, CITE-seq and REAP-seq provide multimodal data and are compatible with existing single-cell sequencing approaches and commercially available systems. However, multiplexing of all these methods is limited by the availability of specific antibodies that can detect the protein of interest, like many other methods relying on antibody–antigen interactions. Given that the proteome of a single mammalian cell comprises >20[thin space (1/6-em)]000 proteins and >100[thin space (1/6-em)]000 epitopes, extending these approaches to the “omics” level may be too impractical.
image file: c8an01503a-f3.tif
Fig. 3 Single-cell proteomics. (a) Conceptual illustration of DNA barcoding strategies for protein profiling41 (top) and DNA-antibody barcodes used in the Abseq44 and CITE-seq45 workflow (bottom). (b) The workflow of single-cell, 42-plex immune effector function profiling with the SCBC platform.9 (c) The workflow of microfluidic single-cell isoelectric focusing.56 (d) Schematic of the nanoliter oil-air–droplet chip and its workflow for single-cell mass spectrometry.59

Fluorescence immunoassay

Although the spectral overlap of fluorophores attached to a single cell limits the potential of multiplexing for all fluorescence-based methods, fluorescence signals are arguably most straightforward to detect with minimal infrastructure requirements (e.g., with a cell phone or even the naked eye). Miniaturized microfluidic devices for fluorescence immunoassays of single cells are thus of great value for clinical and point-of-care situations. A representative device for single-cell fluorescence immunoassays is the single-cell barcode chip (SCBC) developed by Ma and coworkers.47 The device features >1000 microchambers, each of which is patterned with a spatial antibody barcode array for the proteomic analysis of single immune cells. The barcode array consists of parallel stripes of distinctive antibodies created using flow confinement and the DNA-encoded antibody library approach. Functionalized with the barcodes, the device is capable of characterizing the functional heterogeneity in the expression of 12 distinct proteins of human immune cells. Although limited in spectral multiplexing, the approach is essentially limitless in spatial multiplexing, simply by increasing the number of patterned antibody stripes. Lu and coworkers recently demonstrated the use of a modified device for the detection of 42 immune effector proteins secreted from single differentiated macrophage cells (Fig. 3b).48 With appropriate modifications, the SCBC device is capable of profiling intracellular proteins. Shi and coworkers reported a revised SCBC design where an extra valve-controlled microchannel is coupled to the cell microchambers to deliver lysis buffer for on-chip cell lysis.49

Kinetic profiling of proteins at the single-cell level is desired to understand many critical biological processes, such as the development of stem cells. Pioneering work toward this goal is the microengraving approach developed by Love and coworkers.50,51 In this work, nano-liter well arrays containing single cells were capped with replaceable glass slides engraved (functionalized) with antibodies.20 After incubation, cell-secreted proteins of interest were captured by immobilized probes. The glass slide was subsequently removed from the nano-liter well substrate, and a secondary immunoassay was performed off-chip for fluorescence imaging. If desired, the cells of interest could be extracted for further investigations. Although the strategy captures only a small number of secreted proteins, it is capable of performing a kinetic analysis of >100[thin space (1/6-em)]000 single cells. In a following study, Han et al. carried out a 17-hour kinetic characterization of the expression profiles of 4 distinct cytokines secreted from >1000 T cells and observed an asynchronous expression pattern.52 Moreover, the microengraving device can be further modified to study cell–cell interactions.53 Similar to the methods based on the nano-liter well array platform, Schubert et al. developed a method for the absolute quantification of prostate-specific antigen (PSA) expressed in single prostate cancer cells.54 The method is an extension of the single molecule array (SiMoA) technique developed by Rissin and colleagues.55 SiMoA relies on antibody-coated beads to capture single protein molecules and a nano-liter well platform to isolate single beads for signal amplification and fluorescence detection.

Protein electrophoresis

Western blotting has been the bedrock for specific protein detection in biochemistry research. Recently, the Herr group brought western blots to the single-cell level with microfluidic techniques. The approach, scWesterns,2 relies on a nano-liter well array for single-cell isolation and lysis. The array is built on a photoactive polyacrylamide hydrogel layer so that cellular proteins can be electrophoretically separated in place and then photo-captured for immunoblotting. scWestern allows ∼1000 cells to be analyzed in a single experiment and features a demonstrated multiplexing capability of 11 protein targets per cell. Aside from its high-throughput and multiplexing, a key advantage of the approach is that it minimizes the impact of antibody cross-reactivity due to the employment of electrophoretic separation. In a later work, the resolving power of scWestern was further enhanced by using pore-gradient microgel arrays, enabling a 25–289 kDa protein mix to be resolved with 1 mm resolution. More recently, the Herr group extended the isoelectric focusing technique to the single-cell level (scIEF) with a 3D microfluidic device to investigate protein isoforms in individual mammalian cells (Fig. 3c).56 The device comprises a bottom polyacrylamide gel layer with patterned nano-liter wells for single-cell isolation and a lid layer to form a pH gradient. To reduce diffusive protein loss and achieve high sensitivity, the authors used a dense hydrogel lid layer so that the protein analytes are confined in the thin bottom layer during the focusing experiment. The protein capture efficiency was estimated to be ∼17.7%. Isoforms of both endogenous cytoskeletal and nuclear proteins can be discriminated in a single experiment by using different fluorescent dyes. Further improvements in the throughput and multiplexing capability are required to extend the utility of this technique.

