Isolating and accurately identifying target cells is a crucial prerequisite for single-cell analysis. The efficacy of single-cell isolation technologies is predominantly evaluated based on three critical parameters: efficiency, defined as the capacity to isolate a specific number of target cells within a given time frame; separation purity, referring to the proportion of target cells isolated relative to non-target cells post-separation; and recovery rate, indicating the quantity of target cells successfully retrieved post-separation compared to the initial count present in the sample (Hu et al., 2016; Gross et al., 2015). The selection of an appropriate isolation method is essential for reliable single-cell analyses, particularly when studying rare cell populations or investigating cellular heterogeneity, as distinct techniques exhibit varying effectiveness across these parameters.

Contemporary cell separation technologies can be broadly classified into two main categories based on their isolation properties. The first category encompasses techniques that rely on the physical attributes of cells, such as morphology, size, density, and deformability. This category includes methodologies like density gradient centrifugation, which separates cells based on their buoyant density; size-based filtration, which employs membranes with specific pore sizes to isolate cells of interest; manual cell picking, which involves the direct selection of individual cells using micropipettes; and a part of microfluidics-based capture, which utilizes the unique flow properties of cells in microchannels for isolation (Gross et al., 2015; Malter, 2016; Liu and Singh, 2013; Huang et al., 2008; Cha et al., 2022). A key advantage of these techniques is the ability to isolate cells in a label-free manner, without the need for internal or external markers, minimizing potential interference with cellular processes and preserving the native state of the cells (Gross et al., 2015).

The second category relies on the biological properties of cells, employing affinity-based methods. These methods, including fluorescence-activated cell sorting (FACS), magnetic-activated cell sorting (MACS), and laser capture microdissection (LCM), utilize the presence of specific surface proteins and their affinity for corresponding antibodies or probes (Potashnikova et al., 2018; Miltenyi et al., 1990; Vandewoestyne and Deforce, 2010). FACS utilizes fluorescently labeled antibodies to mark cells of interest, enabling their sorting according to fluorescence intensity. MACS uses antibody-coated magnetic beads to separate target cells from a heterogeneous population by use of an external magnetic field. While, LCM employs a laser to selectively isolate cells from tissue sections. These techniques offer high specificity and purity, enabling the isolation of rare cell populations based on their unique biological markers and specific morphology (Gross et al., 2015).

In this section, we briefly summarize the foundational principles, inherent advantages, limitations, and potential applications of the most prevalent cell separation methodologies (Table 1). Each technique’s unique attributes contribute to its suitability for specific research contexts, empathizing the importance of carefully evaluating the requirements of the study and accordingly selecting the most appropriate method. By understanding the strengths and weaknesses of each approach, researchers can optimize their experimental designs and ensure the reliable isolation of target cells for downstream single-cell analysis.

Fluorescence-Activated Cell Sorting (FACS), an advanced flow cytometry technique with sorting capabilities, and it is widely recognized as the most sophisticated and accessible method for identifying, categorizing, and isolating specific cell types from heterogeneous cell populations (Sharon et al., 2013; An and Chen, 2018; Herzenberg et al., 2002). This versatile technique differentiates cells based on a multitude of parameters, including size, granularity, and fluorescence intensity patterns arising from labeling with specific fluorescent markers. FACS enables concurrent quantitative and qualitative multi-parametric analysis of individual live or fixed cells at impressive throughput rates, rendering it an essential tool for single-cell isolation and characterization.

The FACS process begins with the preparation of a single-cell suspension, wherein target cells are labeled with fluorescent markers. Fluorophore-conjugated monoclonal antibodies, which recognize specific cell surface markers, are the primary choice of fluorescent probes utilized in this technique (Pan and Wan, 2020). As cells traverse through the cytometer’s fluidic system, they are hydrodynamically focused into a single-file stream and individually exposed to one or more laser beams. The interaction of the laser light with the cells generates fluorescence signals, which are collected by dedicated detectors (Adan et al., 2017). The forward scatter (FSC) and side scatter (SSC) signals provide information about the cell’s relative size and granularity, respectively, while the fluorescence detectors identify and characterize cells based on their predefined fluorescence signatures (Adan et al., 2017).

