March 15, 2015 (Vol. 35, No. 6)

Kate Marusina Ph.D.

Flow Cytometry Is Breaking Through Single-Cell Barriers on Several Fronts

Flow cytometry, a technology that takes measurements at the single-cell level, can assess phenotypic and functional parameters across diverse cell types and tissues. As this technology develops, many questions that seem difficult or even impossible to address may soon be posed and answered.

Flow cytometry advocates include the Southeast Flow Cytometry Interest Group (SEFCIG), an organization associated with the International Society for the Advancement of Cytometry. SEFCIG is dedicated to the promotion and advancement of techniques and applications of flow cytometry in all relevant scientific disciplines.

SEFCIG is particularly interested in supporting educational and networking efforts for scientists that use flow cytometry. For example, the organization recently held its annual meeting in Atlanta to bring together scientists from 10 southeastern states and beyond. Below is a review of a few key presentations from the meeting highlighting exciting advances in science enabled by flow cytometry.

“Stem cells from many types of cancer are intrinsically capable of forming multicellular organoids (tumorspheres or spheroids) when grown in vitro,” said Peter Geck, M.D., a research assistant professor in the department of integrative physiology and pathobiology at Tufts University. “We thought that a similar process of forming spheroids might also take place in vivo.”

Dr. Geck’s team developed an elegant methodology for identifying cellular spheroids in whole blood using light scatter flow cytometry. Most traditional flow cytometry applications use fluorescent labels to detect cells or organelles using excitation laser beams. However, when the cells pass the laser, they also refract (or scatter) the laser light. Generally, larger cells will refract more light than small objects, resulting in higher forward scatter signals.

“We used forward light scatter in combination with fluorescent labeling to identify spheroids that potentially carry cancer stem cells,” continued Dr. Geck.

While considerable research efforts were directed toward identifying circulating tumor cells (CTCs), their correlation with the development of new cancer foci (metastasis) is unclear. Only cancer stem cells give rise to metastasis, but they do not appear to fare well in highly oxygenated environments, such as blood.

Dr. Geck’s team was able to confirm that spheroid bodies create hypoxic conditions at the core of the spheroids. Spheroid structure may protect the inner stem cells from oxygen in the bloodstream.

Light scatter flow cytometry was used to compare blood samples of 43 patients with confirmed metastatic disease and nonmetastatic controls. Remarkably, spheroids were detected only in the blood of patients with metastatic disease, precisely in the signal range predicted by the model stem cells spheroids. The data supports the hypothesis that cancer spreads throughout the body by releasing cancer stem cells that form “embryonic” structures small enough to permeate circulation and capable of withstanding oxygen-rich environments.

Dr. Geck believes that flow cytometry could be used as a diagnostic tool to predict or diagnose metastatic disease. Moreover, cell sorters can be deployed to collect spheroids from blood thereby gaining direct access to pure cancer stem cells. Once collected, they can be cultured and tested for chemoresistance, providing useful information for personalized treatments.


Researchers at Tufts acquired this phase contrast image of a breast cancer spheroid from blood. The blood sample was taken from an invasive ductal breast cancer patient with distant metastasis. The spheroids were isolated by filtering 6 mL blood through a 20 micrometer Celltrics filter unit (Partec). The phase contrast image (20×) shows a single multicellular formation of about 50–55 micrometers. A few red blood cells are also visible.

Golden Intracellular Markers

Cancer stem cells constitute less than 5% of tumor bulk. The most reliable methodologies for identification of stem cells include dye efflux assay and enzymatic aldefluor assay. “Expression of gold standard transcription factors provides universal confirmation of stem cell identity,” said Steve McClellan, chief, flow cytometry and imaging core laboratories, Mitchell Cancer Institute, University of South Alabama. “Up to now, access to these biomarkers required destroying the cells for PCR, immunostaining or Western blots.”