Mass spectrometry

MS-based label-free proteomics is a dominant method for large-scale protein analysis and a promising candidate to meet the ultimate demand of “omics”-level protein profiling of single cells. For years, the route of extending MS to the single-cell level was hindered by its detection limit. A single mammalian cell contains thousands of different proteins, with a total mass of <1 ng, which is ∼100 times less than that used in routine MS analysis. Recent advances in MS techniques have lowered the detection limit to the range of sub-attomole to zeptomole, which is satisfactory for analyzing individual cells.14–16 Budnik and colleagues successfully quantified >700 protein species from single Jurkat and U-937 cells using an approach called SCoPE-MS.57 However, the approach relies on test tubes for handling single-cell samples and thus suffers from a low throughput. Microfluidics enables the handling of large numbers of single-cell samples with ease and enhanced sensitivity by reducing the sample volume and processing loss. Kuster et al. developed a microfluidic approach that couples microdroplets and matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS), in which 2600 droplets are printed in a nano-liter well array patterned onto a standard stainless steel MALDI-MS plate for multiplexed MS analysis.58 With slight modifications (e.g., involving the cell encapsulation step), the approach is capable of performing single-cell MALDI-MS in a high-throughput manner. Li and coworkers reported a clever microfluidic design for single-cell liquid chromatography mass spectrometry (Fig. 3d).59 The design features an on-chip oil–air–droplet sandwich structure that serves as a microreactor for multistep sample preparation with minimal processing losses. The design also includes a self-aligned monolithic device for interfacing the microfluidic chip with a LC column. Using the system, the authors detected 51 and 335 protein species from a single HeLa cell and a mouse oocyte, respectively. The method can be further scaled by using an array format and microfluidics automation. The performance of liquid chromatography can be further enhanced with the integration of nanofluidic techniques, by the pioneer research teams including Kitamori's group.60 For instance, Ishibashi and coworkers demonstrated highly efficient chromatographic separation in extended nanochannels.61 Compared to conventional high-performance liquid chromatography, the technique features a noticeably faster separation (by 2 orders of magnitude), a much smaller injection volume (by 9 orders of magnitude), and a higher separation efficiency (by 1 order of magnitude).

Single-cell metabolomics

Single-cell metabolomics is not particularly new; in the late 1990s, Fung et al. profiled a handful of metabolites from single rat peritoneal mast cells using laser vaporization MS.62 In the following decade, a few more studies on single-cell metabolomics were reported that relied on a variety of analytical techniques. However, these methods enabled the detection of only a minor subset of the metabolome present at high concentrations due to sensitivity limitations and, importantly, analyte loss during the preparation steps. Moreover, the methods suffered from poor throughput due to the lack of advanced handling techniques. In recent years, the field has witnessed rapid advancement with the assistance of microfluidic techniques. Here, we highlight three of the common categories of microfluidic approaches for single-cell metabolomics, including fluorescence-mediated analysis, electrochemical assays and mass spectrometry, with a focus on new and evolving methods.

Fluorescence-mediated analysis

The most popular fluorescence-based method for profiling individual cells is perhaps flow cytometry. Although powerful, flow cytometry is an open-fluid system that cannot be used to study the cell secretome. Droplet microfluidics offers an elegant solution to this limitation by confining single cells in tiny-volume aqueous droplets so that the secreted analyte is present at a high concentration. Recently, using a microdroplet approach, Shen and coworkers demonstrated sensitive fluorescence detection of hydrogen peroxide (H2O2) secreted by single cells (Fig. 4a).63 H2O2 is a key representative of reactive oxygen species (ROS), which is involved in a variety of cellular processes, including cell adhesion and DNA breaking. To detect H2O2 at the single-cell level, the authors encapsulated different types of cells in droplets together with horseradish peroxidase-templated gold nano-clusters (HRP-AuNCs). HRP-AuNCs enable the quantitative and sensitive detection of H2O2 molecules by catalysis-induced fluorescence quenching. The approach was used to assay four different cell lines, HUVEC, MCF7, U937 and Mut6 cells, with a demonstrated detection limit of 200–400 attomoles. Droplet-based platforms also enable the detection and screening of single cells based on their secretomes. A representative example is the microfluidic double emulsion-based FACS (MDE-FACS) approach developed by Terekhov and colleagues (Fig. 4b).64 The approach combines flowing droplet modules using FACS, followed by NGS or LC-MS for a complete round of ultrahigh-throughput (>108) cell screening and genotype/phenotype analysis. In this approach, single cells are compartmentalized in emulsions along with specific fluorescence-generating machinery for probing of a targeted function. Specific phenotypes can activate the machinery in droplets, which results in the presence of a fluorescence signal. The active phenotypes are subsequently sorted using FACS for downstream sequencing or LC-MS. In another example, Del Ben and coworkers developed a droplet-based method for detecting single circulating tumor cells (CTCs) based on the characterization of single-cell metabolism.65 The detection of CTCs in droplets relies on the Warburg effect (cancer cells secrete large amounts of lactate during their glycolysis-based metabolism, which leads to acidification of the local environment), and the in-droplet pH is sensed using a ratiometric dye. The approach allows the detection of as few as 10 tumor cells among 200[thin space (1/6-em)]000 white blood cells and presents a proof-of-concept indication that CTCs can be detected based on cancer cell metabolism.
image file: c8an01503a-f4.tif
Fig. 4 Single-cell metabolomics. (a, b) Schematic illustration of (a) detecting single-cell-secreted hydrogen peroxide in droplets with Au nanoclusters63 and (b) single-cell metabolism measurement in droplets.64 (c) Workflow of the integrated droplet-microextraction-ESI-MS approach (left) and a set of metabolites detected from single MCF-7 cells (right).78 (d) Graphical summary of the single-cell MALDI-MS approach using microarrays (left) and metabolite time study at single-cell and population levels (right).82

