A Ghastly Overview of Ghost Cytometry

Flow cytometry has become widespread and indispensible tool for medical and scientific research. Given the current demand and future growth projection, the global flow cytometry market is expected to exceed more than $4.93 billion (US) by the year 2022 (Forbes, 2018). Researchers in Japan have recently developed a new method called ghost cytometry that they claim can provide the same functions as flow cytometry, but in a more timely and cost effective manner.  In this blog, we explore basic principles of ghost cytometry, and see how it compares to flow cytometry.

Flow cytometry is a cell analysis technique that utilizes laser-based technology to count, sort, and profile cells from a heterogeneous mixture.  In flow cytometry, the cells travel individually through a rapidly flowing fluid stream until they intersect the path of laser(s), creating reflections and refractions of light known as light scatter as seen in the first part of Figure 1. The electronic detection apparatus of the flow cytometer measures light scatter, along with the fluorescence emission from excited fluorophores associated with cellular antigens on individual cells as they travel through the light, as seen in Figure 2. These measurements detect multiple physical and chemical properties of cells, allowing them to be separated into distinct populations as visualized in the second part of Figure 1

Figure 1: Flow cytometry visualization. A single cell suspension passes through a laser. Light scatter properties determine the size and granularity of the cell population, in this case, human leukocytes. Graphic courtesy of Hillary Reif.

Figure 2: How cells tagged with flourescent antibodies interact with lasers in a flow cytometer. Scatter properties allow the measurement of the size and granularity of the cells. Antibodies tagged with flourochromes emit light which measures specific antigens associated with the cell. Graphic courtesy of Hillary Reif.

Flow cytometry is used for not only cell counting and cell sorting, but also for the detection of cell surface and intracellular antigens. Flow cytometers have the capability to measure thousands of particles per second, making it an incredibly fast and powerful tool with many applications. Flow cytometry is used for multiple immunoassays, the measurement of intracellular or extracellular antigens (immunophenotyping), phosphoylation- signaling, biomarker detection, mRNA detection, receptor occupancy (RO) assays, and to measure the cell cycle and apoptosis markers. With these capabilities it is commonly used for both research and clinical purposes (diagnosis, prognosis, monitoring treatment, and drug response).

Ghost cytometry has the ability to classify cells based on size and shape without producing an actual image of the cell/object, and without relying on molecular markers to identify cell type (Kwon, 2018). To separate cells based on size and morphology, ghost cytometry uses spatial information gained from a cell labeled with fluorophores moving across pseudorandom optical structures, Figure 3. This allows ghost cytometry to acquire data faster, as acquisition depends on the speed that the cell moves across it’s structures, rather than how fast the equipment can move along the cell (Ota, et al, 2018). As the cell passes over the structure, a light source excites the fluorophores on the cell (Kwon, 2018).  This excitation is then converted into a signal that is recorded by a single pixel detector. A temporal waveform is then calculated by comparing the intensity of the light emitted by the random light structures with the intensity of the flurophore excitement as the labeled cell passes over the light structure. These intensities are again recorded with the single-pixel detector, and the measured intensities are then integrated into a unique signature, Figure 4. In “image mode” of ghost cytometry, a 2D image can be reconstructed based on the information gathered from the single pixel detector. However, in “image-free mode” machine-learning methods are applied to compress the temporal waveform to classify the cells (Ota, et al, 2018).

Figure 3: Basic principle of ghost cytometry. As a cell/object moves over the pseudorandom static light structure the flourophores are excited, and their emissions (F1 and F2) are compressively mapped to calculate the total fluorophore emission of a(n) object/cell. Graphic courtesy of Hillary Reif.

Figure 4: Breakdown of how measurements are taken in the ghost cytometer. As the fluidic stream of cells travel through a 3D hydrodynamic flow focusing structure, the photomultiplier tube (PMT) illuminates the cells. The analog signals from this illumination excitation are measured at the PMT and digitized and analyzed by the field programmable gate array (FPGA). The analysis is then used at the piezoelectric actuator (PZT) to separate the cells. Graphic courtesy of Hillary Reif.

One potential drawback of ghost cytometry is that it only analyzes single cells, while flow cytometry analyzes both a population of cells and a single cell (Han, et al, 2016). However, ghost cytometry could potentially reduce cost, as it is a machine-learning technique and therefore would not require as much training for the analyst.  Some controversy exists around the use of ghost cytometry:  :  reviewers found that   cell image shown  was not compared to validated images of the same cell, which makes it difficult to claim the cell image was accurately reconstructed.  Also, the high throughput (3000 cells/s), and high signal to noise ratio (SNR) claimed for Ghost Cytometry in the original paper (Ota, 2018), were found to be insufficiently supported by the data(Di Carlo, et al, 2019). Although flow cytometry has the ability to assess the viability of cells; there is no mention of the ability or inability of ghost cytometry to do this.

However, ghost cytometry has the potential to be used for cancer treatment and diagnosis, as its capabilities could be used to determine healthy cells from cancerous cells based on morphology rather than cell markers (Kwon, 2018). The analysis of these single cells from tissue biopsies would allow medical professionals to determine healthy cells from cancerous cells.  Ghost cytometry can also be used to perform studies on the life cycle of cells, as currently flow cytometry methods cannot separate cells based on life stages such as mitosis and meiosis (Kwon, 2018). Microfluidic cytometers used in ghost cytometry are typically smaller, less expensive, and portable compared to the typical flow cytometer (Shrirao, et al, 2018). Furthermore, ghost cytometry is faster than its preceding method, ghost imaging, as it has the capability to utilize the objects speed, to make its processing speed faster. Ghost cytometry creates compressed images of the cells that require less storage than flow cytometry data, which typically has large files (Ota et al, 2018; Han et al, 2018).

Figure 5: Cartoon representation of Ghost Cytometry. Graphic courtesy of Hillary Reif.

Through these blog articles, FCSL strives to inform the scientific community of cutting edge technology and methods in the field of flow cytometry.  FCSL is a contract flow lab that provides high throughput and high capacity flow cytometry services, running multiple flow cytometers with up to 10 color antibody panels daily.  We are proficient in processing a multitude of specimen types including whole blood, frozen PBMCs, bone marrow, and other tissues. We also have extensive cell culture and tissue processing capabilities. Our flexibility in handling so many specimen types allow for the support of a wide range of flow cytometry assays including: immunophenotyping/lymphocyte subset analysis, receptor occupancy, functional assays, phosphor-signaling assay, and cell viability/apoptosis measurements. Our expert staff is always available to help guide you through these tests and we welcome clients to visit our facility. We encourage sponsor engagement throughout the process. Contact us for more information!

 

References:

  1. Ota, R. Horisaki, Y. Kawamura, M. Ugawa, I. Sato, K. Hashimoto, R. Kamesawa, K. Setoyama, S. Yamaguchi, K. Fujiu, K. Waki, H. Noji, Ghost Cytometry. Science 360 1246-1251(2018). doi:10.1126/science.aan0096pmid:29903975