The characterization and application of the Cellaca™ MX high-throughput cell counter

Bioprocessing applications for cells and biologics have vastly increased the number of samples necessary to test for cell therapy and immunotherapy. The cell counting time is a significant bottleneck for conventional counting methods, which can be abolished by using a rapid, ultra-precise, high-throughput system.

This article will outline and show the application of Cellaca™ MX high-throughput cell counter in both bright field and fluorescence imaging modes. The system was exposed to a number of characterization experiments using microbeads, Jurkat, and CHO-S cells.

We evaluated the consistency and precision of bead/cell counting from a count-to-count, plate-to-plate, and instrument-to-instrument level by assessing counting on several consumables and instruments.

The precise results were obtained by comparing directly as many as 32 Cellaca™ MX instruments over a prolonged period of time (1 year) utilizing stable bead reference samples, considerably boosting confidence in the cell counting results.

We further determined the system under ISO Cell Counting Standard Part 2 guidance to establish the quality of the cell counting method. A comparison of the system was also conducted between the traditional hemocytometer and single-sample-based automatic cell counters.

Finally, the use of Cellaca™ MX for measuring a 4-log range of T cell concentrations is presented. The Cellaca™ MX high-throughput cell counter can quickly produce cell counts at 1 and 3 minutes per 24 counts in bright field and fluorescence, respectively. Its use can reduce cell counting time significantly and effectively eradicate this bottleneck for downstream assays.

There has been a significant increase in preclinical and clinical research and development for cell therapy and immunotherapy in the last two decades following the U.S. Food and Drug Administration’s approval of numerous checkpoint inhibitors (e.g., CTLA-4, PD-1, and PD-L1) and chimeric antigen receptor (CAR) T cell therapy (e.g., Kymriah® and Yescarta®).

The rapid growth of research and production of therapeutic cells and biologics have emphasized the need to evaluate more antibody candidates, testing conditions, and patient samples.1,2

Cell count and viability measurements are crucial for research, development, and generating therapeutic cells and antibodies.

For instance, a greater number of variable conditions for media/feed optimization or other environmental factors are necessary when testing effects on Chinese hamster ovary (CHO) cells to improve the quality and quantity of therapeutic biologics production.

Personalized medicine, including CAR T cell therapies, necessitates the evaluation of an enormous number of patient samples. Finally, numerous cell-based assays with multiple conditions designed for both cell and immunotherapies demand precise cell counts to fully understand the results.3–5

When the number of testing conditions and cell samples grows, bottlenecks can occur when relying on conventional cell counting technologies and methods that inhibit throughput and precision.

Although repetitive and labor-intensive, the manual hemocytometer has been the gold standard for cell counting for more than a century. Yet, it has gradually been replaced by cost-effective bench-top automated cell counters in recent years.6–8

These image-based, single-sample cell counters may need up to 2 minutes per sample because of the necessity to manually change cell counting chambers during each run.9 Fluidic/bright field (BF) image-based cell counters usually use an automatic carousel that can retain multiple samples, but they may also necessitate sample times up to 2 minutes due to the fluidics operation time.9–11

Maintaining precision and consistency when cell counting is crucial to ensure high-quality and reproducible results. There are a number of cell counting systems used in research and development and throughout the manufacturing of therapeutic cell and antibody products.

It is critical that the same system models produce consistent and comparable results to instill confidence in the cell counting methods.9,12 Viability staining used in the cell counting process is a key element also worth considering.

Trypan blue (TB) is frequently used in conventional cell counting methods but has been known to generate cell counting and viability variations.13 Using TB with primary cells (e.g., peripheral blood mononuclear cells, mouse splenocytes) and apheresis samples that include red blood cells, platelets, and debris can lead to high, imprecise counting results.14

TB can also rupture dying or dead primary cells, so dead cells may be undercounted, leading to overvaluation of cell viability. Finally, TB can provoke cytotoxicity when the sample is incubated for longer periods of time.15,16

On the other hand, fluorescence (FL)-based counting methods using dyes, such as acridine orange (AO) and propidium iodide (PI), can precisely identify nucleated cells with negligible cytotoxic effects. The capacity to use FL-based methods for cell count and viability measurements can eradicate the issues related to TB.

There is a critical need to enhance throughput, precision, consistency, speed, and versatility of cell counting systems to satisfy growing cell counting demands.

This work demonstrates the use of the Cellaca™ MX high-throughput cell counter (Nexcelom Bioscience, Lawrence, MA) to enhance the cell counting consistency and efficiency necessary for the cell and immunotherapy workflow.17,18

This was done by characterizing the cell counting consistency and precision via repeated measurements using BF and FL beads across numerous lots of manufactured instruments. Additionally, similar consistency and precision measurements were conducted with Jurkat and CHO-S cells to replicate cell therapy and bioprocessing experiments.

The standardization document “ISO 20391-2:2019 Biotechnology – Cell Counting – Part 2: Experimental Design and Statistical Analysis to Quantify Counting Method Performance” (ISO Cell Counting Standards Part 2), which was recently published, was used to validate the quality of the cell counting result from Cellaca™ MX.19–21

The high-throughput cell counting method was contrasted against the results acquired using a hemocytometer and two other single-sample image-based cell counters (Cellometer Auto2000 and Vision, Nexcelom Bioscience).

The characterization, evaluation, and comparison results demonstrated a considerable improvement in speed, causing a reduction in counting time to around 4 and 8 minutes (10 and 20 seconds per sample) for 24 cell samples in BF and FL, respectively.

The Cellaca™ MX demonstrated consistent high cell counting precision, as well as high cell counting quality equivalent to the Celigo® Image Cytometer (Nexcelom Bioscience). The precision results were obtained by assessing a large set of Cellaca™ MX instruments over a 1 year period, which considerably boosts confidence in the cell counting results.

The system also measured cell concentrations that were analogous to three methodologies (hemocytometer, Cellometer® Auto2000 and Cellometer® Vision), which may offer an initial protocol for users of single-sample image-based cell counters to compare and move to implement a high-throughput system.

Finally, primary T cell counting in fluorescence reached linear results over a 4-log concentration range. These results show the Cellaca™ MX’s high-throughput cell counter capacity with the potential to improve upon the efficiency, consistency, and versatility of single-sample cell counters.

This is especially relevant to the cell and gene therapy sector, as well as researchers working with multiple mouse samples when collecting PBMCs from many patients or optimizing conditions for adeno-associated virus (AAV). Overall, production can benefit from eradicating the bottleneck of cell counting time.

This platform is an invaluable tool for BF- and FL-based cell counting assays that are of particular relevance in the research and development workflow, as well as cell and immunotherapy product manufacturing.

Methodology

Cell culture & sample preparation

Jurkat cells (ATCC, Manassas, VA) were cultivated in RPMI 1640 (Gibco, Gaithersburg, MD) with the addition of 10% fetal bovine serum (Access Biologicals, Vista, CA) and 1% penicillin-streptomycin (Gibco) in T-75 culture flasks at 37 °C  and exposed to less than 5% CO2.

CHO-S cells (Gibco) were cultivated in CD CHO medium with the addition of 1% GlutaMAX-1 (Gibco) and HT Supplement in T-75 culture flasks at 37 °C and exposed to less than  8% CO2.

Before measuring cell concentration and viability using a Cellometer® Spectrum, the Jurkat and CHO-S cells were directly obtained from the flasks and stained 1:1 with a mixture of 20 µL of AO/PI (ViaStain™ AO/PI Staining Solution – CS2-0106-5 mL, Nexcelom Bioscience).

