How fixation can mask epitopes: When cells are fixed, the proteins get stuck together in different ways, making it harder for antibodies to reach and bind to the spots (epitopes) they normally would.
In both mass and flow cytometry, the goal is to measure specific immune populations. To measure these populations, we need to detect specific targets, often proteins, within individual cells. This detection relies on labeled antibodies binding to these targets, making them visible to the instrument. Antibodies bind to specific regions on a protein, known as epitopes. However, when samples are chemically fixed to stabilize them for storage, the fixation creates bonds (or methylene bridges) between nearby amino acids in a protein, a process known as cross-linking. These bonds, illustrated in red below, stabilize the protein but can block access to certain epitopes. As a result, some antibodies can no longer recognize their target because their binding region is now “masked.”
Fortunately, proteins have multiple epitopes, and other antibodies may bind to regions unaffected by fixation. These antibodies remain capable of detecting the protein in both fixed and unfixed states. Alternatively, fixation methods without formaldehyde such as methanol-based fixation can be utilized.
When working with fixed samples, it is essential to rigorously test each antibody in your panel to ensure its performance is not compromised by fixation. Proper testing helps ensure accurate and reliable detection of your targets in cytometry assays.
You might have heard of an “immune reset” for Chimeric Antigen Receptor (CAR) therapies in autoimmune diseases. But what does it mean, exactly?
Here’s the 2024 definition from Georg Schett et al’s paper “”Advancements and challenges in CAR T cell therapy in autoimmune diseases”:
“…deep depletion of B cells, including autoreactive B cell clones, could restore normal immune function, referred to as an immune reset.”
Autoimmune diseases, such as systemic lupus erythematosus (SLE), represent some of the hardest conditions to treat. These diseases need lifelong medication with limited success in achieving full remission. But CAR-T is changing that.
In 2022, Schett et al posted this incredible result: “Remission of SLE according to [disease remission] criteria was achieved in all five patients after 3 months and the median (range) Systemic Lupus Erythematosus Disease Activity Index score after 3 months was 0…Drug-free remission was maintained during longer follow-up (median (range) of 8 (12) months after CAR T cell administration).”
From lifelong drugs, to a “one and done” cure.
The concept of an immune reset: “one and done” cure
To make a CAR T-cell therapy, you engineer T cells to target and destroy B cells expressing a target. In the 2022 paper, that was the CD19 antigen. The CD19 antigen turns out to be a key player in many autoimmune diseases. By deeply depleting “autoreactive” B cells, this therapy not only halts disease progression but resets the immune system. (An autoreactive B cell is a type of immune cell that mistakenly recognizes the body’s own tissues as foreign and initiates an immune response against them.)
This approach could yield a one-time cure.
How is an immune reset measured?
It’s an emerging area, so we’re seeing two ways:
Using Cytometry: “Bad” B cells go away, “good” B cells appear
Bad B cell eliminated: Within two days of CAR T-cell infusion, the bad B cells, i.e. CD19+B cells disappeared. Here’s Figure 2c from the 2022 paper:
Good B cells appeared: B cells reappeared after an average of 110 ± 32 days but with a big change. These new B cells were predominantly naive (CD21+CD27−) with no memory or autoreactive profiles. (Remember, autoreactive is bad since those kinds of cells attack your own body.) This is Figure 4C from the 2022 paper. BL means Baseline, RC means reconstitution.
Using ELISA: Autoantibody Clearance
Reduction in Anti-dsDNA Levels: Autoantibodies, such as anti-double-stranded DNA (dsDNA), became undetectable in all patients by the three-month mark in the study.
Comprehensive Autoantibody Decline: Beyond dsDNA, other pathogenic autoantibodies, including those targeting nucleosomes and single-stranded DNA, showed similar declines, reflecting a broad reset of immune activity.
Wrapping up
CAR-T has the possibility of being a lifetime “one-and-done” cure to autoimmune disease. It seems that achieving an “immune reset” is a critical step. The science is still evolving, so watch this space!
Goal: Quantitatively measure that blood specimens collected in TokuKit for cytometry remain stable over a period of time.