The sorting mechanism of FACS relies on an electrostatic deflection system. After laser excitation, the cell stream is fragmented into individual droplets, with each droplet encapsulating a single cell. Droplets can be formed by using high-frequency (cycles/second, Hz) vibration of the nozzle at an optimal amplitude. Based on the detected fluorescence and scatter signals, the droplets containing the desired single cells are assigned an electrical charge—either positive or negative by an electrical charging ring placed just at the point where the stream breaks into droplets. Next, as the droplets pass through an electric field, the charged droplets are deflected into designated collection tubes or well plates for downstream analysis or culture, while the uncharged droplets are directed into a waste container (Figure 1A) (Rambault et al., 2021). State-of-the-art FACS systems, such as the FACS-Aria™ III, boast an impressive throughput rate of up to 100,000 droplets per second, enabling the analysis of approximately 70,000 events per second (Gross et al., 2015). Moreover, FACS has proven adept at sorting individual cells from complex mixtures containing thousands of cells, utilizing up to 18 distinct surface markers for precise identification and separation (Hu et al., 2016). Additionally, an advanced FACS system developed by Dr. Li’s group has demonstrated over 90% recovery efficiency and more than 80% cell viability (Dong et al., 2017).

One of the key advantages of FACS over other cell sorting techniques is its capability to analyze and sort cells simultaneously based on multiple parameters, encompassing both surface and intracellular markers (Bennett et al., 2016). This multi-parametric capability enables highly specific cell isolation and characterization, allowing researchers to identify and isolate rare cell populations with unique phenotypic profiles (Maes et al., 2020).

After single cell isolation, subsequent analysis is necessary. The high throughput and purity of FACS make it an invaluable tool in various applications, including immunophenotyping (An and Chen, 2018; Pan and Wan, 2020), cell cycle analysis (Potashnikova et al., 2018), subpopulation analysis (Antoniadi et al., 2022), cancer diagnosis (Sharon et al., 2013), and the isolation of rare cell populations for downstream single-cell analyses, such as single-cell RNA sequencing and single-cell proteomics (Zhu et al., 2019).

Despite its unparalleled versatility, FACS has certain limitations. The technique requires a relatively large initial cell number, typically in the range of 10 ^ 4 to 10 ^ 6 cells, which can hinder the isolation of individual cells from rare subpopulations (Gross et al., 2015; Shields et al., 2015). Moreover, the high-speed fluidic system and the potential for shear stress during sorting may compromise the viability and functional integrity of sorted cells, particularly for fragile cell types (Gross et al., 2015). To mitigate these issues, researchers have developed gentle sorting techniques, such as low-pressure sorting and the use of cell-friendly collection buffers, to improve cell viability and recovery (Teteris et al., 2012). Furthermore, recent advancements in microfluidics-based cell sorting technologies have enabled the isolation of single cells from rare populations with minimal sample requirements and reduced shear stress (Stavrakis et al., 2019).

Magnetic-Activated Cell Sorting (MACS) is another widely employed technique for the passive separation of cell populations, leveraging their cluster of specific extracellular markers to isolate specific cell types (Rambault et al., 2021; Shen et al., 2021). MACS has demonstrated the capability to achieve isolation of particular cell groups with a purity exceeding 90%, ideally even higher than 95% (Miltenyi et al., 1990; Desikan et al., 2022) and high recovery rate of more than 90% (Willasch et al., 2010). The principle underlying MACS involves the use of magnetic beads conjugated with targeting molecules such as antibodies, enzymes, lectins, or streptavidin. These conjugated beads are designed to selectively target surface molecules primarily on live cells, thereby enabling the efficient isolation of the desired cell population.