Cellometer® Auto2000 cell counter

The Cellometer® Auto2000 previously described uses one BF and two FL channels to quantitatively measure the concentration and feasibility of a target cell sample.22 The excitation (EX)/emission (EM) filter sets to observe AO/PI fluorescence for counting are 470/535 nm and 540/605 nm for the green and red channels, respectively.

Target cell samples were stained 1:1 with AO/PI, pipetted (20 μL) into a Nexcelom disposable counting chamber (CHT4-SD100), and then placed into the system for imaging at four locations and evaluated using the default counting parameters. Image acquisition and analysis were generally <2 minutes per sample.

Cellometer® Vision & Spectrum cell analyzers

As described in previous publications, the Cellometer® Vision and Spectrum are both platforms equipped with one BF and two FL channels to quantitatively measure target cell sample concentration and feasibility.23–28  

The interchangeable EX/EM filter sets to identify AO/PI fluorescence was 475/534 nm and 527/655 nm for the green and red channels, respectively. Preparation of the target cell samples was conducted as described in the Cellometer® Auto2000. Image acquisition and analysis were generally <2 minutes per sample.

Celigo® Image Cytometer

The Celigo® Image Cytometer for performing high-throughput cell-based assays in conventional microtiter plates as described previously;29–33 it possesses a single BF and four FL imaging channels in blue (EX: 377/50 nm, EM: 470/22 nm), green (EX: 483/32 nm, EM: 536/40 nm), red (EX: 531/40 nm, EM: 629/53 nm) and far red (EX: 628/40 nm, EM: 688/31 nm).

The Celigo software application “Expression: Target 1 + 2” was utilized to directly count AO-stained Jurkat and CHO-S cells in the Cellaca™ plate with a 12 x 2 Cellaca™ plate profile. The instrument setup was assembled to acquire images in Target 1 (BF) and Target 2 (Green) with the AO exposure time set to 4,000 μs.

Utilizing hardware-based autofocus was used to focus in the BF channel, and focus offsets were applied for the Green (+26 μm) channel.

Optimization of the preset ANALYZE parameters enabled automatic counting of the cells but disregarded debris and nonspecific particles.

For AO-stained Jurkat cells and CHO-S cells, the ANALYZE parameters for the green channel are displayed in Supplementary Table 1. The ANALYZE parameters for the BF channel are also displayed in Supplementary Table 1.

The BF channel was not analyzed, it was only used for visualization. The counting results were exported into an EXCEL (Microsoft Corp., Redmond, WA) template to calculate the respective cell concentrations directly.

Cellaca™ MX high-throughput cell counter

The Cellaca™ MX FL5 high-throughput cell counter uses one BF, four EX (365, 470, 527, and 620 nm), and five EM (452, 534, 605, 655, and 692 nm) filter combinations. The optical system utilizes an epi-fluorescence setup with an imaging resolution of around 1.27  µm2 per pixel.

Using either TB, AO, or AO/PI, target cell samples were stained 1:1  in the mixing wells on the Cellaca™ plates directly. Subsequently, 50 μL of the stained cell samples were moved into the loading wells on the Cellaca™ plates in either a 3 x 8 (CHM24-B100-020) or 12 x 2 (CHM24-A100-020) format with a total of 24 sample chambers.

The plate was then placed into the high-throughput cell counting system for acquisition and analysis of the images. The default cell counting analysis algorithms were selected for TB, AO, or AO/PI to count cells and measure feasibility.

The system can image and analyze 24 samples in BF and FL at 1 and 3 minutes, respectively, without the need to autofocus. When applying autofocus, the times corresponded to 4 and 8 minutes, respectively. The results were contrasted against those acquired using the Cellometer® Auto2000, Vision and Celigo® Image Cytometer.

UV-cured bead reference plates

Preparation of robust, stable reference samples for assessing counting performance was carried out using microbeads and UV-curing polymer.

Three types of microbeads were applied: a non-fluorescent 5.0-μm poly latex microbead product (SPI Supplies, West Chester, PA), and a mixture of 70% green (Dragon Green, 7.5-μm) and 30% red fluorescent (Envy Green, 10-μm) microbeads from Bangs Laboratories Inc. (Fishers, IN).

Microbead suspensions were left to evaporate inside conical tubes. After the microbeads dried, around 2–3 mL of viscous UV-curable polymer solution was introduced to the conical tubes. The conical tubes were subsequently wrapped in black fabric for light protection and rotated for up to 3 weeks on a rotisserie-style tube rotator (RKVS, Laurel, MD) for resuspension of the beads.

Small metal weights were introduced to the tubes to eliminate any beads stuck to the sides. After resuspension, dilution of the concentration of the beads was conducted by introducing more UV-curable polymer.

The final bead solutions were pipetted into the loading wells of the Cellaca™ plates and left to flow into the counting chambers via capillary action. Using high-intensity UV light, the filled plates were then illuminated for 30 seconds to cure the optically clear polymer and lock the beads into position.

To avoid any potential photodamage to the beads, the finished plates were then stored in the dark.

UV-cured bead counting consistency and precision in BF

Preparation of four UV-cured reference plates (Cellaca™ plates 12 x 2) was conducted with non-fluorescent beads at two concentrations (5 x 106 beads/mL and 1 x 106 beads/mL, 2 plates each).

The BF UV-cured reference plates were evaluated on 32 Cellaca™ MX instruments produced over a period of 10 months. In another experiment, one of the plates was successively evaluated 20 times on a single instrument to acquire the analysis-to-analysis and scan-to-scan variation for this assay.

CHO cell counting consistency & precision in BF

A 600-μL aliquot of CHO-S cells was acquired and mixed in a microtube with 600 μL of 0.2% TB solution (STEMCELL Technologies, Vancouver, Canada). The resultant 1.2 mL of TB-stained cell sample was used to fill up 20 Cellaca™ plate counting chambers, evenly divided between 2 Cellaca™ plates at 50 μL/well.

Both plates were imaged and analyzed immediately on 5 Cellaca™ MX instruments in quick succession for a total of 100 measurements. All instruments utilized equivalent counting parameters as detailed in the default Cellaca™ assay for “CHO Trypan Blue Viability.”

UV-cured bead counting consistency & precision in FL

Pilot experiments demonstrated the fact the green fluorescent beads were considerably brighter than red. Thus green beads were used for the remainder of this study. Two UV-cured reference plates (Cellaca plates 12 x 2) were primed using a 6-point dilution series of the fluorescent bead mixture with four replicates for each dilution.

It is crucial to note that the viscosity of the UV curable polymer solution inhibited precise dilution ratios. The concentrations of green beads at their highest and lowest were ~5 x 106 beads/mL and 1.5 x 105 beads/mL, respectively.

The two reference plates were measured in 13 Cellaca™ MX instruments manufactured over a period of 6 months. In a subsequent experiment, one of the plates was evaluated 20 times on a single instrument for the study of the analysis-to-analysis and scan-to-scan precision.

Jurkat cell counting consistency and precision in FL

A 600-μL aliquot of Jurkat cells was collected and mixed 1:1 with the AO/PI solution. The resulting 1.2 mL of AO/PI-stained cell sample was used to fill 20 Cellaca™ plate counting chambers, divided evenly between 2 Cellaca™ plates at 50 μL/well.

Both plates were imaged and analyzed immediately on the same five Cellaca™ MX instruments used in the CHO experiment. The instruments were operated in FL mode using the in-built AO/PI viability assay with default counting parameters.