Of course, we can do this the old-fashioned way, and just wait 24 months. And we are. However, there’s another method, using the Arrhenius equation, which allows us to “accelerate the aging” process. As we measure the “old-fashioned” way, we’ll update our stability assessment accordingly.
The normal storage temperature for TokuKit is -80°C. For accelerated aging, we used -30°C. We derived the days from the Accelerated Aging calculator, referenced by the American Society for Testing and Materials (ASTM), as follows:
For this, we used TAA = -30C, TRT = -80C, Q10 = 2, and Desired Real Time (RT) = 723 days.
Example 1: in our mass cytometry experiment, we stored the sample at -80°C for one year to naturally age the specimen. We then transferred the sample to -30°C for 12 days to mimic the second year of -80°C storage. This is the “accelerated” part of the experiment.
Example 2: in our spectral flow experiment, we stored the sample at -80°C for 1 week to establish a baseline. We then transferred one aliquot to -30°C for 23 days. This yields the “accelerated” to 2 years, mimicking two years at -80°C.
You just had your IND approved. Now it’s time to look at pharmacodynamic measures. At this point, you want to compare different flow cytometry options. You come back with quotes from two vendors, with these prices per specimen for a gated panel.
Vendor 1: $800 Vendor 2: $1,250
Naive method
Which price is better? Well, this isn’t a trick question: $800 is cheaper than $1,250. $1,250 is 56% more expensive than $800. So yes, Vendor 1 is a better deal. This is the naive way of comparison.
Panel
Cost per specimen
8 marker panel
$800
25 marker panel
$1,250
% Change from 25 to 8
56%
Naive method of comparison
But what if we told you Vendor 1’s panel had 8 markers, and Vendor 2’s panel had 25 markers?
Per marker method
Now, you might compare on a per marker basis.
This is a better way to compare. In this case, you would get $100 per marker for Vendor 1, and $50 per marker for Vendor 2. And in this case, the 25 marker panel is 50% lower than the 8 marker panel.
Panel
Cost per specimen
Number of markers
Price per marker
8 marker panel
$800
8
$100
25 marker panel
$1,250
25
$50
% Change from 25 to 8
56%
-50%
Per marker method of comparison
But, you’re a drug developer and you care about detail in the immune state. You know the immune state means that the more markers you have on your panel, the greater “depth” you can capture in your immune lineage or “tree.”
Price per subset method
That brings us to price per subset.
The 8-marker panel yields 52 subsets while the 25-marker panel produces 298 subsets. This difference dramatically changes the cost-benefit tradeoff. All you need to do is divide the headline price by the number of subsets.
Calculating price-per-subset
Panel 1: $800 / 52 subsets = $15.38 per subset Panel 2: $1,250 / 298 subsets = $4.19 per subset
Let’s see that table again. Now, we can see that using price per subset, the 25 marker panel posts a 73% improvement, compared to an 8 marker panel.
Panel
Cost per specimen
Number of markers
Price per marker
Number of subsets
Price per subset
8 marker panel
$800
8
$100
52
$15.38
25 marker panel
$1,250
25
$50
298
$4.19
% Change from 25 to 8
56%
-50%
-73%
Per subset method of comparison
A simple change in the denominator yields striking differences in the affordability for each panel. Of course, price isn’t the only attribute that matters, but it’s an essential variable that drug developers care about when picking a flow cytometry provider.
How can you use this measure?
If you get different flow cytometry quotes, you might want to use this calculation to get an apples-to-apples comparison.
Are there any limitations to the measure?
This measure just focuses on similar outputs like FCS and gated files. That is, it assumes both methods generate similar outputs. But realistically, drug developers don’t just stop at an FCS file. They need to invest the time or money to turn the FCS file into statistically significant differences. So, this measure doesn’t account for the bioinformatics time or money to turn these FCS files into a true analysis.
Moreover, this measure doesn’t capture sample failure rates, which can impact the actual cost per usable subset. For example, if 10% of samples fail before they even hit a machine with one panel, that would increase your cost per subset downstream.