In addition to droplet-based platforms, fluorescence-mediated assays have been coupled with other common microfluidic formats for single-cell metabolomic analysis. Xue et al. reported a modified SCBC design for the quantification of metabolites from single cells.66 Instead of specific antibodies as in the proteomic assays, the barcode chip was functionalized with streptavidin to capture glucose–biotin conjugates released by single cells. Stratz and coworkers demonstrated a valve-based device for capturing and cultivating single Saccharomyces cerevisiae cells and for monitoring NAD(P)H concentrations.67 Cole and colleagues used the hybrid platform PDM (as previously discussed) to monitor intracellular calcium levels. However, as stated in the previous section, for each of these fluorescence-based methods, the multiplexing capability is a limiting issue due to the need for labels. They are particularly suitable for studying a group of metabolites that are unique to particular metabolic pathways.

Electrochemical assays

High-sensitivity and label-free electrochemical assays combined with microfluidic devices are ideal for the analysis of the extracellular environment and, thus, the cell secretome. Cheng et al. developed a microchip featuring an array of 15 microchambers with multiple integrated microelectrodes for electrostimulation and electrochemical measurements of single human cardiac muscle cells (cardiomyocytes).68 The device enables the multiplexed electrochemical profiling of the lactate, intracellular pH and calcium levels. Weltin and coworkers presented a multiparametric microfluidic system for the dynamic monitoring of human cancer cell metabolism, which allows the detection of pH, oxygen, glucose and lactose, with integrated electrochemical sensors.69 Respiration activity is a crucial indicator of cellular metabolism. Recent advances in scanning electrochemical microscopy (SECM) have enabled the microelectrochemical visualization of oxygen consumption in individual living cells;70 however, the method suffers from low throughput and is thus not suitable for analyzing large numbers of cells. Koide et al. developed a scalable platform for monitoring the respiration kinetics in single cells.71 The platform features a nano-liter well array for accommodating single cells and integrated electrodes in each well for electrochemical oxygen detection. An electrochemical sensor has also been coupled with separation methods (e.g., capillary electrophoresis) to detect a larger set of metabolites.8,72,73 Although electrochemical devices are only suitable for detecting electroactive species, and are thus limited in their multiplexing capability, they can be fabricated with low-cost materials (e.g., paper)74 and do not require expensive instruments for signal detection. Therefore, electrochemical devices are particularly favored in point-of-care situations.

Mass spectrometry

Benefiting from increased sensitivity and microfluidic techniques, MS has become a key enabling tool for single-cell metabolomics. A major contribution of microfluidics is that it interfaces single cells and MS in an efficient and high-throughput manner. Cells can be isolated in microfluidic devices and lysed in place. If desired, the cells can be cultivated in vitro to monitor the cell secretion behavior. The lysate or secreted analyte can be purified with an integrated micro-column and sent to the electrospray ionization (ESI) module directly, which eliminates sample losses during the processing steps. Wei and coworkers demonstrated the use of such a device to dynamically profile glutamate secretion from cultured neuronal cells.75 Gao et al. integrated solid-phase extraction (SPE) columns on-chip for sample preconcentration and monitored vitamin E and other related metabolites of human lung epithelial cells.76 Chen and coworkers investigated the metabolism of breast cancer cells (MCF-7) using a chip-ESI-MS system.77 However, for each device presented above, there are a few challenges to overcome before reaching single-cell resolution. For example, the cell culture medium contains various components, such as nucleotides, lipids and proteins, which are negligible at the population level but can be disastrous when dealing with single cells. Zhang and coworkers developed a droplet-based microextraction approach for the removal of such matrix interference (Fig. 4c).78 In the approach, single cells are wrapped and the cellular compounds are extracted in 2 nL droplets preloaded on the tip of a glass micropipette. After evaporation and redissolving in picoliters of an assisting solvent, the tip was directly applied to the MS instrument for ESI-MS. In this manner, the authors successfully detected tens of different metabolites from single MCF-7 cells.

It is known that the cellular metabolome reacts to environmental stimulations quite immediately. Thus, in situ (and even non-destructive) measurements are favored, as they cause minimal perturbation to the cells and the environment. Pan et al. developed an interesting design, the Single-probe, for real-time and in situ metabolomic analysis of individual living cells.79 The Single-probe consists of a dual-bore quartz micropipette for in situ extraction of cell constituents and a nano-ESI emitter for ionization. This microfluidic assembly allows the identification of >10 distinct metabolites (e.g., ATP) from single HeLa cells. Another approach for in situ metabolomic analysis is single-cell MS imaging. It is impressive that the spatial resolution of the state-of-the-art MS imaging methods (e.g., nano-SIMS)80 has reached well below 100 nm. Unless used to analyze secreted metabolites, MS is destructive to the cells being analyzed. With respect to non-destructive measurements, Kalfe and coworkers developed a planar microfluidic waveguide device for in vitro NMR metabolomics of single tumor spheroids.81 It would be exciting to extend microfluidic NMR to the single-cell level and to develop more non-destructive strategies for single-cell metabolomics.