To calculate the precision of the cell counting, the same experiment was conducted nine more times on various days, with two to four instruments in each experiment. Including those used in the 5-instrument experiment, 15 different instruments were used in the study.

The Jurkat cell concentration ranged from 5  x  105 to 2.75 x 106 cells/mL, with feasibility ranging from 25% to 100%.

Bead and cell counting consistency and precision calculations

Counting assay precision from count-to-count in each experiment was calculated for each plate/instrument combination separately, and the resulting coefficients of variation (CVs) were consolidated following the equation:

CV Pooled

Where nk is the number of measurements of the kth experiment, and CVk is their CV. Counting assay precision from plate-to-plate was calculated by averaging all measurements for each plate and calculating the CV for the consequent plate averages (n = 2/experiment).

The results for the 10 experiments were then consolidated in the same way. Counting assay precision from instrument to instrument was conducted in a similar way. System-wide cell counting precision was characterized as the CV for the whole collection of data for each cell sample, including all wells, plates, and instruments.

The resulting CVs for the 10 experiments were then consolidated as before. There was no correction for the dependence of variation on cell concentration; instead, all CVs were equally weighted in the pooling, irrespective of concentration.

Counting assay precision from scan to scan was calculated by consolidating the CVs of 24 wells (BF) and 6 concentration groups of 4 wells (FL), scanning each well 20 times on the same instrument. The precision for analysis-to-analysis was conducted by analyzing the same images a further 20 times and calculating the consolidated CV.

Comparison of Cellaca™ MX to hemocytometer in BF

Using ~10 mL of non-fluorescent 5-μm poly latex beads in water at a concentration of ~2 x 106 beads/mL, three 15-mL conical tubes were filled. The bead concentration in each tube was evaluated by a trained operator using a conventional hemocytometer and a light microscope.

Forty manual counts were conducted for each tube, each comprised of four squares on the hemocytometer. Subsequently, each tube was used to fill all 24 wells in 6 Cellaca™ plates (total of 144 Cellaca™ counting chambers) and counted on the Cellaca™ MX utilizing the BF concentration assay, with the contrast parameter set to 0.6.

Additionally, 3 samples of high-viability CHO cells were acquired from the culture at ~2  x  106 cells/mL and 0.5  x  106 cells/mL. The CHO cell samples were stained 1:1 with TB at 50 µL in the mixing wells on the Cellaca™ plates and then placed into 24 loading wells. Up to 4 samples were counted manually using the hemocytometer.

Comparison of Cellaca™ MX to Cellometer® Vision & Cellometer® Auto2000 in FL

Preparation of the Jurkat cells was conducted in three conical tubes at concentrations approximately 6 x 105, 1.2 x 106, and 1.9 x 106 cells/mL in a volume of 4–7 mL/tube and rotated gently using a tube rotator.

The initial cell concentrations were measured by staining cells with AO/PI and counted directly in Cellometer® Auto2000. The Jurkat cell suspensions were sequentially evaluated, with all measurements for one tube completed before moving on to the next concentration.

For each series of measurements, the conical tube was gently inverted five times, and a 15-μL cell sample was immediately aliquoted and mixed 1:1 with AO/PI. After staining, 20 μL was loaded into one of the two counting chambers in a Nexcelom cell counting slide. The staining and loading procedures were repeated to prepare 12 chambers on 6 slides.

Jurkat cells from the same tube were then used to load 12 counting chambers on a Cellaca™ plate by mixing 50 μL of cells with 50 μL of AO/PI in the mixing well and then transporting 50 μL of stained cells into the loading well. The slides that had been prepared were imaged and analyzed using two Cellometer® Auto2000 and two Cellometer® Vision instruments.

The prepared plates were imaged and analyzed using two Cellaca™ MX instruments. Green and red FL channels were employed for all instruments with equivalent exposure and cell counting settings on both instruments of each type.

ISO Cell Counting Standards Part 2

The ISO Cell Counting Standards Part 2 protocol was applied for the comparison of the cell counting performances of the Cellaca™ MX and Celigo® Image Cytometer. The Bland-Altman comparative method was also applied to assess statistical bias between the two cell counting methods.34–37

Once sample preparation was complete, Jurkat and CHO-S cells were gathered and placed into two separate 15-mL conical tubes and acclimated to ~5 x 106 cells/mL to generate the stock concentration for use in the ISO Cell Counting Standards Part 2.

Subsequently, cell samples with varying dilution fractions (DF: 1.0, 0.9, 0.7, 0.5, 0.3, 0.1) were independently produced (n = 3 per DF), and then stained 1:1 with AO (ViaStain™ AO Staining Solution – CS1-0108-5mL, Nexcelom Bioscience) in microtubes.

The stained cell sample from the first microtube of each DF sample was pipetted into the first Cellaca™ plate (n = 4 per plate). Once prepared, the first plate was directly imaged and assessed using the Cellaca™ MX and Celigo® Image Cytometer.

The second and third Cellaca™ plates were prepared, imaged, and analyzed following the same steps. A total of 12 observations were carried out per DF sample in these 3 plates.

For AO-stained Jurkat cells, the Cellaca™ MX analysis parameters for the green channel were set to: “Min Diameter = 3,” “Max Diameter = 25,” “Roundness = 0.1,” and “Intensity Threshold = 15.” For AO-stained CHO-S cells, the ANALYZE parameters for the green channel were set to: “Min Diameter = 2,” “Max Diameter = 40,” “Roundness = 0,” and “Intensity Threshold = 20.”

The cell counting and concentration results were evaluated by applying an in-house developed software program to calculate the coefficient of determination (R2 ), consolidated CV for each DF sample, and proportionality index (PI) as indicated in the ISO Cell Counting Standards Part 2.

The results were then compared directly using the Bland-Altman comparative analysis method to establish the bias, limits of agreement (LoAs), and bias confidence interval (CI).

Bland-Altman statistical analysis

Comparison of cell counting methods was conducted using mean-difference or Bland-Altman plots.34,35 Since the variance of replicate cell counting measurements is generally proportional to the average concentration, the use of percent differences was adopted rather than total differences for the vertical axis of the plot.37

The bias of one measurement method in relation to the other is calculated by averaging the percent differences across all concentrations. The LoAs are calculated to contain roughly 95% of the percent differences, using the standard sample deviation as an estimate for the population standard deviation.

Dividing the LoA by the square root of the number of samples produces an estimated 95% CI on the bias. If the value of zero falls in the bounds of this CI, the bias was determined as being inconsequential.

Application of high-throughput T cell counting

The T cell culture employed in this experiment was gifted by a current collaborator. Cells were acquired from eight T-25 culture flasks and consolidated in a 50-mL conical tube at a total volume of 40 mL.

The cells were then centrifuged at 1200 RPM for 10 minutes and resuspended in 1.5 mL of RPMI media to a concentration around 3.5 x 107 cells/mL. Fifteen dilutions were produced in microtubes by conducting 1:2 serial dilutions with RPMI media. These serial dilutions were then rendered down to 1:16,384 DF of the original sample.

An equivalent volume of AO/PI (500 μL) was introduced to each dilution to produce a 1:1 staining of the cell samples (500 μL) in the Cellaca™ plate mixing wells (12 x 2 format). Four replicates of 50 μL of the stained cell samples were transferred directly into loading wells for each dilution for an overall number of 60 counting chambers on 3 Cellaca™ plates.