MACS offers two modes of separation: positive and negative. In positive separation, the heterogeneous cell mixture is incubated with magnetic beads coated with antibodies specific to the surface markers of the target cells (Pan and Wan, 2020). When exposed to an external magnetic field, the labeled cells become magnetically responsive and are retained within the field, while unlabeled cells can be easily removed through a washing process (Figure 1B). This approach is particularly useful when the target cell population is rare or when a high purity of the isolated cells is required (Nemescu et al., 2020). In contrast, negative separation is employed when targeting cell-specific substances are unavailable or when the aim is to deplete a specific cell population from the sample. In this mode, the unwanted cells are labeled with more than one antibody-conjugated magnetic beads and subsequently separated from the mixture, thereby isolating the unlabeled target cells in the solution (Pan and Wan, 2020).

Compared to FACS, which can separate cells based on the expression of multiple specific molecules simultaneously, MACS has a more limited capacity, sorting cells solely into positive and negative populations based on a single surface marker. Consequently, the purity level achievable with MACS is typically lower than that of FACS (Zeb et al., 2019). However, MACS offers several advantages, including its relative simplicity, cost-effectiveness, and the ability to process large sample volumes quickly. Additionally, MACS is generally gentler on the cells than FACS and results in lower cell loss, as it does not require high-pressure fluidic systems, making it more suitable for isolating fragile cell types (Sutermaster and Darling, 2019). To enhance the performance of MACS, an advanced method known as integrated Dielectrophoretic-Magnetic Activated Cell Sorter (iDMACS) was developed in 2009 (Kim and Soh, 2009). This innovative approach combines dielectrophoretic forces and magnetic particles. In iDMACS, cells are first labeled with magnetic beads and then subjected to a dielectrophoretic force, which helps to further separate the labeled and unlabeled cells based on their distinct dielectric properties. This additional separation step improves the purity of the isolated cell population (Kim and Soh, 2009). Recently, a novel purification method of human iPSC-derived cells at large scale by combining the miR-switch and MACS (miR-switch-MACS) was developed in 2022. This method achieved over 99% purity of chromogranin A-positive cells following puromycin selection to eliminate untransfected cells, surpassing the commonly used 95% purity threshold (Tsujisaka et al., 2022).

MACS enables efficient enrichment of target cell populations based on their unique surface markers, rendering it a valuable tool for a wide range of biological and medical research applications where high-purity cell samples are indispensable (Nemescu et al., 2020; Li et al., 2023). Its ability to process large sample volumes rapidly and its relative simplicity make it an attractive choice for many cell isolation tasks. Moreover, MACS can be used in conjunction with other cell separation techniques, such as FACS or density gradient centrifugation, to achieve even higher purity and specificity (Berteli et al., 2017). As the field of single-cell analysis advances, the innovation of new magnetic separation technologies and the fusion of MACS with microfluidic platforms show potential in applications of this robust cell isolation technique (Wang X. et al., 2019). Wang et al., combined Microfluidic chip with magnetic-activated cell sorting technology. Compared with EpCAM dependent traditional CTCs isolation system like CytoQuest™ CR system, their antigen-independent platform maintained a high detection sensitivity while achieving higher CTC counts consistent with clinical observations.

Laser Capture Microdissection (LCM) is an advanced technology for the precise isolation of pure cell populations or individual cells from solid tissue samples mounted on a microscope slide, as established by Emmert-Buck et al. (1996). This technique enables the targeted capture of cells of interest from heterogeneous tissue sections, allowing for subsequent molecular analyses such as PCR (Kandathil et al., 2013), microarrays (Wang et al., 2010), and proteomics (Dilillo et al., 2017).

The procedure for LCM begins with the visualization of target cells under a microscope. The tissue sample is typically prepared as a thin section and mounted on a special microscope slide. The operator then identifies the cells of interest and marks the section to be excised on the display by outlining it with a line. Subsequently, a laser follows this path to cut the tissue, isolating the desired cell or compartment (Figure 1C). The laser cutting procedure remains consistent across various LCM systems, but there are different methods for retrieving the dissected tissue, including contact-based extraction (Vandewoestyne et al., 2013), gravity-assisted microdissection (Gross et al., 2015), and laser pressure catapulting (Vogel et al., 2007). Contact-based extraction involves utilizing a thermoplastic film placed in contact with the tissue section. When activated by the laser, the film melts onto the targeted cells, enabling their separation from the surrounding tissue. In gravity-assisted microdissection, the laser cuts the tissue, and the dissected cells fall into a collection tube due to gravity. Conversely, laser pressure catapulting employs a focused laser pulse to propel the cut tissue into a collection device positioned above the sample. Typically, fixed cells—most commonly used in LCM—are subjected to a single downstream analysis. However, different downstream applications require distinct recovery protocols (such as variations in lysis solvents, extraction reagents, and treatment conditions) to obtain DNA, RNA, proteins, or metabolites (Guo et al., 2023).