McClellan’s team was the first to deploy SmartFlare™, a new RNA detection technology provided by EMD Millipore to count and isolate live cancer stem cells from freshly resected tumors. SmartFlare enables what was previously considered impossible—using flow cytometry to sort live cells based on their endogenous intracellular protein markers, all without disrupting normal cellular processes.

The technology is based on tiny gold nanoparticles conjugated to the capture oligonucleotides. The complementary reporter oligos are conjugated with fluorochromes. They bind to the capture oligos in such a way that fluorochromes are quenched in proximity of gold particles.

The particles are endocytosed by live cells using the cells’ own endocytosis machinery. If a particle encounters a complementary mRNA strand, the native mRNA displaces the reporter strand, thereby unquenching the fluorochrome. Fluorescent signal is directly proportionate to the amount of captured mRNA. After a period of time, the particles exit the cell without harming it. Gold particles conjugated with nonspecific oligos are used for establishing background fluorescence.

“Currently, SmartFlare offers an exclusive opportunity to follow the fate of cancer stem cells as the patient goes through several rounds of anticancer therapy,” continued McClellan, who added that his team is preparing for a clinical trial that will quantify stem cells in biopsies obtained from ovarian cancer patients before and after a novel investigational treatment.

McClellan explained that the majority of conventional chemotherapies debulk the tumor but are unable to kill cancer stem cells. After several rounds of chemotherapy, these stem cells become even more resistant. SmartFlare-based cell sorting would enable precise isolation of stem cells, followed by genetic testing and selection of targeted molecular treatments.

McClellan’s team has already been able to isolate the intact stem cells and confirm their identity by the presence of other classical surface markers and by functional tests. “As EMD Millipore expands the range of colors,” McClellan added, “this technology will revolutionize our ability to study intracellular processes in live cells.”


Researchers at the University of South Alabama used SmartFlare RNA detection technology by EMD Millipore to isolate and count live cancer stem cells from freshly resected tumors. The method avoids the use of potentially destructive surface markers and functional assays.

Complex Immune Responses

“Immune responses are built on complex interactions between proteins,” said Pratip Chattopadhyay, Ph.D., a staff scientist in the immunotechnology section of the vaccine research center at the NIH. “Thousands of molecules govern the transition from resting immune cells to the dynamic, functionally active cells that fight disease. To better understand these processes, multiple biomarkers must be measured simultaneously.”

Dr. Chattopadhyay is a pioneer in multicolor flow cytometry, performing the first 18-color experiments in 2003, and more recently leading the development of 28-color technology. To do this work, he uses a wide palette of fluorescent dyes including classic organic fluorophores, tandem dyes, quantum dots, and new inorganic “Brilliant” dyes, which he combines into antibody panels designed to measure the frequency and function of a wide variety of blood cells.

By taking these measurements, Dr. Chattopadhyay aims to identify the cell types most important for different immune responses. For example, in 2010, his team described cell types whose frequency early in HIV disease predicted how quickly untreated HIV-positive individuals would develop AIDS.

“I quickly realized that our classical data analysis tools were quite labor intensive and a bit subjective and couldn’t sift through the entire dataset,” Dr. Chattopadhyay recalled. Working closely with bioinformaticians, he developed new tools to analyze complex flow cytometry datasets, revealing rare, previously unappreciated cell types that also predicted HIV progression.

“Now, we’re applying these tools and technologies to vaccine development,” continued Dr. Chattopadhyay. “We hope to find unique immune cell populations, emerging soon after vaccination, that predict whether a vaccine will work. The goal is to provide a screening tool early in the vaccine development process, which more clearly identifies promising vaccine candidates than current, overly simplistic, immunogenicity measurements.”

Interestingly, although polychromatic flow cytometry is the most common tool for “deep” profiling of cells, Dr. Chattopadhyay is also evaluating new technologies that promise to offer even more information about immune processes. These include molecular tools, such as qPCR platforms that simultaneously measure 96 gene transcripts from single cells, and RNA sequencing.