Before conducting any dilutions, sample transfers, or sub-sampling to produce replicates, each tube was vortexed gently to limit cell settling and guarantee uniform distribution. Each plate was placed into the Cellaca™ MX high-throughput cell counter for image acquisition and analysis utilizing the built-in AO/PI feasibility assay with default counting parameters.

Results & discussion

Variation and precision considerations in cell counting methods

In line with the ISO Cell Counting Standards guidelines, assessing a cell counting method is characteristic of the entire process, including cell type, cell suspension, aliquoting, mixing, sample preparation, consumables, reagents, parameters, instruments, algorithms, and every single step; from the original cell culture flask or sample tube to the final data on the screen.

Any variation at a single point may result in considerable performance differences, and the new process may be thought to be a unique cell counting method to be evaluated separately. It is, therefore, crucial to determine the intended purpose and scope of a cell counting method evaluation.

For instance, if the purpose is to establish the expected precision, then the evaluation experiments should include all variation sources to be anticipated for the cell counting method (e.g., multiple operators, instruments, days, reagent lots, etc.).

In this context, results should only be seen as examples of cell counting method evaluations as opposed to exact method performance predictions. Precision on several levels was taken into account, including count-to-count, plate-to-plate, instrument-to-instrument, and system-wide, which will be described briefly for each level of precision.

Cell counting typically involves analyzing a sample of the suspension from a much larger volume; the inherent variability in the number of cells captured in each analyzed volume is a source of random variation among replicate counts.

For a standard cell counting process, sampling variation typically leads to higher CVs among replicate counts for lower-concentration cell suspensions. This random error, also known as Poisson noise or shot noise, is incorporated in the count-to-count precision.

Other sources of variation can exist with minor variations in counting chamber dimensions, which leads to variations in the analyzed sample volume. The cell counting assay precision for count-to-count or intra-plate precision can be detailed as the amount of variation that a user can anticipate for the same cell sample counted on a single instrument with a single plate.

Minimal differences in the cell counting chamber can also occur between the Cellaca™ plates, which is another source of variation for experiments involving multiple plates. The cell counting assay precision for plate-to-plate or inter-plate precision is calculated by averaging all other experimental variables to establish plate consistency.

Likewise, measuring the same sample on two different instruments can produce slightly different average results. The cell counting assay precision for instrument-to-instrument is the expected variation when a single cell sample is measured across multiple instruments.

The different sources of error do not add linearly; rather, they partially cancel each other out, producing an overall CV that is less than the sum.

System-wide cell counting precision is the variation to be expected when repeated measurements are comprised of a cell sample using a random chamber in a random plate on a random instrument, which is characterized by including all sources of variation.

The system-wide precision offers an indication of the confidence levels a user can have that a cell counting result is close to the ‘true' value, although accuracy cannot be determined because of the lack of a true live-cell reference standard.20,21,38

One of the work’s most significant strengths is the high number of Cellaca™ MX instruments assessed over a prolonged time period, demonstrating not only high repeatability but also high intermediate precision as described in the ICH Q2 (R1) guidance document.39

It is also crucial to note that the precision level needed for cell counting is predicated on the ‘fit-for-purpose’ principle described in the ISO Cell Counting Standard Part 1, which is reliant on the intended purpose of the cell counting results.

Cell counting variation and precision characterization in BF

The capabilities of the Cellaca™ MX in relation to BF cell analysis were established by counting UV-cured beads and TB-stained CHO cells on several instruments. It is crucial to note that a small sample of instruments may be more or less consistent than the general instrument population.

Therefore, as many instruments as possible were included in this study to best determine the cell counting assay precision on an instrument-to-instrument level. Such experiments can be complicated by sample instability; cell suspensions can deteriorate over the course of a prolonged experiment, and microbead solutions may even evaporate from counting chambers.

To circumvent these problems, stable reference plates, consisting of microbeads locked in optically transparent UV-curable polymers, were created. Bead counting was conducted using 32 Cellaca™ MX instruments manufactured over 10 months.

Experimental design and results comparing 32 Cellaca MX instruments for BF counting of 5-μm microbeads. (a) Experiment design workflow diagram: (1) Two concentrations of microbeads are suspended in a UV-curable transparent polymer. Each concentration was loaded into all 48 counting chambers of 2 CellacaTM Plates. The plates were then exposed to UV light to lock the beads into place. (2) The plates were then imaged on 32 Cellaca™ MX instruments, and the beads in each chamber were counted. (b) Comparison of the bead concentration measured for one plate of each concentration by the 32 Cellaca™ MX instruments (n = 24 each). Error bars are 1 SD

Figure 1. Experimental design and results comparing 32 Cellaca MX instruments for BF counting of 5-μm microbeads. (a) Experiment design workflow diagram: (1) Two concentrations of microbeads are suspended in a UV-curable transparent polymer. Each concentration was loaded into all 48 counting chambers of 2 CellacaTM Plates. The plates were then exposed to UV light to lock the beads into place. (2) The plates were then imaged on 32 Cellaca™ MX instruments, and the beads in each chamber were counted. (b) Comparison of the bead concentration measured for one plate of each concentration by the 32 Cellaca™ MX instruments (n = 24 each). Error bars are 1 SD. Image Credit: Nexcelom Bioscience LLC

Table 1. Bead counting consistency and precision characterization results for BF applications. Source: Nexcelom Bioscience LLC

Precision level Beads total conc. (CV)
4.9 x 106 beads/mL 1.1 x 106 beads/mL
Analysis-to-analysis 0.0% 0.0%
Scan-to-scan 1.0% 0.5%
*Count-to-count 4.2% 5.6%
Plate-to-plate 1.6% 3.3%
Instrument-to-instrument 3.6% 4.9%
*System-wide 5.7% 7.6%

*The count-to-count and system-wide variation include random error and sample preparation error.

The counting results are displayed in Figure 1 and summarized in Table 1, demonstrating a count-to-count CV of 4.2%, a plate-to-plate CV of 1.6%, an instrument-to-instrument CV of 3.6%, and overall system-wide precision of 5.7% at 4.9 x 106 beads/mL.

One of the most frequently used BF cell counting applications in bioprocessing and cell line development is the TB exclusion assay for CHO and HEK293 cells. It is used to evaluate target cell feasibility based on cell membrane permeability, where live cells act as objects with bright centers and dead cells are dark and diffuse.

Experimental design and results comparing five Cellaca™ MX instruments for TB counting of CHO cells. (a) Experimental design workflow diagram: (1) Equal volumes (600 μL) of 0.2% TB solution and CHO cell suspension were prepared and (2) mixed thoroughly by pipetting up and down. (3) Two Cellaca™ plates were loaded with the mixture (10 counting chambers each). (4) Both plates were imaged on five Cellaca™ MX instruments, and the cells in each chamber were counted. (b) Live cell concentration results for the five instruments. All 20 counts from each instrument are summarized, with error bars of 1 SD. (c) Viability measurement results from the five instruments. Error bars represent 1 SD.