Laser Capture Microdissection (LCM) offers several significant advantages. One of the most notable is its ability to maintain both precision and versatility when working with both fixed and live samples (Espina et al., 2007). LCM enables the accurate separation of even small numbers of cells, as well as single-cell isolation with high purity. With recent advancements in optical resolution, LCM can now also isolate cell organelles (Satori et al., 2012). This feature allows researchers to study cells in their native tissue environment, preserving the spatial relationships and potential interactions between different cell types. Additionally, LCM minimizes damage to adjacent cells following the initial dissection, ensuring that the collected cells are not contaminated by unwanted material. LSM, when combined with other techniques, has become more widely used in single-cell protein analysis. Gordon and Gousset demonstrated the integration of Laser Capture Microdissection (LCM) with mass spectrometry (MS) to identify low-abundance proteins in complex samples, achieving high protein extraction efficiency and excellent sample quality (Gordon and Gousset, 2021). However, LCM has some limitations, including a high operational barrier due to the need for specialized equipment and trained personnel, relatively low throughput compared to other cell isolation methods, and potential UV-induced damage to DNA or RNA during operation (Espina et al., 2007).

To overcome these limitations, recent advancements in LCM technology have focused on improving the speed, automation, and gentleness of the cell capture process. For example, the development of infrared (IR) LCM systems has enabled the use of longer wavelengths of light, reducing the risk of UV-induced damage to biomolecules (Vandewoestyne et al., 2013). Moreover, the integration of LCM with other technologies, such as microfluidics and single-cell sequencing, has expanded its applications and enhanced its throughput (Zhang et al., 2022).

Integration with immunohistological staining enhances LCM as a robust tool for analyzing solid samples at the single-cell level. By employing specific antibodies to mark cells of interest, researchers can visualize and isolate these specific cell types based on their protein expression profiles. This capability renders LCM an invaluable tool for diverse research efforts that necessitate the isolation of specific cells for subsequent molecular analysis, such as studies on tumor heterogeneity, neurodegenerative disorders, and developmental biology (Vandewoestyne and Deforce, 2010; Fink et al., 2006; Decarlo et al., 2011).

Manual cell picking, also known as micromanipulation, represents a straightforward, convenient, and effective approach for isolating single cells. Micromanipulation systems typically involve the use of an inverted microscope in conjunction with movable ultrathin glass capillaries controlled by motorized micromanipulation stages (Figure 1D). The cell sample is usually suspended in a dish or well-plate, allowing for direct visualization and precise isolation of individual live or fixed cells. The operator can observe and photograph the cells under the microscope, facilitating isolation based on morphological characteristics or fluorescent labeling.

A primary advantage of manual cell picking is its capability to isolate cells in a live state, preserving their viability and functionality for subsequent applications like single-cell sequencing, clonal expansion, or functional assays (Hu et al., 2016). This feature sets it apart from techniques like LCM, which primarily isolate single cells from sections of fixed tissue. Manual cell picking enables researchers to select and extract specific cells of interest based on their morphology, behavior, or response to stimuli, providing a powerful tool for studying cellular heterogeneity and elucidating the roles of individual cells within a population (Malter, 2016; Fröhlich and König, 2000; Li et al., 2011).

Microfluidics is a rapidly growing field of science and engineering that focuses on the manipulation and control of fluids at the microscale level. This technology enables the development of highly miniaturized platforms, often referred to as “lab-on-a-chip” devices, capable of handling biological samples in extremely small volumes, typically in the nanoliter range. The miniaturization offered by microfluidics provides several advantages in the study of biological systems, particularly in the realm of single-cell analysis and manipulation (Grigorev et al., 2023; Sims and Allbritton, 2007; Lecault et al., 2012; Liu D. et al., 2022).