Among many variations of mass spectrometry, MALDI-MS is a favored method for metabolomic analysis due to its high sensitivity. The planar configuration of the MALDI target plate makes it readily compatible with nano/microwell-based platforms for single-cell measurement at a high throughput. Ibanez et al. developed microarrays for a mass spectrometry (MAMS) platform that allows thousands of individual cells to be analyzed in a single MS experiment (Fig. 4d).82 The MAMS chip features arrays of hydrophilic wells surrounded by an omniphobic material to enable automated isolation of single cells. Using the platform, the authors monitored the metabolism of single yeast cells and observed intrinsic variabilities among yeast cells that are not accessible by population-based metabolomics. Xie et al. reported a similar platform, the microdot array chip, to monitor phospholipids in single human epithelial cells (A549).83 Yang et al. revised the fabrication process of the microdot array by using the contact printing technique.84 Finally, it is notable that the coverage of these single-cell MS platforms must be considerably increased before they can become truly useful, which will require further advancement in MS sensitivity and perhaps the combination of multiple analytical tools.

Extensions and conclusion

The remarkable advancements of single-cell proteomic and metabolomic tools have been beneficial for the study of various cell types of great value. The investigation of circulating tumor cells, for example, has witnessed substantial development in the past decade. CTCs are present at low concentrations in the peripheral blood of patients with solid tumors. It is believed that the isolation, ex vivo culture and characterization of CTCs may provide a non-invasive approach for monitoring the changing patterns of drug susceptibility. More importantly, the proteomic and metabolomic analyses of CTCs hold great potential to reveal the biology of cancer metastasis and identify signaling pathways for therapeutic interventions. Due to the extremely low concentrations of CTCs in blood (1 CTC per 106–107 leukocytes), the efficient enrichment and capture of CTCs are vital in all CTC analysis pipelines. Microfluidics has shown its usefulness in CTC capture with various innovative devices utilizing different principles, including mechanical85,86 and electrical87 trapping and antigen–antibody interaction. Notably, Wang and coworkers reported a microfluidic silicon nanowire array integrated with upconversion nanoprobes for highly efficient CTC isolation and detection.88 In their work, the nanoparticles were conjugated with anti-EpCAM antibodies to specifically detect tumor cells in blood, which can be subsequently pulled down with an external magnetic field. The use of a silicon nanowire array further increases the capture efficiency by 50%, leading to an overall capture efficiency of ∼90%. Other than their use in isolation and capture, microfluidic tools have provided new biological insights into CTC proteomics and metabolomics. Deng et al. developed an integrated microfluidic chip for the secretion profiling of single CTCs.89 The device leverages the SCBC technology (previously discussed) to profile a panel of secreted markers, such as IL-8 and VEGF, and select phenotypic subsets of the CTC population. The analysis shows that ∼76% of single cells are able to secrete detectable VEGF, characteristic of viable tumor cells. Yang and colleagues developed a microfluidic platform to profile multiple proteins of single CTCs.84 The device is capable of analyzing 15[thin space (1/6-em)]000 individual cells and integrates a beads-on-barcode antibody microarray to profile a panel of 12 different proteins. Using the device, it is estimated that A549, H1650, and H1975 cells have an average of ∼7700, ∼98[thin space (1/6-em)]000, and 290[thin space (1/6-em)]000 EpCAM molecules, respectively. Sinkala and coworkers profiled protein expressions in CTCs with microfluidic single-cell western blotting and discovered two distinct GAPDH subpopulations within the patient-derived CTCs.90 In addition to their applications for CTCs, microfluidic techniques have been accelerating the pace of stem cell research. Using the SCoPE-MS approach, Budnik et al. quantified the proteome heterogeneity of stem cells during the differentiation process and observed distinct regulatory patterns that were in agreement with the mRNA profiles.91 Kobel et al. tracked single stem cell division in a microfluidic trap device.92 Lecault and coworkers performed a high-throughput analysis of single hematopoietic stem cell proliferation using a microfluidic cell culture array device.93

In conclusion, we have provided a concise overview of how microfluidics has been helpful in the development of proteomic and metabolomic tools for single cells. With the emergence of single-cell proteomics and metabolomics, there is still a great need for methodological improvements, in multiplexing capacity, detection sensitivity, assay throughput and cost-effectiveness. All of these would facilitate widespread applications and approaching the ultimate goal of “omics” level analysis. Moreover, with increased volumes and dimensions of datasets gained with advanced analytical tools such as single-cell mass spectrometry, it is necessary to develop appropriate tools for “big data” analysis. In the next decade, this scientific community shall expect an increasing number of methodological breakthroughs in the field of single-cell proteomics and metabolomics.

Conflicts of interest

There are no conflicts to declare.


The research received funding from the National Natural Science Foundation of China (21575095 and 21874096), the Major State Basic Research Development Program (2013CB932702); a project supported by Collaborative Innovation Center of Suzhou Nano Science and Technology, the 111 project, and the Priority Academic Program Development of Jiangsu Higher Education Institutions. J. L. is supported by the “1000 Youth Talents” plan of the Global Expert Recruitment Program.