Figure 2. Experimental design and results comparing five Cellaca™ MX instruments for TB counting of CHO cells. (a) Experimental design workflow diagram: (1) Equal volumes (600 μL) of 0.2% TB solution and CHO cell suspension were prepared and (2) mixed thoroughly by pipetting up and down. (3) Two Cellaca™ plates were loaded with the mixture (10 counting chambers each). (4) Both plates were imaged on five Cellaca™ MX instruments, and the cells in each chamber were counted. (b) Live-cell concentration results for the five instruments. All 20 counts from each instrument are summarized, with error bars of 1 SD. (c) Viability measurement results from the five instruments. Error bars represent 1 SD. Image Credit: Nexcelom Bioscience LLC

Table 2. CHO cell counting and viability precision characterization results in BF. Source: Nexcelom Bioscience LLC

Precision
level
CHO total
conc. (CV)
CHO live
conc. (CV)
CHO viability
(CV)
*Count-to-count 5.5% 5.7% 0.9%
Plate-to-plate 3.4% 3.2% 0.3%
Instrument-to-instrument 1.7% 2.0% 0.7%
*System-wide 7.0% 7.3% 1.3%

*The count-to-count and system-wide variation include random error and sample preparation error.

TB assay precision was evaluated using the Cellaca™ MX by measuring the same TB-stained CHO-S cell sample on five instruments. The live-cell concentration and feasibility counting and precision results are displayed in Figure 2 and Table 2.

The count-to-count CV results were 5.5%, 5.7%, and 0.9% for total cell concentration, live-cell concentration, and feasibility, respectively. The plate-to-plate CV results measured from two plates were 3.4%, 3.2%, and 0.3% for total cell concentration, live-cell concentration, and feasibility, respectively.

The respective instrument-to-instrument CV results were 1.7%, 2.0% and 0.7% for total cell concentration, live cell concentration and feasibility. Overall system-wide precision values across the 20 chambers, 2 plates, and 5 instruments were 7.0%, 7.3%, and 1.3% for total cell concentration, live-cell concentration, and feasibility, respectively.

Cell counting variation and precision characterization in FL

A similar procedure was used to evaluate Cellaca™ MX performance for FL applications. Over a period of 6 months, stable fluorescent bead reference plates were produced in six concentrations that ranged from 1.5 x 105 to 5 x 106 beads/mL and measured in 13 Cellaca™ MX instruments.

Experimental design and results comparing 13 Cellaca™ MX instruments for FL-based counting of 7.5-μm microbeads. (a) Experimental design workflow diagram. (1) Microbeads were suspended in a UV-curable transparent polymer in a serial dilution of 6 concentrations. Each concentration was loaded into 8 Cellaca™ counting chambers, 4 in each of 2 plates. (2) After the polymer was cured, the plates were imaged on 13 Cellaca™ MX instruments in FL mode, and the beads counted. (b) Combined data for both plates (8 counts/concentration) for all 13 instruments. Error bars are 1 SD.

Figure 3. Experimental design and results comparing 13 Cellaca™ MX instruments for FL-based counting of 7.5-μm microbeads. (a) Experimental design workflow diagram. (1) Microbeads were suspended in a UV-curable transparent polymer in a serial dilution of 6 concentrations. Each concentration was loaded into 8 Cellaca™ counting chambers, 4 in each of 2 plates. (2) After the polymer was cured, the plates were imaged on 13 Cellaca™ MX instruments in FL mode, and the beads counted. (b) Combined data for both plates (8 counts/concentration) for all 13 instruments. Error bars are 1 SD. Image Credit: Nexcelom Bioscience LLC

Table 3. Bead counting consistency and precision characterization results for FL applications. Source: Nexcelom Bioscience LLC

Precision level Measured precision (CV) by concentration (beads/mL)
5.0 x 106 2.7 x 106 1.4 x 106 0.7 x 106 0.4 x 106 0.2 x 106
Analysis-to-analysis 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Scan-to-scan 0.28% 0.36% 0.42% 0.47% 0.75% 0.8%
*Count-to-count 3.1% 3.6% 2.3% 4.3% 4.9% 8.3%
Plate-to-plate 0.8% 1.8% 0.3% 0.1% 0.3% 4.2%
Instrument-to-instrument 1.7% 1.0% 0.8% 0.7% 0.7% 1.7%
*System-wide precision 3.2% 3.5% 2.2% 3.8% 4.3% 7.8%

*The count-to-count and system-wide variation include random error and sample preparation error.

The cell counting and precision results are displayed in Figure 3 and Table 3. The count-to-count, plate-to-plate, instrument-to-instrument, and system-wide precision results were determined for each bead concentration.

The concentration series offers an example of increasing CVs for lower numbers of counted objects, which is an effect of random counting variation as detailed in previous sections. It only impacts the count-to-count precision and later the system-wide precision.

Experimental design and results comparing 5 Cellaca™ MX instruments for FL-based counting of Jurkat cells. (a) Experimental design workflow diagram: (1) Equal volumes (600 μL) of AO/PI dye mixture and Jurkat cell suspension were prepared and (2) mixed thoroughly by pipetting up and down. (3) Two Cellaca™ plates were loaded with the resulting mixture (10 counting chambers each). (4) Both plates were imaged on 5 Cellaca™ MX instruments, and the cells in each chamber were counted. (b) Live cell concentration results for the 5 instruments. All 20 counts from each instrument are summarized, with error bars of 1 SD. (c) Viability measurement results from the 5 instruments. Error bars represent 1 SD.

Figure 4. Experimental design and results comparing 5 Cellaca™ MX instruments for FL-based counting of Jurkat cells. (a) Experimental design workflow diagram: (1) Equal volumes (600 μL) of AO/PI dye mixture and Jurkat cell suspension were prepared and (2) mixed thoroughly by pipetting up and down. (3) Two Cellaca™ plates were loaded with the resulting mixture (10 counting chambers each). (4) Both plates were imaged on 5 Cellaca™ MX instruments, and the cells in each chamber were counted. (b) Live cell concentration results for the 5 instruments. All 20 counts from each instrument are summarized, with error bars of 1 SD. (c) Viability measurement results from the 5 instruments. Error bars represent 1 SD. Image Credit: Nexcelom Bioscience LLC

Table 4. Jurkat cell counting and viability precision characterization results in FL. Source: Nexcelom Bioscience LLC

Precision
level
Jurkat total
conc. (CV)
Jurkat live
conc. (CV)
Jurkat viability
(CV)
*Count-to-count 5.8% 5.9% 3.8%
Plate-to-plate 1.7% 1.7% 0.9%
Instrument-to-instrument 3.4% 2.2% 1.8%
*System-wide 7.0% 6.6% 4.4%

*The count-to-count and system-wide variation include random error and sample preparation error.

To further establish the FL mode on the Cellaca™ MX, the concentration and feasibility of a sample of Jurkat cells stained with AO/PI were assessed on the same five Cellaca™ MX instruments used for the CHO TB experiment. The live-cell concentration and feasibility results are drawn up in Figure 4, and the precision results are presented in Table 4.

The AO/PI assay has shown higher cell counting quality for primary cell samples (containing RBC residues, platelets, and debris) when contrasted against the TB assay. 8,13,14,29 Critcally, Mascotti et al. (2000) demonstrated that incubating cells with TB for a prolonged period of time can be harmful to the cells and lead to lower feasibility, while AO/PI assay did not.

Since Cellaca™ MX can quickly measure cell concentration and feasibility for 24 samples in less than 4 and 8 minutes, respectively, both the TB and AO/PI assays and AO/PI assay will not be influenced by the counting time.

Cell counting quality characterization using ISO Cell Counting Standard Part 2

Accuracy is one of the key parameters as described in the ICH Q2 (R1) guidance document.39–41 While precision details how well a measurement method agrees with itself, accuracy quantifies how well it is in accordance with a known reference standard.

Accuracy can be hard to establish for cell counting due to the lack of stable live-cell reference standards. Cells in a sample are continuously changing and are always sub-sampled from a larger volume, and both of these factors insert uncertainties during counting.