Microfluidic systems offer a significant advantage in generating localized high protein concentrations. By confining cells and their secreted proteins within small volumes, requirement for amplification techniques to detect low-abundance proteins is not that necessary with microfluidic devices, enabling more precise quantification of biomolecules derived from single cells (Liu and Singh, 2013). This feature is particularly valuable in the study of rare cell types, such as hematopoietic stem cells and circulating tumor cells (CTCs), where sample availability is often limited (Yao et al., 2014). Furthermore, the integration of microfluidic cell isolation and detection strategies enables extensive parallelization, allowing for the quantitative analysis of hundreds to thousands of single cells simultaneously (Yu et al., 2014).

Microfluidic technologies encompass a diverse array of principles and techniques for single-cell isolation (Figure 1E). It is suitable for both live and fixed cells, with a recovery rate exceeding 90% and a purity greater than 90% for single-cell isolation. However, these values may vary depending on the experimental conditions (Yeo et al., 2016; Liu et al., 2018; Wang K. et al., 2019; Pritchard et al., 2019). Two of the most prominent approaches are microwell arrays and droplet microfluidics.

Endogenous gene tagging with fluorescent proteins or fluorophore-conjugated antibody labeling techniques have enabled highly sensitive quantification of endogenous protein levels. The fluorescence intensity obtained from these methods directly correlates with the quantity of protein molecules present, thereby enabling the assessment of relative protein abundance.

Fluorescence labeling of target proteins combined with FACS is considered the gold-standard approach for profiling proteins at the single-cell level (Hu et al., 2016). This method involves staining single cells with fluorescent-labeled antibodies or tagging them with a fluorescent protein, which are then detected as they pass through the flow chamber during FACS analysis. The fluorophores are excited by a laser, and the emitted fluorescence is subsequently measured. Through meticulous system calibration using protein-coated beads, the fluorescent intensities can be translated into single-cell protein expression levels (Adan et al., 2017).

FACS facilitates high-throughput detection (∼10⁴ cells/s) and simultaneous measurement of ∼20 multiplexing membrane and intracellular protein parameters (Liu L. et al., 2020). However, this technique has some limitations. Dynamic cell monitoring over time poses a challenge, and the multiplexing capacity is constrained by spectral overlap, potentially impacting the accuracy of protein measurements. In the multi-colour detection system, each excitation laser, including green (532 nm), blue (488 nm) and violet (407 nm) ranges and red (633 nm), is configured with multiple detectors (for the green laser, fluorochromes PE, TRPE, Cy5PE, Cy5.5PE and Cy7PE could be applied and separated). If additional fluorochromes become available that have distinct emission spectra after excitation by the green laser, the detection capacity could be increased even further. Furthermore, physical stressors experienced by cells during FACS procedures could influence their status. A critical aspect is the requirement for a large number of cells (1 × 10^6/mL) in conventional flow cytometry sample preparation (Xie and Ding, 2022).

To address these challenges, flow cytometers incorporating microfluidics have been proposed as a solution (Liu et al., 2011; Wu et al., 2012; Su et al., 2011). Wu et al. introduced a microfluidic platform that integrates cell preparation and multi-color flow cytometry components. This on-chip system offers multiplexed and orthogonal data for profiling signaling pathways across various cell types, including primary cells (Wu et al., 2012). The integration of microfluidics with flow cytometry enables the analysis of smaller sample sizes, reduces the risk of sample contamination, and allows for the precise control of the cellular microenvironment during analysis. Besides, a recent study showed a workflow for generating double emulsions and performing multicolor cell sorting using a commercial FACS instrument. This workflow achieves a double emulsion detection rate exceeding 90%, enabling multicellular encapsulation and high-throughput immune cell activation sorting for the first time (Ding et al., 2024).