  1. E. Z. Macosko, A. Basu, R. Satija, J. Nemesh, K. Shekhar, M. Goldman, I. Tirosh, A. R. Bialas, N. Kamitaki, E. M. Martersteck, J. J. Trombetta, D. A. Weitz, J. R. Sanes, A. K. Shalek, A. Regev and S. A. McCarroll, Cell, 2015, 161, 1202–1214 CrossRef CAS PubMed .
  2. A. J. Hughes, D. P. Spelke, Z. Xu, C. C. Kang, D. V. Schaffer and A. E. Herr, Nat. Methods, 2014, 11, 749–755 CrossRef CAS PubMed .
  3. A. M. Klein, L. Mazutis, I. Akartuna, N. Tallapragada, A. Veres, V. Li, L. Peshkin, D. A. Weitz and M. W. Kirschner, Cell, 2015, 161, 1187–1201 CrossRef CAS PubMed .
  4. J. R. Heath, A. Ribas and P. S. Mischel, Nat. Rev. Drug Discovery, 2016, 15, 204–216 CrossRef CAS PubMed .
  5. R. Zenobi, Science, 2013, 342, 1243259 CrossRef CAS PubMed .
  6. S. M. Prakadan, A. K. Shalek and D. A. Weitz, Nat. Rev. Genet., 2017, 18, 345–361 CrossRef CAS PubMed .
  7. S. S. Rubakhin, E. V. Romanova, P. Nemes and J. V. Sweedler, Nat. Methods, 2011, 8, S20–S29 CrossRef CAS PubMed .
  8. A. Amantonico, P. L. Urban and R. Zenobi, Anal. Bioanal. Chem., 2010, 398, 2493–2504 CrossRef CAS PubMed .
  9. Y. Lu, L. Yang, W. Wei and Q. Shi, Lab Chip, 2017, 17, 1250–1263 RSC .
  10. A. M. Streets, X. Zhang, C. Cao, Y. Pang, X. Wu, L. Xiong, L. Yang, Y. Fu, L. Zhao, F. Tang and Y. Huang, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 7048–7053 CrossRef CAS PubMed .
  11. X. Han, R. Wang, Y. Zhou, L. Fei, H. Sun, S. Lai, A. Saadatpour, Z. Zhou, H. Chen, F. Ye, D. Huang, Y. Xu, W. Huang, M. Jiang, X. Jiang, J. Mao, Y. Chen, C. Lu, J. Xie, Q. Fang, Y. Wang, R. Yue, T. Li, H. Huang, S. H. Orkin, G. C. Yuan, M. Chen and G. Guo, Cell, 2018, 173, 1307 CrossRef CAS PubMed .
  12. S. C. Bendall, E. F. Simonds, P. Qiu, A. D. Amir el, P. O. Krutzik, R. Finck, R. V. Bruggner, R. Melamed, A. Trejo, O. I. Ornatsky, R. S. Balderas, S. K. Plevritis, K. Sachs, D. Pe'er, S. D. Tanner and G. P. Nolan, Science, 2011, 332, 687–696 CrossRef CAS PubMed .
  13. M. Wilhelm, J. Schlegl, H. Hahne, A. M. Gholami, M. Lieberenz, M. M. Savitski, E. Ziegler, L. Butzmann, S. Gessulat, H. Marx, T. Mathieson, S. Lemeer, K. Schnatbaum, U. Reimer, H. Wenschuh, M. Mollenhauer, J. Slotta-Huspenina, J. H. Boese, M. Bantscheff, A. Gerstmair, F. Faerber and B. Kuster, Nature, 2014, 509, 582–587 CrossRef CAS PubMed .
  14. M. E. Belov, M. V. Gorshkov, H. R. Udseth, G. A. Anderson and R. D. Smith, Anal. Chem., 2000, 72, 2271–2279 CrossRef CAS PubMed .
  15. B. N. Walker, C. Antonakos, S. T. Retterer and A. Vertes, Angew. Chem., Int. Ed., 2013, 52, 3650–3653 CrossRef CAS PubMed .
  16. A. Oikawa and K. Saito, Plant J., 2012, 70, 30–38 CrossRef CAS PubMed .
  17. T. W. Murphy, Q. Zhang, L. B. Naler, S. Ma and C. Lu, Analyst, 2017, 143, 60–80 RSC .
  18. B. B. Lake, S. Chen, B. C. Sos, J. Fan, G. E. Kaeser, Y. C. Yung, T. E. Duong, D. Gao, J. Chun, P. V. Kharchenko and K. Zhang, Nat. Biotechnol., 2018, 36, 70–80 CrossRef CAS PubMed .
  19. N. Habib, I. Avraham-Davidi, A. Basu, T. Burks, K. Shekhar, M. Hofree, S. R. Choudhury, F. Aguet, E. Gelfand, K. Ardlie, D. A. Weitz, O. Rozenblatt-Rosen, F. Zhang and A. Regev, Nat. Methods, 2017, 14, 955–958 CrossRef CAS PubMed .
  20. A. J. Torres, A. S. Hill and J. C. Love, Anal. Chem., 2014, 86, 11562–11569 CrossRef CAS PubMed .
  21. Y. Liu and L. Yobas, Appl. Phys. Lett., 2015, 106, 174101 CrossRef .
  22. M. A. Unger, H. P. Chou, T. Thorsen, A. Scherer and S. R. Quake, Science, 2000, 288, 113–116 CrossRef CAS PubMed .