Other biological questions also present challenges to the meaning of cell counting accuracy, such as:

  1. What is a live cell?
  2. What cells are dividing, dying or dead?
  3. How would life and death be defined when considering attributes such as compromised membranes, enzymatic activity, and initiation of apoptosis?

Instead of defining the accuracy of a cell counting method, the guidance from ISO 20391-2 can be applied to contrast the proportionality of multiple cell counting methods using a dilution series design.

While the ‘actual’ live cell concentration of a sample may not be known, it is reasonable to assume that doubling the sample volume by the relevant dilution should reduce the live cell concentration by half.

The quality of a cell counting method can therefore be associated with its capacity to generate a number that is inversely proportional to the dilution of the sample. Consequently, it is useful to follow guidance documents to assess the quality of the cell counting method for the intended purposes of the downstream assays.

For a legitimate comparison between methods using such an evaluation, the two methods should be evaluated simultaneously using the same cell samples. The ISO Cell Counting Standards method characterization introduces three primary quality indicators: R2, CV, and PI, as implemented by the National Institute of Standards and Technology.20

Although different metrics can be used as a proportionality index, the PI was determined using the smoothed scaled absolute value of residuals.20

It is important to note, however, that this metric is not normalized to the number of DFs or biological replicates in the experimental design, and its value should not be equated across experiments.

Determining proportionality is crucial when assessing the quality of a cell counting method. Conversely, it is just as important to characterize how the two results of the two methods agree.

For this purpose, Bland-Altman comparative analyses were applied to visualize variations between two methods across a range of measurement values. Bland-Altman plots (mean-difference plots) display variations between results from two methods with respect to their average values.

Since variance in cell counting is relative to cell sample concentration, the percent difference, rather than absolute difference, was employed to acquire the same approximate variance across the concentration range.

Bland-Altman analysis returns a value for the bias between two methods with its accompanying CI, as well as the LoAs, demonstrating the range of differences to be anticipated for a single measurement.

Evaluation of cell counting methods using ISO guidelines and comparisons with two cell counting methods using Bland-Altman analysis. CHO cells were stained with AO, and 12 replicates were prepared in Cellaca plates for each of 6 DFs. (a) The plated cells were counted using both the Cellaca™ MX and Celigo® imaging cytometer, and the R2, PI, and CVs were calculated for each method. (b) Bland-Altman plot demonstrating a small bias between the two cell counting methods using the concentration data. (c) The R2, PI, and CVs calculated for Jurkat cells. (d) Bland-Altman plot comparing the two cell counting methods using Jurkat cells.

Figure 5. Evaluation of cell counting methods using ISO guidelines and comparisons with two cell counting methods using Bland-Altman analysis. CHO cells were stained with AO, and 12 replicates were prepared in Cellaca plates for each of 6 DFs. (a) The plated cells were counted using both the Cellaca™ MX and Celigo® imaging cytometer, and the R2, PI, and CVs were calculated for each method. (b) Bland-Altman plot demonstrates a small bias between the two cell counting methods using the concentration data. (c) The R2, PI, and CVs were calculated for Jurkat cells. (d) Bland-Altman plot comparing the two cell counting methods using Jurkat cells. Image Credit: Nexcelom Bioscience LLC

In this experiment, ISO Cell Counting Standard Part 2 guidance was followed to determine and compare cell counting quality between the Cellaca™ MX and Celigo® for CHO and Jurkat cells stained with AO. Both instruments were utilized for the measuring of samples from the same Cellaca™ plates for direct comparison. The results are displayed in Figure 5.

The CHO cell concentration measurement was in the range of 5 x 105 to 6 x 106 cells/ mL. The Cellaca™ MX results demonstrated values of 2.7–7.0% for CV, 0.998 for R2  and 0.44 for PI.

The results from Celigo® were 2.7–6.4% for CV, 0.996 for R2, and 0.35 for PI. The Bland-Altman comparison produced a bias of -5.1% ± 0.9% (95% CI) within the two methods (Celigo® counting higher), and the LoAs spanned -12.5% to 2.3%.

For Jurkat cells, the Cellaca™ MX results exhibited values of 2.2–7.5% for CV, 0.997 for R2, and 0.44 for PI. The Celigo® produced 1.8–7.6% for CV, 0.997 for R2, and 0.42 for PI. Bland-Altman analysis generated a bias of 1.5% ± 0.6% across the two methods (Cellaca™ MX counting higher), with an LoA range of -3.9% to 6.9%.

For evaluation of cell counting methods using the same experimental design, a lower PI is deemed to be more optimal or proportional. The Celigo® produced slightly more proportional results than the Cellaca™ MX, although the two PI values are not statistically different.

Across the experiments, the R2 and CV values were comparable for both instruments. There was a minor but statistically significant bias between the two methods, shown by the value of zero falling outside the 95% CI. Curiously, the Cellaca™ MX was inclined to count slightly higher than the Celigo® for lower concentrations.

Direct comparison to other cell counting methods

Characterizing the proportionality of a cell counting method is a robust way of addressing the lack of live-cell reference standards, but a comparison to an independent method is typically employed in practice.

Here comparisons of counts acquired using the Cellaca™ MX in relation to those from the manual hemocytometer method, the Cellometer® Auto2000, and the Cellometer® Vision were carried out.

Bright field beads and CHO cells were employed to compare Cellaca™ MX and hemocytometer. Three independent samples of 5 μm beads were assessed, each with 40 counts for the manual counting method and 144 counts (6 plates) for the Cellaca™ MX method.

Comparison of the Cellaca™ MX to other cell counting methods. (a) Comparison of 144 Cellaca™ MX counts to 40 manual counts for each of three 5-μm bead suspensions. (b) Comparison results for CHO cell counting between Cellaca™ MX and hemocytometer for 3 samples. (c) Live cell concentration measurements from two Cellaca™ MX instruments, two Cellometer™ Vision instruments, and two Cellometer™ Auto2000 instruments for Jurkat suspensions prepared in three concentrations. (d) Viability data for the same comparison shown in panel B. Viability was determined by AO/PI staining and dual-fluorescence imaging by all six instruments

Figure 6. Comparison of the Cellaca™ MX to other cell counting methods. (a) Comparison of 144 Cellaca™ MX counts to 40 manual counts for each of three 5-μm bead suspensions. (b) Comparison results for CHO cell counting between Cellaca™ MX and hemocytometer for 3 samples. (c) Live cell concentration measurements from two Cellaca™ MX instruments, two Cellometer™ Vision instruments, and two Cellometer™ Auto2000 instruments for Jurkat suspensions were prepared in three concentrations. (d) Viability data for the same comparison shown in panel B. Viability was determined by AO/PI staining and dual-fluorescence imaging by all six instruments. Image Credit: Nexcelom Bioscience LLC

The values for the Cellaca™ MX were ~1.2% lower, ~2.35% lower, and ~0.71% higher than the manual counts for the three samples. The CVs for the Cellaca™ MX were 3.7%, 3.6%, and 3.9%, whereas the hemocytometer CVs were 4.5%, 4.3%, and 4.3% (Figure 6A).

It is crucial to note that beads were selected for comparison to hemocytometer to guarantee the stability of samples over the prolonged manual counting period.

The CHO cell samples were measured on both Cellaca™ MX (n = 24) and hemocytometer (n = 4), which produced comparable cell counting results. The values for the Cellaca™ MX were ~5.3% higher, ~0.5% higher, and ~12.1% lower than the manual counts for the 2 high and 1 low concentration samples.