Despite these challenges, FACS remains a powerful tool for single-cell protein analysis, offering high throughput and multiplexing capabilities. Researchers continue to explore and develop new techniques that address these limitations, aiming to enable more comprehensive and accurate analysis of protein dynamics and interactions at the single-cell level.

Single-cell Western blotting (scWB) is another innovative technique that combines microfluidics and conventional Western blotting to analyze protein expression at a single-cell resolution (Hughes et al., 2014). It overcomes the issue of cross-reactivity by introducing a step of electrophoretic separation before antibody probing (Figure 2A). The scWestern analysis utilizes a microscope slide coated with a thin photoactive polyacrylamide gel. This gel is micropatterned with an array of thousands of microwells, allowing for the simultaneous analysis of 10³–10⁴ single cells/chip. Simply stated, the scWestern approach integrates all the essential steps of Western blotting into a dense array format. In this technique, single cells are placed into the microwells and lysed in situ. The proteins from the lysed cells are then separated by gel electrophoresis within each microwell. Following electrophoresis, the proteins are immobilized onto the PA gel through photoinitiated blotting. Finally, fluorescent-labeled antibodies are used for protein detection within each microwell.

scWB has emerged as a promising tool for investigating protein expression in individual cells, offering valuable insights into cellular heterogeneity and dynamics at a higher resolution (Rosàs-Canyelles et al., 2020; Kim et al., 2021; Grist et al., 2020). scWB enables detection of more than 10 proteins, including both membrane and intracellular proteins, in each cell during 4 h. The sensitivity of the scWB method is comparable to that of flow cytometry (FC). However, scWB encounters challenges in quantifying secreted proteins, low-abundance proteins, and small molecular weight proteins (Duncombe et al., 2016). To address these limitations, several new methods have been suggested and developed. For example, the integration of scWB with ultrasensitive detection methods, such as single-molecule fluorescence microscopy or mass spectrometry, has the potential to improve the sensitivity and dynamic range of protein measurements (Yen et al., 2014; Lomeli et al., 2021).

In addition to using fluorescent antibodies for protein expression measurement, employing DNA amplification and reporters offers improved accuracy and higher sensitivity. With protein signal outputs encoded by various DNA sequences, the multiplexing capacity is virtually unlimited when utilizing DNA reporters. Several approaches have emerged to enable simultaneous measurement of both mRNA and protein at the single-cell level (Reimegård et al., 2021; Stoeckius et al., 2017; Peterson et al., 2017; Mimitou et al., 2019).

One such approach is Single-Cell Protein and RNA Co-profiling (SPARC) (Reimegård et al., 2021), proposed by Dr. Gallant’s group in 2021. SPARC enables the simultaneous measurement of overall mRNA levels and targeted 89 intracellular proteins in individual cells (Figure 2B). Single cells were sorted on a BD FACS ARIAIII into a 96-well plate and lysed for protein analysis. Protein quantification is accomplished through multiplex, homogeneous protein extension assay (PEA), which is an affinity-based detection technique employing pairs of antibodies linked with oligonucleotides (Reimegård et al., 2021). In PEA, each antibody pair is designed to recognize a specific protein target, and when both antibodies bind to the same protein, the oligonucleotides are brought into close proximity, allowing for their extension and amplification. The resulting DNA amplicons are then quantified using next-generation sequencing or qPCR, providing a digital readout of protein abundance (Darmanis et al., 2016).

CITE-seq, pioneered by Stoeckius et al., is another well-established method in single-cell analysis (Stoeckius et al., 2017). This innovative technique utilizes a digital, sequencing-based platform to quantify single-cell protein expression levels. It involves conjugating antibodies to oligonucleotides, incorporating a unique barcode for antibody identification and facilitating PCR amplification (Figure 2C). The antibody-oligo complexes are then applied to single-cell suspensions. After incubation, a washing step removes unbound antibodies before proceeding to single-cell RNA sequencing (scRNA-seq). CITE-seq primarily allows for the analysis of cell surface proteins, providing valuable information about cell type, state, and function. Recently, Liu et al., extended co-indexing of transcriptomes and epitopes (CITE) to the spatial dimension and demonstrated high-plex protein and whole transcriptome co-mapping to measure 273 proteins and transcriptome in human tissues with 10–50 μm Spatial resolution (Liu et al., 2023). The simultaneous measurement of mRNA and surface proteins enables a more comprehensive characterization of cellular heterogeneity and can reveal novel cell subpopulations and their functional properties.