In sum
The price-per-subset metric is a more realistic way to compare cytometry services. Cytometry’s value is to read more and more detail of the immune “tree”, and by using this measure, you can make better apples-to-apples comparisons between cytometry panels.
What are the advantages of barcoding in mass cytometry? In short, time and money for drug developers.
Here’s an example with 120 specimens. Let’s compare what happens without barcoding and with barcoding.
Parameter
Without barcoding
With barcoding
Processing method
Individual processing per specimen
Samples barcoded and pooled
Number of runs
120 runs (1 per specimen)
6 runs (20 specimens per run)
Data Quality
Variable
Consistent and reliable; reduced cell loss
Figure 1: Comparison of cytometry with and without barcoding
How does barcoding work? Read on to find out.
Example of barcoding in mass cytometry on 120 specimen project
To illustrate how barcoding works in mass cytometry, let’s consider a simple example on the 120 specimen project using metal isotopes for tagging samples.
Goal: Barcode 20 different samples for simultaneous analysis. Remember, we had a 120 specimen project, with 20 specimens in each of the six runs.
First, the barcode criteria must be established:
Purity: each isotope must be >90% pure Titration: each isotope is titrated to the concentration that gives the best signal-to-noise ratio. This is similar to what we do with antibodies. We also ensure each isotope produces similar “brightness.” Combination: after each isotope is titrated, combine three isotopes together into one barcode and test again the optimal concentration to be stained with cells
Second, we pick the barcoding isotopes: We’ll use six palladium isotopes not used in the 40+ marker antibody panel. To construct the actual barcode, we’ll pick three from those six, to avoid signal overlap between the channels. ¹⁰²Pd ¹⁰⁴Pd ¹⁰⁵Pd 106Pd 108Pd 110Pd
Now, the fun part starts.
Third, we build a barcoding scheme:
By using different combinations of these three isotopes, we can create unique barcodes for each sample. Here’s an example:
Mass in Angstroms
Specimen ID
102
104
105
106
108
110
01
1
1
1
0
0
0
02
1
1
0
1
0
0
03
1
1
0
0
1
0
04
1
1
0
0
0
1
05
1
0
1
1
0
0
06
1
0
1
0
1
0
07
1
0
1
0
0
1
08
1
0
0
1
1
0
09
1
0
0
1
0
1
10
1
0
0
0
1
1
11
0
1
1
1
0
0
12
0
1
1
0
1
0
13
0
1
1
0
0
1
14
0
1
0
1
1
0
15
0
1
0
1
0
1
16
0
1
0
0
1
1
17
0
0
1
1
1
0
18
0
0
1
1
0
1
19
0
0
1
0
1
1
20
0
0
0
1
1
1
Figure 2: Barcodes by Sample
Fourth, onto the actual processing steps in the lab:
Barcode individual samples: Each of the 20 samples is incubated separately with a specific combination of the barcoding isotopes from the table above. The isotopes bind to proteins uniformly across all cell types in the sample.
Pool samples: After barcoding, all 20 samples are combined into a single tube. The pooled sample is then stained with a panel of metal-tagged antibodies targeting specific cellular markers (using different isotopes not overlapping with the barcodes).
Acquire the data: The pooled, stained sample is run through the mass cytometer. The cytometer detects the metal isotopes from both the barcodes and the antibody-bound markers.
Deconvolve the data: Specialized software analyzes the data, identifying cells based on the presence of barcoding isotopes. Cells are assigned back to their original samples according to the unique combination of isotopes detected. For example, a cell positive for 102Pd, 104Pd for 110Pd is assigned to Specimen 04.
And then, analyze! Once deconvoluted, each sample’s data can be analyzed individually for immunophenotyping. This includes assessing the expression of various markers to characterize cell populations.
So what just happened?
A drug developer who sends 120 specimens, can get them done in 6 runs, not 120. That means lower batch variation, and better quality data for your patients.
Bonus: what if we need more barcodes?
You might have noticed that we used 6 isotopes to generate barcodes, which yields 20 combinations (6 Choose 3). As the number of specimens per run increases, you can add more isotopes to generate higher combinations (i.e. 10 isotopes can generate 120 combinations).