  23. J. Shemesh, T. Ben Arye, J. Avesar, J. H. Kang, A. Fine, M. Super, A. Meller, D. E. Ingber and S. Levenberg, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 11293–11298 CrossRef CAS PubMed .
  24. G. Amselem, C. Guermonprez, B. Drogue, S. Michelin and C. N. Baroud, Lab Chip, 2016, 16, 4200–4211 RSC .
  25. L. Kang, M. J. Hancock, M. D. Brigham and A. Khademhosseini, J. Biomed. Mater. Res., Part A, 2010, 93, 547–557 Search PubMed .
  26. E. Um, E. Rha, S. L. Choi, S. G. Lee and J. K. Park, Lab Chip, 2012, 12, 1594–1597 RSC .
  27. K. Leung, A. Klaus, B. K. Lin, E. Laks, J. Biele, D. Lai, A. Bashashati, Y. F. Huang, R. Aniba, M. Moksa, A. Steif, A. M. Mes-Masson, M. Hirst, S. P. Shah, S. Aparicio and C. L. Hansen, Proc. Natl. Acad. Sci. U. S. A., 2016, 113, 8484–8489 CrossRef CAS PubMed .
  28. Y. Zhu, L. N. Zhu, R. Guo, H. J. Cui, S. Ye and Q. Fang, Sci. Rep., 2014, 4, 5046 CrossRef CAS PubMed .
  29. S. S. Bithi and S. A. Vanapalli, Sci. Rep., 2017, 7, 41707 CrossRef CAS PubMed .
  30. D. Di Carlo, Lab Chip, 2009, 9, 3038–3046 RSC .
  31. R. H. Cole, S. Y. Tang, C. A. Siltanen, P. Shahi, J. Q. Zhang, S. Poust, Z. J. Gartner and A. R. Abate, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 8728–8733 CrossRef CAS PubMed .
  32. K. Leung, H. Zahn, T. Leaver, K. M. Konwar, N. W. Hanson, A. P. Page, C. C. Lo, P. S. Chain, S. J. Hallam and C. L. Hansen, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 7665–7670 CrossRef CAS PubMed .
  33. Y. Yang, H. S. Rho, M. Stevens, A. G. Tibbe, H. Gardeniers and L. W. Terstappen, Lab Chip, 2015, 15, 4331–4337 RSC .
  34. G. J. Shah, A. T. Ohta, E. P. Chiou, M. C. Wu and C. J. Kim, Lab Chip, 2009, 9, 1732–1739 RSC .
  35. A. Ali-Cherif, S. Begolo, S. Descroix, J. L. Viovy and L. Malaquin, Angew. Chem., Int. Ed., 2012, 51, 10765–10769 CrossRef CAS PubMed .
  36. D. J. Collins, B. Morahan, J. Garcia-Bustos, C. Doerig, M. Plebanski and A. Neild, Nat. Commun., 2015, 6, 8686 CrossRef CAS PubMed .
  37. T. P. Hunt, D. Issadore and R. M. Westervelt, Lab Chip, 2008, 8, 81–87 RSC .
  38. I. A. Eydelnant, U. Uddayasankar, B. Li, M. W. Liao and A. R. Wheeler, Lab Chip, 2012, 12, 750–757 RSC .
  39. A. Banerjee, E. Kreit, Y. Liu, J. Heikenfeld and I. Papautsky, Lab Chip, 2012, 12, 758–764 RSC .
  40. L. Xu, I. L. Brito, E. J. Alm and P. C. Blainey, Nat. Methods, 2016, 13, 759–762 CrossRef CAS PubMed .
  41. F. Lan, B. Demaree, N. Ahmed and A. R. Abate, Nat. Biotechnol., 2017, 35, 640–646 CrossRef CAS PubMed .
  42. K. J. Son, D. S. Shin, T. Kwa, J. You, Y. Gao and A. Revzin, Lab Chip, 2015, 15, 637–641 RSC .
  43. J. R. Moffitt, J. B. Lee and P. Cluzel, Lab Chip, 2012, 12, 1487–1494 RSC .
  44. P. Shahi, S. C. Kim, J. R. Haliburton, Z. J. Gartner and A. R. Abate, Sci. Rep., 2017, 7, 44447 CrossRef CAS PubMed .
  45. M. Stoeckius, C. Hafemeister, W. Stephenson, B. Houck-Loomis, P. K. Chattopadhyay, H. Swerdlow, R. Satija and P. Smibert, Nat. Methods, 2017, 14, 865–868 CrossRef CAS PubMed .
  46. V. M. Peterson, K. X. Zhang, N. Kumar, J. Wong, L. Li, D. C. Wilson, R. Moore, T. K. McClanahan, S. Sadekova and J. A. Klappenbach, Nat. Biotechnol., 2017, 35, 936–939 CrossRef CAS PubMed .
  47. C. Ma, R. Fan, H. Ahmad, Q. Shi, B. Comin-Anduix, T. Chodon, R. C. Koya, C. C. Liu, G. A. Kwong, C. G. Radu, A. Ribas and J. R. Heath, Nat. Med., 2011, 17, 738–743 CrossRef CAS PubMed .
  48. Y. Lu, Q. Xue, M. R. Eisele, E. S. Sulistijo, K. Brower, L. Han, A. D. Amir el, D. Pe'er, K. Miller-Jensen and R. Fan, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, E607–E615 CrossRef CAS PubMed .