The concentration CVs for Cellaca™ MX were 4.2%, 6.4% and 15.7%, whereas the hemocytometer CVs were 5.8%, 6.8% and 23.2% (Figure 6B). Comparison of concentration and feasibility measurements were performed for two Cellaca™ MX, Cellometer® Vision, and Cellometer® Auto2000 instruments employing AO/PI-stained Jurkat cells in three concentrations.

All six instruments generated reasonably consistent results with CVs ranging from 1.4 to 8.0% for cell concentration and feasibility (Figure 6C & D). The live-cell concentrations and feasibility results demonstrated comparable results and CVs for each sample and instrument (Supplementary Table 2).

LoD and LoQ characterization using T cell dilution series

A conventional serial dilution experiment can swiftly characterize the linear range of a cell counting method with multiple DFs. The dilution series may achieve low concentrations sufficient enough to establish the method’s limit of detection (LoD) and limit of quantification (LoQ).

As the results are unique to a specific method, each cell type, assay or instrument should be independently evaluated. T cells were subjected to serial dilution and stained with AO/PI and then counted on the Cellaca™ MX.

Live cell concentration for a 15-point 2X dilution series of T cells stained with AO/PI and imaged using dual-fluorescence mode on the Cellaca™ MX. The dilution series extended well beyond the instrument’s manufacturer-suggested concentration range, but good linearity extended down to the point of only ~10 cells visible in each counting chamber

Figure 7. Live cell concentration for a 15-point 2X dilution series of T cells stained with AO/PI and imaged using dual-fluorescence mode on the Cellaca™ MX. The dilution series extended well beyond the instrument’s manufacturer-suggested concentration range, but good linearity extended down to the point of only ~10 cells visible in each counting chamber. Image Credit: Nexcelom Bioscience LLC

The 15-point dilution series in Figure 7 includes concentrations from 2.3 x 103 to 2.8 x 107 cells/mL, exceeding the range as specified by the manufacturer. At the low end, only ~10 cells were observed in each image, implying a lower concentration limit for single-image counts on the instrument for this assay.

The LoQ and LoD concentrations determined were around 5.5 x 104 cells/mL and 2.3 x 103 cells/mL, respectively.

Conclusion

Reducing the cell counting bottleneck is a key factor when attempting to streamline preclinical and clinical research and cell and biologics bioprocessing. This work demonstrated the application capacity of the Cellaca™ MX high-throughput cell counter.

The system can rapidly count cells in BF and FL directly between 1 and 3 minutes, respectively. Cell counting performance was determined for BF and FL applications using beads and CHO-S and Jurkat cells.

We analyzed and quantified the Cellaca™ MX platform’s abilities, including count-to-count, plate-to-plate, and instrument-to-instrument precision, displaying overall variation series experiment while assessing the R2, CV and PI in accordance with the guidance of ISO Cell Counting Standard Part 2.

The Cellaca™ MX can be compared to a conventional hemocytometer and single-sample-based automatic cell counters.

Finally, we investigated the sensitivity of the instrument in relation to high and low-concentration T cell samples and visualized result linearity across 4 logs of cell concentration.

It is recommended that comparable characterization experiments in accordance with the guidance of ISO Cell Counting Standards should be conducted to comprehensively evaluate other cell counting methods.

The Cellaca™ MX high-throughput cell counter is a unique, advanced system that can provide reliable data rapidly with the kind of exceptional precision required for cell and immunotherapy applications.