“I have a 13 marker panel that analyzes T-cells. And my drug just hit the T-cells, why bother looking at any other immune cell types, like B cells, or NK cells? Even if those other cells are affected, I don’t know how to interpret those results.”
The reason is: those extra subsets might predict the difference between success and failure for a drug.
More to the point, if we took a conventional flow panel, you get up to ~108 subsets with a 13-marker panel, almost exclusively T cells. Using mass cytometry, with a 42-marker panel, you get to 792. With spectral flow, it’s around 245. This diagram shows what is missed by the lower-parameter methods. And these are cell subsets we ourselves have found in our own work and collaborations, associated with either clinical benefit, severe immune-related adverse events or peak toxicity.
We know there is a “characteristic” time-series curve associated with clearing infections, like COVID. It would be difficult for a drug developer to analyze a “productive” immune response if whole curves were missing. For example, imagine tracking one curve, like the antibodies. If you tracked just antibodies, you could be misled into thinking someone had an average COVID infection or generic infection.
There is good reason to believe this sort of curve is true in cancer immunology. To that point, this is a nice time-series diagram from Nature Medicine of different cell types in the tumor microenvironment. The authors summarized it well: “understanding what cell types can be modulated and when may enable the next biggest improvements in immunotherapy.”
These distinct subtypes are not of academic interest: these drive real-world responses.
Pre-fixation versus post-fixation. Sounds complicated. What do those terms mean, anyway?
Pre-fixation: The pre-fixation time is from drawing blood to putting it into our first buffer “stable lyse,” starting the blood fixation process. This stage needs to be done within five hours from the time of blood draw.
(The time between adding Buffer 1 and adding Buffer 2 is the “fixation time” which is a separate measurement that should be the recommended 15 min. At this point, the sample is “fixed”: meaning the cells have been “snapshotted” in place.)
Drug developers care about this measure because it can tell them how long a phlebotomist has to put the blood sample into a collection kit. A five hour window gives drug developers the flexibility to collect blood samples from multiple patients throughout the morning and process them all in the afternoon, or collect in the afternoon and process at the end of the day.
Post-fixation: The amount of time once the sample has been fixed and frozen, to when the sample is actually run on the cytometer and data is produced. This can be on the order of months, or even years.
Drug developers care about this measure because it can tell them how long the sample is stable for. For trials that run for 24 months, drug developers want to know that they can get the same quality of data from the 24 month-old samples from their first patient as their most recent patient.
The answer depends on the confidence you need, and the variability you can accept. Variability is expressed by the coefficient of variation (CV) is simply the standard deviation divided by the mean. The higher this number, the more “variable” the measurement. The lower the number, the less “variable” the measurement. Intuitively, a population that appears a lot, say 10% of the time, needs fewer cells than a population that occurs 0.001% of the time.
For purposes of this answer, let’s consider two populations, intermediate Monocytes (inMono) and T cells. These figures are derived from our mass cytometry validation report: https://teiko.bio/technology/, however we believe these reference ranges are applicable regardless of the instrument type (spectral cytometry or mass cytometry).
Cell Population
Median
Inter-run CV%
Intra-run CV%
inMono
0.47%
16.09%
4.63%
T cells
37.28%
0.39%
1.27%
Table 1: Reference Range Values for PBMCs from Healthy Subjects, % of non-granulocytes
Luckily, this question has been addressed in the work of Keeney et al. To wit, “for cell-based assays such as flow cytometry, a simple calculation can be used to determine the size of the database/sample that will provide a given precision: r = (100/CV)2; where r is the number of events meeting the required criterion, and CV is the coefficient of variation of a known positive control.”