  49. Q. Shi, L. Qin, W. Wei, F. Geng, R. Fan, Y. S. Shin, D. Guo, L. Hood, P. S. Mischel and J. R. Heath, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 419–424 CrossRef CAS PubMed .
  50. J. C. Love, J. L. Ronan, G. M. Grotenbreg, A. G. van der Veen and H. L. Ploegh, Nat. Biotechnol., 2006, 24, 703–707 CrossRef CAS PubMed .
  51. A. O. Ogunniyi, C. M. Story, E. Papa, E. Guillen and J. C. Love, Nat. Protoc., 2009, 4, 767–782 CrossRef CAS PubMed .
  52. Q. Han, N. Bagheri, E. M. Bradshaw, D. A. Hafler, D. A. Lauffenburger and J. C. Love, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 1607–1612 CrossRef CAS PubMed .
  53. Y. J. Yamanaka, C. T. Berger, M. Sips, P. C. Cheney, G. Alter and J. C. Love, Integr. Biol., 2012, 4, 1175–1184 RSC .
  54. S. M. Schubert, S. R. Walter, M. Manesse and D. R. Walt, Anal. Chem., 2016, 88, 2952–2957 CrossRef CAS PubMed .
  55. D. M. Rissin, C. W. Kan, T. G. Campbell, S. C. Howes, D. R. Fournier, L. Song, T. Piech, P. P. Patel, L. Chang, A. J. Rivnak, E. P. Ferrell, J. D. Randall, G. K. Provuncher, D. R. Walt and D. C. Duffy, Nat. Biotechnol., 2010, 28, 595–599 CrossRef CAS PubMed .
  56. A. M. Tentori, K. A. Yamauchi and A. E. Herr, Angew. Chem., Int. Ed., 2016, 55, 12431–12435 CrossRef CAS PubMed .
  57. B. Budnik, E. Levy, G. Harmange and N. Slavov, bioRxiv, 2018 DOI:10.1101/102681 .
  58. S. K. Kuster, S. R. Fagerer, P. E. Verboket, K. Eyer, K. Jefimovs, R. Zenobi and P. S. Dittrich, Anal. Chem., 2013, 85, 1285–1289 CrossRef CAS PubMed .
  59. Z. Y. Li, M. Huang, X. K. Wang, Y. Zhu, J. S. Li, C. C. L. Wong and Q. Fang, Anal. Chem., 2018, 90, 5430–5438 CrossRef CAS PubMed .
  60. R. Ishibashi, K. Mawatari and T. Kitamori, J. Chromatogr. A, 2012, 1238, 152–155 CrossRef CAS PubMed .
  61. R. Ishibashi, K. Mawatari and T. Kitamori, Small, 2012, 8, 1237–1242 CrossRef CAS PubMed .
  62. E. N. Fung and E. S. Yeung, Anal. Chem., 1998, 70, 3206–3212 CrossRef CAS PubMed .
  63. R. Shen, P. Liu, Y. Zhang, Z. Yu, X. Chen, L. Zhou, B. Nie, A. Zaczek, J. Chen and J. Liu, Anal. Chem., 2018, 90, 4478–4484 CrossRef CAS PubMed .
  64. S. S. Terekhov, I. V. Smirnov, A. V. Stepanova, T. V. Bobik, Y. A. Mokrushina, N. A. Ponomarenko, A. A. Belogurov Jr., M. P. Rubtsova, O. V. Kartseva, M. O. Gomzikova, A. A. Moskovtsev, A. S. Bukatin, M. V. Dubina, E. S. Kostryukova, V. V. Babenko, M. T. Vakhitova, A. I. Manolov, M. V. Malakhova, M. A. Kornienko, A. V. Tyakht, A. A. Vanyushkina, E. N. Ilina, P. Masson, A. G. Gabibov and S. Altman, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 2550–2555 CrossRef CAS PubMed .
  65. F. Del Ben, M. Turetta, G. Celetti, A. Piruska, M. Bulfoni, D. Cesselli, W. T. Huck and G. Scoles, Angew. Chem., Int. Ed., 2016, 55, 8581–8584 CrossRef CAS PubMed .
  66. M. Xue, W. Wei, Y. Su, J. Kim, Y. S. Shin, W. X. Mai, D. A. Nathanson and J. R. Heath, J. Am. Chem. Soc., 2015, 137, 4066–4069 CrossRef CAS PubMed .
  67. S. Stratz, P. E. Verboket, K. Hasler and P. S. Dittrich, Electrophoresis, 2018, 39, 540–547 CrossRef CAS PubMed .
  68. W. Cheng, N. Klauke, G. Smith and J. M. Cooper, Electrophoresis, 2010, 31, 1405–1413 CrossRef CAS PubMed .
  69. A. Weltin, K. Slotwinski, J. Kieninger, I. Moser, G. Jobst, M. Wego, R. Ehret and G. A. Urban, Lab Chip, 2014, 14, 138–146 RSC .
  70. C. H. Chen, L. Jacobse, K. McKelvey, S. C. Lai, M. T. Koper and P. R. Unwin, Anal. Chem., 2015, 87, 5782–5789 CrossRef CAS PubMed .