The integration of DNA reporters with single-cell protein analysis has greatly expanded the multiplexing capacity and sensitivity of these techniques. The ability to encode protein signals with unique DNA sequences allows for the simultaneous measurement of hundreds to thousands of proteins in individual cells, overcoming the limitations of spectral overlap encountered in fluorescence-based methods (Table 3). Furthermore, the amplification of DNA reporters through PCR or sequencing enables the detection of low-abundance proteins that may be missed by other techniques (Reimegård et al., 2021).

Chemiluminescence (CL) is known for its simple and cost-effective optical systems, as it does not require an external light source. This characteristic circumvents issues related to stray light and the instability of light sources, ultimately resulting in low backgrounds and high sensitivity.

The Enzyme-linked immunospot assay (ELISPOT) is a powerful technique developed in the 1980s for quantitatively detecting individual cells that secrete a specific protein of interest (Alexander et al., 2013). This method, based on an immuno-sandwich assay, is valuable for enumerating cell subsets with distinct secretory functions and for monitoring cellular responses to external stimuli or drugs (DiPiazza et al., 2016; Barabas et al., 2017). In ELISPOT, cells are cultured on a surface coated with antibodies that target the protein of interest. The secreted protein is then captured by the immobilized antibodies. A second, enzyme-conjugated antibody is then added, which binds to the captured protein. Upon addition of a substrate, the enzyme catalyzes a reaction that produces a visible spot on the membrane, with each spot representing an individual secreting cell.

The ELISPOT assay is well-known for its high sensitivity in detecting secreted proteins, with a detection limit of approximately six spots per 100,000 cells (Moodie et al., 2010). It enables high-throughput detection (∼10⁶ cells/run) and simultaneous measurement of 1-3 secreted proteins. It has been particularly useful in studying immune responses, such as the detection of cytokine-secreting T cells or antibody-secreting B cells, and in assessing the impact of various interventions on cell function (Leehan and Koelsch, 2015). The ability to quantify the frequency of secreting cells and the relative amount of protein secreted by each cell makes ELISPOT a valuable tool for understanding the heterogeneity of cellular responses.

Recent advancements in ELISPOT technology have focused on improving its multiplexing capabilities and compatibility with other single-cell analysis techniques. For instance, FluoroSpot, a type of fluorescent ELISPOT, allows for the concurrent detection of multiple secreted proteins by utilizing fluorescent dyes instead of enzymes for signal generation (Axelsson, 2022). Additionally, the integration of ELISPOT with microfluidic devices has allowed for the isolation and analysis of individual secreting cells, providing a more comprehensive profile of their functional properties (Huang et al., 2012).

Recent advances in mass spectrometry (MS) have made single-cell MS one of the most powerful approaches to obtain the protein profile of a single cell (Slavov, 2021; Petelski et al., 2021). Mass spectrometry-based methods can generally be divided into two main categories: targeted and untargeted approaches.

The Proximity Ligation Assay (PLA) represents a relatively novel method for studying protein-protein interactions (PPIs) with high specificity, applicable in both cells and tissue sections (Söderberg et al., 2006). In the assay, a pair of oligonucleotide-labeled antibodies is used to recognize two target proteins. When the two proteins are in close proximity (less than 30–40 nm apart) due to their interaction, the oligonucleotides are ligated, subsequently extended, and amplified to generate sequence products. Detection is accomplished by adding complementary fluorescently labeled oligonucleotides or by measuring the output of DNA sequencing using qPCR.