References

  1. Seimetz D, Heller K, Richter J. Approval of First CAR-Ts: Have we Solved all Hurdles for ATMPs? Cell Medicine 2019; 11: 1–16.
  2. Vaddepally RK, Kharel P, Pandey R, Garje R, Chandra AB. Review of Indications of FDA-Approved Immune Checkpoint Inhibitors per NCCN Guidelines with the Level of Evidence. Cancers 2020; 12(3): 738.
  3. Bartheld CSv, Bahney J, Herculano-Houzel S. The search for true numbers of neurons and glial cells in the human brain: A review of 150 years of cell counting. J. Comp. Neurol. 2016; 524(18): 3865–95.
  4. Briggs C. Quality counts: new parameters in blood cell counting. Int. J. Lab. Hematol. 2009; 31(3): 277–97.
  5. Väyrynen JP, Vornanen JO, Sajanti S, Böhm JP, Tuomisto A, Mäkinen MJ. An improved image analysis method for cell counting lends credibility to the prognostic significance of T cells in colorectal cancer. Virchows Archive 2012; 460: 455–65.
  6. Chan LLY, Lai N, Wang E, Smith T, Yang X, Lin B. A rapid detection method for apoptosis and necrosis measurement using the Cellometer imaging cytometry. Apoptosis 2011; 16(12): 1295–303.
  7. Berkes CA, Chan LLY, Wilkinson A, Paradis B. Rapid quantification of pathogenic fungi by Cellometer image-based cytometry. J. Microbiol. Methods 2012: 468–76.
  8. Chan LL, Wilkinson AR, Paradis BD, Lai N. Rapid Image-based Cytometry for Comparison of Fluorescent Viability Staining Methods. J. Fluorescence 2012; 22(5): 1301–11.
  9. Cadena-Herrera D, Lara JEE-D, Ramírez-Ibañez ND et al. Validation of three viable-cell counting methods: Manual, semi-automated, and automated. Biotechnol. Rep. 2009; 7: 9–16.
  10. Huang L-C, Lin W, Yagami M et al. Validation of cell density and viability assays using Cedex automated cell counter. Biologicals 2010; 38(3): 393–400.
  11. Louis KS, Siegel AC. Cell Viability Analysis Using Trypan Blue: Manual and Automated Methods. In: Stoddart M (Ed.). Mammalian Cell Viability. Methods in Molecular Biology (Methods and Protocols). Volume 740: Humana Press; 2011.
  12. Chen A, Leith M, Tu R, Tahim G, Sudra A, Bhargava S. Effects of diluents on cell culture viability measured by automated cell counter. PLOS One 2017; 12(3) :e0173375.
  13. Mascotti K, McCullough J, Burger SR. HPC viability measurement: trypan blue versus acridine orange and propidium iodide. Transfusion 2000; 40: 693–6.
  14. Chan LLY, Laverty DJ, Smith T, Nejad P, Hei H, Gandhi R, Kuksin D, Qiu J. Accurate measurement of peripheral blood mononuclear cell concentration using image cytometry to eliminate RBC-induced counting error. J. Immunol. Methods 2013; 388(1–2): 25–32.
  15. Chan LL-Y, Kuksin D, Laverty DJ, Saldi S, Qiu J. Morphological observation and analysis using automated image cytometry for the comparison of trypan blue and fluorescence-based viability detection method. Cytotech. 2015; 67(3): 461–73.
  16. Chan LL-Y, Rice WL, Qiu J. Observation and quantification of the morphological effect of trypan blue rupturing dead or dying cells. PLoS One 2020; 15(1): e0227950.
  17. Hanamsagar R, Reizis T, Chamberlain M et al. An optimized workfow for single cell transcriptomics and repertoire profling of purifed lymphocytes from clinical samples. Sci. Rep. 2020; 10(2219).
  18. Zhu L, Yang P, Zhao Y et al. Single-Cell Sequencing of Peripheral Mononuclear Cells Reveals Distinct Immune Response Landscapes of COVID-19 and Influenza Patients. Immunity 2020; 53: 685–96.
  19. Biotechnology – Cell counting – Part 2: Experimental design and statistical analysis to quantify counting method performance. International Organization for Standardization 2019.
  20. Sarkar S, Lund SP, Vyzasatya R et al. Evaluating the quality of a cell counting measurement process via a dilution series experimental design. Cytotherapy 2017; 19(12): 1509–21.
  21. Sarkar S, Pierce L, Lin-Gibson S, Lund SP. Standards Landscape in Cell Counting: Implications for Cell & Gene Therapy. Cell Gene Ther. Ins. 2019; 5(1): 117–31.
  22. Xu Y, Chan LL-Y, Chen S et al. Optimization of UC-MSCs cold-chain storage by minimizing temperature fluctuations using an automatic cryopreservation system. Cryobiology 2021; 99: 131–9.
  23. Bieberich F, Vazquez-Lombardi R, Yermanos A et al. A single-cell atlas of lymphocyte adaptive immune repertoires and transcriptomes reveals age-related differences in convalescent COVID-19 patients. bioRxiv 2021.
  24. Sun T, Zhang K, Pangeni RP et al. Histone Methyltransferase G9a Promotes Invasion of Non-small Cell Lung Cancer Through Enhancing Focaladhesionkinase Activationvia NF-KB Signaling Pathway. Research Square 2021.
  25. Zeijlemaker W, Kelder A, Cloos J, Schuurhuis GJ. Immunophenotypic Detection of Measurable Residual (Stem Cell) Disease Using LAIP Approach in Acute Myeloid Leukemia. Curr. Protoc. Cytom. 2019; 91(e66).
  26. Chan LL, Zhong XM, Qiu J, Li PY, Lin B. Cellometer Vision as an Alternative to Flow Cytometry for Cell Cycle Analysis, Mitochondrial Potential, and Immunophenotyping. Cytometry Part A 2011; 79A(7): 507–17.
  27. Chan LLY, Shen D, Wilkinson AR et al. A novel image-based cytometry method for autophagy detection in living cells. Autophagy 2012; 8(9): 1371–82.
  28. Chan LL-Y, Huang J, Hagiwara Y, Aguila L, Rowe D. Discriminating Multiplexed GFP Reporters in Primary Articular Chondrocyte Cultures Using Image Cytometry. J. Fluoresc. 2014; 24(4): 1041–53.
  29. Chan LL-Y, Smith T, Kumph KA et al. A high-throughput AO/PI-based cell concentration and viability detection method using the Celigo image cytometry. Cytotechnology 2016; 68(5): 2015–25.
  30. Zigon ES, Purseglove SM, Toxavidis V, Rice W, Tigges J, Chan LL-Y. A rapid single cell sorting verification method using plate-based image cytometry. Cytometry Part A 2018; 93(10): 1060–5.
  31. Magnotti EL, Chan LL-Y, Zhu Q, Marasco WA. A high-throughput chemotaxis detection method for CCR4+ T cell migration inhibition using image cytometry. J. Immunol. Methods 2020; 112747.
  32. Maldini CR, Love AC, Tosh KW et al. Characterization of CAR T cell expansion and cytotoxic potential during Ex Vivo manufacturing using image-based cytometry. J. Immunol. Meth. 2020; 484–485: 112830.
  33. Wang Y, Chan LL-Y, Grimaud M, Fayed A, Zhu Q, Marasco WA. High-Throughput Image Cytometry Detection Method for CAR-T Transduction, Cell Proliferation, and Cytotoxicity Assays. Cytometry Part A 2020:
  34. Altman DG, Bland JM. Measurement in Medicine: The Analysis of Method Comparison Studies. The Statistician 1983; 32(3): 307–17.
  35. Bland JM, Altman DG. Statistical Methods For Assessing Agreement Between Two Methods of Clinical Measurement. The Lancet 1986; 327(8476): 307–10.
  36. Carkeet A. A Review of the Use of Confidence Intervals for Bland-Altman Limits of Agreement in Optometry and Vision Science. Optom. Vis. Sci. 2020; 97(1): 3–8.
  37. Tholudur A, Giron L, Alam K et al. Comparing Automated and Manual Cell Counts for Cell Culture Applications. BioProcess Int. 2006.
  38. Biotechnology — Cell counting — Part 1: General guidance on cell counting methods. International Organization for Standardization 2018.
  39. Validation of Analytical Procedures: Text and Methodology Q2(R1). In: Guideline IHT; 2005.
  40. Guidance for Industry: Process Validation: General Principles and Practices. In: Services USDoHaH, Administration FaD, (CDER) CfDEaR, (CBER) CfBEaR, (CVM) CfVM; 2011.
  41. Guidance for Industry: Analytical Procedures and Methods Validation for Drugs and Biologics. In: Services USDoHaH, Administration FaD, (CDER) CfDEaR, (CBER) CfBEaR; 2015.

About Nexcelom Bioscience

Nexcelom Bioscience is a developer and marketer of image cytometry products for cell analysis in life science and biomedical research. Products range from cell viability counters (Cellometer) to high throughput microwell image cytometry workstations (Celigo), used in thousands of research laboratories in academic institutes, and pharmaceutical and biotech companies. The company contributes to the life science industry through innovation and expertise in the science of cell counting.

The product family includes instruments, consumables, and reagents. Nexcelom customers engage in a wide variety of research, such as cancer research, immunology, stem cell research, and neuroscience. Nexcelom offers different Cellometer models to count and analyze cell lines and primary cells, through brightfield and fluorescence imaging modes. In addition, Celigo is a powerful high image quality, high-throughput image cytometry system for adherent and suspension cells in microwell plates.

Nexcelom Bioscience is a fast-growing company in a huge market. With its headquarters and manufacturing facilities in the Boston area, the company currently has over 80 global employees, who are fast-paced, customer-centric, helpful to colleagues and customers, and passionate about their impact in life science.


Sponsored Content Policy: News-Medical.net publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of News-Medical.Net which is to educate and inform site visitors interested in medical research, science, medical devices, and treatments.

Last updated: Nov 2, 2021 at 6:57 AM

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Nexcelom Bioscience LLC. (2021, November 02). The characterization and application of the Cellaca™ MX high-throughput cell counter. News-Medical. Retrieved on December 08, 2021 from https://www.news-medical.net/whitepaper/20211007/The-characterization-and-application-of-the-Cellacae284a2-MX-high-throughput-cell-counter.aspx.

  • MLA

    Nexcelom Bioscience LLC. "The characterization and application of the Cellaca™ MX high-throughput cell counter". News-Medical. 08 December 2021. <https://www.news-medical.net/whitepaper/20211007/The-characterization-and-application-of-the-Cellacae284a2-MX-high-throughput-cell-counter.aspx>.

  • Chicago

    Nexcelom Bioscience LLC. "The characterization and application of the Cellaca™ MX high-throughput cell counter". News-Medical. https://www.news-medical.net/whitepaper/20211007/The-characterization-and-application-of-the-Cellacae284a2-MX-high-throughput-cell-counter.aspx. (accessed December 08, 2021).

  • Harvard

    Nexcelom Bioscience LLC. 2021. The characterization and application of the Cellaca™ MX high-throughput cell counter. News-Medical, viewed 08 December 2021, https://www.news-medical.net/whitepaper/20211007/The-characterization-and-application-of-the-Cellacae284a2-MX-high-throughput-cell-counter.aspx.

Other White Papers by this Supplier