  71. M. Koide, T. Yasukawa, Y. Horiguchi, K. Nagamine, H. Shiku, T. Matsue and T. Itayama, Anal. Sci., 2012, 28, 69 CrossRef CAS PubMed .
  72. P. Dittrich and A. J. Ibanez, Electrophoresis, 2015, 36, 2196–2206 CrossRef CAS PubMed .
  73. A. E. Vasdekis and G. Stephanopoulos, Metab. Eng., 2015, 27, 115–135 CrossRef CAS PubMed .
  74. P. Shah, X. Zhu and C. Z. Li, Expert Rev. Mol. Diagn., 2013, 13, 83–91 CrossRef CAS PubMed .
  75. H. Wei, H. Li, D. Gao and J. M. Lin, Analyst, 2010, 135, 2043–2050 RSC .
  76. D. Gao, H. Wei, G. S. Guo and J. M. Lin, Anal. Chem., 2010, 82, 5679–5685 CrossRef CAS PubMed .
  77. Q. Chen, J. Wu, Y. Zhang and J. M. Lin, Anal. Chem., 2012, 84, 1695–1701 CrossRef CAS PubMed .
  78. X. C. Zhang, Z. W. Wei, X. Y. Gong, X. Y. Si, Y. Y. Zhao, C. D. Yang, S. C. Zhang and X. R. Zhang, Sci. Rep., 2016, 6, 24730 CrossRef PubMed .
  79. N. Pan, W. Rao, N. R. Kothapalli, R. Liu, A. W. Burgett and Z. Yang, Anal. Chem., 2014, 86, 9376–9380 CrossRef CAS PubMed .
  80. D. M. Doughty, M. Dieterle, A. L. Sessions, W. W. Fischer and D. K. Newman, PLoS One, 2014, 9, e84455 CrossRef PubMed .
  81. A. Kalfe, A. Telfah, J. Lambert and R. Hergenroder, Anal. Chem., 2015, 87, 7402–7410 CrossRef CAS PubMed .
  82. A. J. Ibanez, S. R. Fagerer, A. M. Schmidt, P. L. Urban, K. Jefimovs, P. Geiger, R. Dechant, M. Heinemann and R. Zenobi, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 8790–8794 CrossRef CAS PubMed .
  83. W. Xie, D. Gao, F. Jin, Y. Jiang and H. Liu, Anal. Chem., 2015, 87, 7052–7059 CrossRef CAS PubMed .
  84. L. Yang, Z. Wang, Y. Deng, Y. Li, W. Wei and Q. Shi, Anal. Chem., 2016, 88, 11077–11083 CrossRef CAS PubMed .
  85. T. Yeo, S. J. Tan, C. L. Lim, D. P. Lau, Y. W. Chua, S. S. Krisna, G. Iyer, G. S. Tan, T. K. Lim, D. S. Tan, W. T. Lim and C. T. Lim, Sci. Rep., 2016, 6, 22076 CrossRef CAS PubMed .
  86. A. F. Sarioglu, N. Aceto, N. Kojic, M. C. Donaldson, M. Zeinali, B. Hamza, A. Engstrom, H. Zhu, T. K. Sundaresan, D. T. Miyamoto, X. Luo, A. Bardia, B. S. Wittner, S. Ramaswamy, T. Shioda, D. T. Ting, S. L. Stott, R. Kapur, S. Maheswaran, D. A. Haber and M. Toner, Nat. Methods, 2015, 12, 685–691 CrossRef CAS PubMed .
  87. X. Xing and L. Yobas, Analyst, 2015, 140, 3397–3405 RSC .
  88. C. Wang, M. Ye, L. Cheng, R. Li, W. Zhu, Z. Shi, C. Fan, J. He, J. Liu and Z. Liu, Biomaterials, 2015, 54, 55–62 CrossRef CAS PubMed .
  89. Y. Deng, Y. Zhang, S. Sun, Z. Wang, M. Wang, B. Yu, D. M. Czajkowsky, B. Liu, Y. Li, W. Wei and Q. Shi, Sci. Rep., 2014, 4, 7499 CrossRef CAS PubMed .
  90. E. Sinkala, E. Sollier-Christen, C. Renier, E. Rosas-Canyelles, J. Che, K. Heirich, T. A. Duncombe, J. Vlassakis, K. A. Yamauchi, H. Huang, S. S. Jeffrey and A. E. Herr, Nat. Commun., 2017, 8, 14622 CrossRef CAS PubMed .
  91. B. Budnik, E. Levy and N. Slavov, bioRxiv, 2017 DOI:10.1101/102681 .
  92. S. A. Kobel, O. Burri, A. Griffa, M. Girotra, A. Seitz and M. P. Lutolf, Lab Chip, 2012, 12, 2843–2849 RSC .
  93. V. Lecault, M. Vaninsberghe, S. Sekulovic, D. J. Knapp, S. Wohrer, W. Bowden, F. Viel, T. McLaughlin, A. Jarandehei, M. Miller, D. Falconnet, A. K. White, D. G. Kent, M. R. Copley, F. Taghipour, C. J. Eaves, R. K. Humphries, J. M. Piret and C. L. Hansen, Nat. Methods, 2011, 8, 581–586 CrossRef CAS PubMed .
  94. S. Khetan, M. Guvendiren, W. R. Legant, D. M. Cohen, C. S. Chen and J. A. Burdick, Nat. Mater., 2013, 12, 458–465 CrossRef CAS PubMed .

This journal is © The Royal Society of Chemistry 2019