“By coupling these technologies to flow cytometry, we have an incredibly powerful pipeline for characterizing immune cells,” he explained. Given the enormous data available from these cutting-edge single-cell technologies, bioinformatics tools and fully automated data analysis approaches will be more important than ever.

“We need strong partnerships to do this work successfully, with immunologists and bioinformaticians having a solid foundation in each other’s field,” remarked Dr. Chattopadhyay. “This is a great time for scientists in each field to learn and grow.” Soon, such efforts will allow unequivocal identification of important immune subsets, and provide a new understanding of disease biology.

Integrating Imaging and Analysis

ImageStream technology from EMD Millipore is enabling a new generation of flow cytometry. The ImageStreamX, for example, uniquely combines multiparameter fluorescent analysis and real-time imaging of individual cells in flow.

This multispectral imaging system acquires up to 12 images per cell in three different imaging modes: brightfield, darkfield, and fluorescence. The light is transmitted in discreet spectral bands to a CCD camera. With this technique, a cell image can be optically decomposed into a set of sub-images (each of its own fluorescent color). Each image portrays localization of a fluorescent probes; image overlap enables visualization of colocalization of probes on the cell surface or in the cellular interior.

“While conventional flow cytometry dot plots yield important information, in many cases they are not sufficient to confirm a scientific hypothesis,” said Bob Smith-McCollum, director of marketing, molecular and cellular biology systems, EMD Millipore. “Mystery populations would remain such until sorted, stained, and visualized under a microscope. ImageStream not only saves time and labor spent on microscopy, but it also greatly expands the range of questions to be tested.”

Scientists can use ImageStream to interrogate dynamic events such as protein trafficking, cell signaling, autophagy, and changes in cell morphology. In a recently published study by Orla Maguire, Ph.D., a research scientist at the Roswell Park Cancer Institute, ImageStream was used to measure NF-κB localization. This protein is active in a variety of cancers and is actively explored as a target for anticancer therapies.

Conventional techniques fail to provide statistically robust estimates of NF-κB nuclear localization, and they are unable to achieve reasonable sensitivity in detecting rare events. ImageStream is able to capture high-resolution images of cells at the rate of up to 5,000 cells/second and to provide quantitative assessment of the cellular localization of NF-κB. Highly reproducible measurements produced by the instrument make it ideal for evaluation of therapies aiming to modulate NF-κB pathway activity.

Smith-McCollum underscored the decade-long effort that went into improving sensitivity of ImageStream as well as into creating over 85 proprietary scoring algorithms. The algorithms compare images pixel by pixel and provide scoring based on statistical metrics.

“Our Feature Finder App determines which algorithms produce statistically significant differences in scores. This application helps scientists determine how cells differ by morphology, phenotype, or functional characteristics,” continued Smith-McCollum. “This would never be possible using conventional flow cytometry or microscopy alone.”


Amnis developed the ImageStreamX imaging flow cytometer. The instrument is meant to combine the speed, sensitivity, and phenotyping abilities of flow cytometry with the detailed imagery and functional insights of microscopy.

High-Dimensional Cytometry Data

Multiparametric analysis of single cells is a key technology for understanding cellular heterogeneity. “While advanced immune staining methods and new instrumentation provide tremendous ability for multidimensional data capture, methods for analyzing this data remain inadequate,” contended Peng Qiu, Ph.D., an assistant professor of the department of biomedical engineering at the Georgia Institute of Technology and Emory University.

Typically, selection of cell subsets for analysis involved “gating,” which is manual or automated clustering of data on flow cytometry dot plots. Dr. Qiu explained that gating adopts a discrete mindset and aims to define distinct cell populations.

Yet clear distinctions may be hard to find among cells in a tissue sample. Such cells are likely to exhibit continuous progression from one state to another, such as development of blood cells from bone marrow precursor cells, or development of a tumor from cancer stem cells.

“The need to represent all transitional cellular states without knowing their relationships inspired us to develop the spanning-tree progression analysis of density normalized events (SPADE) algorithm,” remarked Dr. Qiu. The name of the algorithm highlights key innovations of this computational method.