We’ve adapted Keeney’s table below:
Desired Coefficient of Variation (%)
1
5
10
20
r = number of events of interest
10,000
400
100
25
When occurring at a frequency of:
Fraction
1:n
Total number of events which must be collected
0.1
10
100,000
4,000
1,000
250
0.01
100
1,000,000
40,000
10,000
2,500
0.001
1,000
10,000,000
400,000
100,000
25,000
0.0001
10,000
100,000,000
4,000,000
1,000,000
250,000
0.00001
100,000
1,000,000,000
40,000,000
10,000,000
2,500,000
0.000001
1,000,000
10,000,000,000
400,000,000
100,000,000
25,000,000
Table 2: “Determination of database/sample size that will provide a given precision in rare event analysis”, Keeney, et. al
Now, let’s couple that with the instruments we have at our disposal: mass cytometry and spectral flow cytometry. Based on our experience, mass cytometry and spectral flow have recovery rates of 50% and 90%, respectively.
Table 3: Estimated events for Peripheral Blood Mononuclear Cell collection
Drawn from Patient (mL)
Cells per mL (M)
Number of Cells in Whole Blood (M)
% Granulocytes
Approximate number of Granulocytes (M)
Non-Granulocytes (M)
Instrument
Recovery Rate
Resulting Events (M)
Resulting non-granulocyte events (M)
2
5
10
50%
5
5
Mass Cytometry
50%
5.00
2.50
5
5
Spectral Flow
90%
9.00
4.50
Table 4: Estimated events for Whole Blood collection
For PBMCs, you can get 0.5M – 10.8M events, depending on the volume of blood collected. In the table, we just show the average ranges, i.e. 2.63M – 4.73M. And for whole blood, you can get 4M – 10.8M total events. Since at least 50% of the cells will be granulocytes, you’ll get to ~2M – ~5.4M non-granulocytes. For readability, we’re showing only a smaller range, from 2.5M – 4.5M.
Now, let’s come back to the two populations from Table 1, inMono and T-cells.
Desired CV
1%
5%
Teiko observed inMono Median % of non-Granulocytes
0.47%
Total Number of events that must be collected, based on Keeney Table
1,000,000
40,000
Estimated inMono Population needed to achieve CV
4,700
188
Actual Teiko collected inMono events
302
Above Threshold?
No
Yes
Actual Teiko Intra-Run CV
4.63%
Actual Teiko Inter-Run CV
16.09%
Industry Standard CV Acceptance Criteria
25-30%
Table 5: Desired CV for inMono
For a 5% CV for a 0.47% population, you would need about 188 events to achieve this CV. And We collected 302, so we were above the threshold. Turns out, we were right in line with inter-run CVs, at 4.63%, and Inter-run CVs at 16.09%.
Desired CV
1%
5%
Teiko observed T-cell Median % of non-Granulocytes
37.28%
Total Number of events that must be collected, based on Keeney Table
100,000
4,000
Estimated T-cell Population needed to achieve CV
37,280
1,491
Actual Teiko collected T-cell events
39,210
Above Threshold?
Yes
Yes
Actual Teiko Intra-Run CV
1.27%
Actual Teiko Inter-Run CV
0.39%
Industry Standard CV Acceptance Criteria
25-30%
Table 6: Desired CV for T-cells
Now, T-cells are much more populous. In our precision study, we collected 39,210 events, beyond the 37,280 and 1,491 cells we need to achieve a good coefficient of variation. And, as it turns out, we clocked 1.27% CVs for intra-run and 0.39% CVs for inter-run.
So, hopefully this gives you an intuition for “How many cells do you need to reliably detect a population of interest?” Interested in capturing a specific population?
In short: We validate antibodies by testing different antibody clones, concentrations, and cellular conditions to optimize signal-to-noise ratio and reduce background noise, ensuring accurate identification of cellular markers. For the full details, download our technical whitepaper ‘Panel and Custom Antibody Validation Process‘ to learn more!
Our customers often want to know that when we have a cellular marker on our panel, such as CD45, we can confidently distinguish this cellular marker signal from background noise. To detect a cellular marker, we need a corresponding clone.
But not all clones are created equal. For instance, here’s an example of two different CD45 antibody clones tested on human cells with very different results. Can you see the difference?
Bad signal: cellular population missing or too low
Good signal: cellular population found
To that end, our immunologists use their extensive expertise to reduce mass cytometry panel validation time from months to weeks, using proven techniques.