Dr. Abramson’s group has integrated the PLA assay with imaging flow cytometry (IFC) and developed Proximity Ligation Imaging Cytometry (FLIC) (Avin et al., 2017). This integration enables multiparametric fluorescent and localization analysis of cellular events, including the examination of PPIs and post-translational modifications (PTMs), as well as the assessment of subcellular signal distribution within cells (Figure 2E). FLIC combines the specificity of PLA with the high-throughput and multiplexing capabilities of IFC, allowing for the simultaneous analysis of multiple PPIs and cellular parameters in thousands of individual cells. Vistain et al. developed Prox-seq, a method that enables the detection of intracellular and membrane proteins, protein complexes, and mRNAs in thousands of single cells. Prox-seq combines single-cell RNA sequencing (scRNA-seq) with the proximity ligation assay (PLA). It utilizes pairs of DNA-conjugated antibodies that, when brought into close proximity, facilitate the ligation of DNA oligonucleotides (oligomers) on the antibodies. This ligation generates a PLA product that can be detected through next-generation sequencing. The method allows for the simultaneous measurement of up to 741 protein complexes across 8,700 single peripheral blood mononuclear cells (Vistain et al., 2022).

To investigate PPIs directly with high spatiotemporal resolution, genetically encoded biosensors based on Förster Resonance Energy Transfer (FRET) or Bioluminescence Resonance Energy Transfer (BRET) can be employed. These resonance energy transfer (RET) techniques offer the ability to record single-cell kinetics with millisecond resolution, facilitating the identification of protein complexes and the study of cell-to-cell heterogeneity (van Unen et al., 2016). Recently, Hoshino et al. designed γB2-FRET probes by fusing FRET donor and acceptor fluorescent proteins to a single γB2 molecule, successfully visualizing γB2 homophilic interactions in cultured hippocampal neurons (Hoshino et al., 2023). By utilizing these biosensors, researchers can visualize dynamic changes in PPIs in real-time, providing valuable insights into the molecular mechanisms underlying various cellular processes.

The Single-molecule pull-down (SiMPull) technique provides a method to directly capture native protein complexes from cell lysates, facilitating the identification of complex composition and PPIs at the single-molecule level (Jain et al., 2012). SiMPull involves immobilizing an antibody against a target protein on a glass surface, followed by the application of cell lysates containing the protein complexes. The captured complexes are then visualized using fluorescently labeled antibodies against the interacting partners, allowing for the direct observation of PPIs and the determination of their stoichiometry.

Expanding on this concept, various techniques have been suggested to investigate PPIs at the single-cell level (Ryu et al., 2019a; Zhao et al., 2023; Wang et al., 2018; Wedeking et al., 2015; Stolpner and Dickinson, 2022). Dr. Ha’s group developed an assay that involves lysing bacterial cells in situ and capturing the released proteins on an imaging surface coated with antibodies. This approach allows for the unambiguous assignment of captured proteins to their originating cells (Wang et al., 2018). The developed platform is compatible with high-throughput protein analysis and PPI analysis at the single-cell level through single-molecule imaging (Figure 2F). The ability to link protein interactions to specific cells enables the investigation of cell-to-cell variability in complex formation and function, providing insights into the mechanisms underlying cellular heterogeneity.

For mammalian cells, micro-patterned surfaces have been utilized to capture target cells and proteins. Yoon et al. implemented a method that involved in situ co-IP for individual cancer cells (Ryu et al., 2019a). This process included capturing single cells using a microfluidic device and inducing in situ lysis of the cells while immunoprecipitating EGFRs (the proteins of interest) to the surface. Subsequently, fluorescently labeled downstream interactors of EGFRs were directly added to stimulate protein-protein interactions, which were then visualized using a single-molecule fluorescence microscope. This innovative approach offers insights into the dynamics of protein interactions at the single-cell level, providing valuable information on signaling pathways and heterogeneity in cancer cells (Figure 2G).

The integration of co-IP with microfluidic devices and single-molecule imaging has greatly enhanced the sensitivity and resolution of PPI analysis at the single-cell level. By capturing protein complexes directly from individual cells, these techniques overcome the limitations of bulk co-IP experiments, which average out the heterogeneity in protein interactions across a population of cells. Moreover, the ability to visualize individual protein complexes allows for the quantitative analysis of their composition, stoichiometry, and dynamics, providing a more detailed understanding of their functional roles in cellular processes.