SPADE attempts to give equal representation to all cell populations, rare and abundant alike, via normalization of density (or density-dependent downsampling). Next, SPADE clusters cells with similar phenotypes based on expression of cellular markers. Because of density normalization, all cells, including rare cells, are allowed to form their own clusters. Finally, SPADE generates a tree to approximate the “skeleton” of the cloud that connects all clusters based on their similarity. Such skeleton tree can be colored to visualize gradual changes in biomarker intensities.

“SPADE can also be viewed as a feature extracted tool,” continued Dr. Qiu, “which can be coupled with machine-learning algorithms to make decisions.”

SPADE was used in a DREAM (Dialog for Reverse Engineering Assessments and Methods) challenge on AML (acute myeloid leukemia) prediction. Based on flow cytometry data of 369 subjects and normal/AML status of 179 of these subjects, the challenge was to predict the disease status of the remaining subjects.

Dr. Qiu was able to achieve 100% prediction accuracy using SPADE followed by decision-making algorithms. The SPADE algorithm is uniquely capable of capturing “intermediate” phenotypes, and it provides hints on dynamic changes as cells morph from progenitor cells to defined immune cell types.


This image details a portion of SPADE, an algorithm developed by researchers at the Georgia Institute of Technology for visualizing and analyzing high-dimensional flow cytometry data. The input consists of FCS files (flow cytometry data matrices), and the output consists of tree plots colored by measured markers in the data files.

Bringing Flow Cytometry to the People

“Multicolor flow cytometry is indeed a powerful tool and is answering many important questions,” said Melissa Ma, global product manager, cell biology business unit, Bio-Rad Laboratories. “However, more than half of the researchers using flow cytometry only need one- or two-color sorting to support their discoveries.”

Bio-Rad fills the instrumentation gap for those who need quick, readily accessible, and reliable cell sorting or may not currently have access to flow cytometry resources and expertise. Bio-Rad’s new S3e™ instrument is a benchtop cell sorter that packs an astounding array of technological innovations in a small footprint. Key features enable almost entirely automated instrument setup.

Mike Kissner, field applications specialist, cell biology business unit, Bio-Rad, highlights ProDrop™ Technology, a fully automated drop delay calculation and maintenance feature. To sort fluorescently labeled cells, droplets containing these cells must be charged at an extremely precise instant as they pass through the system. ProDrop automates the calculation of charge timing, and this feature contributes to sorts with consistently high purity and recovery on the S3e.

“Another key innovation is the automatic alignment of the nozzle using a set of five positional adjustments,” noted Ma. “Even a very small micrometer misalignment may interfere with analytical and sorting functions.” The AutoGimbal™ system combines picomotor fine motion position, imaging and software to enable precise alignment of the nozzle tip. “Automated setup and innovative smart features make this instrument exceptionally easy to operate,” added Ma. “As little as an hour is required for a user to learn the basic operating principles.”

Despite the apparent simplicity, the S3e is able to efficiently sort highly heterogeneous populations of cells. Mesenchymal stem cells (MSCs) constitute less than 1/100th of a percent of all hematopoietic cells of bone marrow. The Bio-Rad team was able to identify and isolate progressively differentiated MSC with 90–99% purity depending on the cell subtype. The MSCs retained their viability and were successfully propagated in culture.

“Bio-Rad strives to provide highly compatible products so that the scientists can seamlessly move from one step to another,” asserted Ma. Cell sorting is an integral part of Bio-Rad’s connected workflow that includes real-time PCR, digital PCR, and proteomics products.


The Bio-Rad S3 cell sorter is an automated benchtop instrument featuring either one or two lasers and up to four fluorescence detectors plus forward- and side-scatter detectors. The S3 cell sorter uses ProDrop technology for automated drop delay calculation and droplet break-off monitoring.

Previous articleAutism Genes Found to Be Associated with Brighter Minds
Next articleMerger-Bound Salix Is New Object of Endo’s Desire