Problems with standard approaches: Traditional approaches would use panel design software alone to assign markers to channels, however standard panel design software approaches have crucial limitations. Specifically, these approaches:
only work with markers sold by the manufacturer
leave the antibody clone, antibody concentration, and stimulation condition unspecified for markers not sold by the manufacturer
Imagine you need a marker that’s not sold by the manufacturer: at this point, you need an expert to figure out what the right clone and concentration is. And if you don’t happen to have an expert handy, and you get the wrong clone or concentration: that could yield “bad or missing signals” on a panel, and ultimately unuseable data for precious samples.
Now, let’s compare our approach.
Our process:
We pair the best of software design with expert scientist intuition. Teiko scientists and immunologists have collectively processed over 2,800+ samples and 25 panels, yielding many hard-won lessons about panel design.
Ultimately, for a 44-marker panel, the goal of this process is to produce a 44 X 5 table:
Marker
Channel [44 total]
Antibody Clone
Antibody Concentration (μg/ml)
Cellular conditions for signal detection
CD45
[89Y, 112Cd, …, 209Bi]
[C1, C2, … CN]
[6, 3, 1.5, … 0.1875]
[Stimulated PMA, Stimulated PHA, Unstimulated]
CD8a
[89Y, 112Cd, …, 209Bi]
[C1, C2, … CN]
[6, 3, 1.5, … 0.1875]
[Stimulated PMA, Stimulated PHA, Unstimulated]
…
…
…
…
…
CD45RA
[89Y, 112Cd, …, 209Bi]
[C1, C2, … CN]
[6, 3, 1.5, … 0.1875]
[Stimulated PMA, Stimulated PHA, Unstimulated]
The first thing we do is look at the literature and antibody vendors and see if a marker of interest has been tested before. For each marker, i.e. CD45, we’ll figure out which antibody clone (i.e. HI30) to use based on historical performance.
Then, we’ll determine which channel to use to get the “maximum” signal to noise ratio. In general, Teiko’s experts follow a few rules that sound simple but require a lot of knowledge to execute quickly:
If a marker is weakly expressed on the cells, we put that marker on a “bright” channel to get a good signal.
Similarly, if a marker is strongly expressed on the cells, we put that marker on a “dim” channel to prevent the signal from becoming too “overwhelming”.
Some metals shouldn’t be matched on the same cell population because of concerns for natural impurity, and we won’t put those metals together.
Our experience with channel placement is where our experts can save you weeks of time and frustration.
After finding the right channel, we determine the “optimal” antibody concentration to get the maximum separation between signal and noise. We plot this “stain index” on a curve to find the right concentration.
This is how we generate the stain index:
First, we determine whether the marker needs to be tested in unstimulated or stimulated cells based on whether the marker is expressed on immune cells at baseline. If stimulation is required, we always test two different biological stimulations.
Next, we test six different concentrations of antibody and at each concentration we check:
Strength of the signal
Presence of background noise
“Spillover” (false-positive) signal into other channels
When a marker calls for it, we focus our analysis on a specific population that expresses that marker, so we don’t underestimate the signal.
Finally, we check that our marker performance is comparable to what others have seen in literature.
The end result is a fully populated table and completed validation:
Marker
Channel [44 total
Antibody clone
Antibody Concentration (μg/ml)
Cellular Conditions for Signal Detection
CD45
89Y
C2
1.5
Unstimulated
CD8A
209Bi
C1
0.1875
Unstimulated
…
…
…
…
…
CD45RA
112Cd
C3
3
Stimulated PMA
Illustrative purposes only
What is your process for validating new antibodies and adding them into my customized panel?
Using a process very similar to how we validated our base panels, we verify the optimal staining concentration at which each metal-conjugated antibody demonstrates a detectable and accurate signal on the mass cytometer, while minimizing background signal and spillover into other channels. Once titrated and validated individually, these antibodies are added to the panel and the full panel is then validated.
In sum
That’s a lot of heavy lifting behind the scenes. And we do that in weeks, not months. Teiko scientists performed this process to design all of our current Custom backbone and Standard panels.
