1) As indicated by A. Barani and F. Koch, If you increase the flow rate you increase the introduction of particles through the laser. Your counts should theoretically be proportionnal to the speed at each level (LOW - MED - HIGH).

2) Flow cytometers in general are not made to provide precise volumetric counting (rather relative percentages between particles in the same sample). If you want to make so-called "absolute counts", you have to add counting beads, as suggested by F. Koch. Other options include Flow-Count beads(10 µm from Beckman-Coulter), Tru-Count beads provided in tubes (~4.5 µm from Becton-Dickinson), Cyto-Count (~6µm beads from DAKO),  and Perfect-Count Microspheres (~6µm beads from Cytognos, Spain). All those are made for counting cells in suspension. If your emulsion droplets are smaller, you may prefer 2 to 3 µm counting beads; some are accessible at Spherotech (USA) or Bang's/Life Science, also as polystyrene (that may not sediment as easily as glass).

3) If  your droplets are heterogeneous in size with some going going down to the sub-micron field, as we are used with cell-derived microparticles, then you may enter into more troubles: precisely defining a standardized threshold for scatter-based analysis becomes tricky and needs appropriate tools (e.g. Megamix-Plus from BioCytex, Marseille, F). I would highly recommend using the LOW speed of your instrument and check that you have a linear relationship between your counts and your sample dilution factor in dilution experiments. If you do not have linear correlation, you may be facing some coincidence artifacts: too many particles at the same time in the laser beam and the smallest, normally below the threshold and thus not detected, can generate enough signal to be counted as one event, meaning that you now count droplets that were too small to be normally detected at low speed and/or higher dilution. At the minimum (LOW) speed, the core-stream (sample fluid) is already about 10 µm in diameter . At higher speeds, its diameter increases, the fluids are less laminar (more turbulences between sample fluid/core-stream and sheath fluid) and the small paricles go everywhere.

In conclusion, in order to have absolute counts of your droplets in FCM:

1) standardardize/master your scatter threshold

2) use counting beads of a size and concentration comparable (if not similar) to that of your droplets

3) prefer the LOW speed if that is enough to get significant numbers of events in a few minutes run

4) apply dilution experiments on your samples to check linearity of counts with dilution and thus avoid coincidences and so-called "swarming" effects.

I will assume straight away that you took into consideration the flow rate and the time for which you ran the sample through the flow cytometer and the final volume of the suspension that went through the cytometer.

One possibility is the sheath fluid you use. Is it Isotonic or hypotonic or hypertonic with the emulsion droplets? If it is hypotonic, the phenomenon you observe could be easily explained. If the sheath fluid is hypotonic, the longer the residence time for the emulsion droplets in the sheath fluid, particle size could decrease due to exchange of water between the sheath fluid and the emulsion. Hence you would get lower volumetric yield when you run it at a lower speed.

This occurs frequently with cells in my hands. If the sheath fluid is milli-Q water instead of PBS, with lower flow rates, the cells appear to be bigger. My suggestion would be to use an isotonic solution as the sheath fluid, if it would not affect the regular functioning of the flow cytometer. If that cannot be done, I would suggest that you run the sample at a fixed speed, so that any error in measurement is uniform across the board.

Alternatively, you could try volumetric counts by other methods and compare it with different speeds on the flow cytometer and use the speed that closely matches what you observe by other methods.

I Gilles,

You should not forget that the sheath fluid of your cytometer is the "vector fluid" of you sample. It is a fluid, so it has a flow rate and the more your increase the flow rate, the more you introduce the sample in the flow cell. So the sample quantity analyzed will be increased (with an the same acquisition duration)

I have no experience in emulsion droplets but I agree with Sanath on the importance of the fluid you use during flow analyses. The number of events you read each sec of acquisition should also low enough to permit a correct discrimination between two subsequent events. Finally are you sure that under fast-flow your emulsion is stable?

I Gilles,

I agree with A. Barani.  If you increase the flow rate you increase the introduction of particles through the laser.  I always add fluorescent glass beads to all my samples at a known concentration and use those to calculate the volume of sample run.  This accounts for small changes in the pump speed due to current surges etc. and so my final particle concentrations are not based on running time but rather on the actual volume sampled.   The ones I use are from SpheroTech (2um diameter).  Short answer:  You want a good correlation between the flow rate and the amount of particles counted.  I hope this helps.

This is a link to a basic, still helpful manual. On page 13, the correlation between speed and resolution ( cell number is affected), is explained.

http://www.d.umn.edu/~biomed/flowcytometry/introflowcytometry.pdf

“• A higher flow rate is generally used for qualitative measurements such as immunophenotyping. The data are less resolved, since the cells are less in line in the wider core stream, but are acquired more quickly.”

“• A lower flow rate decreases the width of the sample core and restricts the position of the cells to a smaller area. The majority of cells passes through the center of the laser beam; thus the light illuminating the cells and emitted from the cells is more uniform. A lower rate is generally used in applications where greater resolution iscritical, such as DNA analysis.”

First, you cannot do volumetric counting on any regular flow cytometry that does not use volumetric delivery system such as a syringe. Most don't!. So you need to know that. Sounds like you are not using a syringe driven instrument. Second, if you want to do accurate counts, you would have to purchase a quantification bead and you add that bead to your sample at a precise concentration and count the required number of beads and then you can determine an accurate count. As you increase the "speed" as you call it, you increase the width of the core and increase the potential for multiple particles entering the interrogation region simultaneously.  You should be calibrating your system so that it is set up exactly the same way (use at least 2 different bead populations) and then you can compare results day to day.

1) As indicated by A. Barani and F. Koch, If you increase the flow rate you increase the introduction of particles through the laser. Your counts should theoretically be proportionnal to the speed at each level (LOW - MED - HIGH).

2) Flow cytometers in general are not made to provide precise volumetric counting (rather relative percentages between particles in the same sample). If you want to make so-called "absolute counts", you have to add counting beads, as suggested by F. Koch. Other options include Flow-Count beads(10 µm from Beckman-Coulter), Tru-Count beads provided in tubes (~4.5 µm from Becton-Dickinson), Cyto-Count (~6µm beads from DAKO),  and Perfect-Count Microspheres (~6µm beads from Cytognos, Spain). All those are made for counting cells in suspension. If your emulsion droplets are smaller, you may prefer 2 to 3 µm counting beads; some are accessible at Spherotech (USA) or Bang's/Life Science, also as polystyrene (that may not sediment as easily as glass).

3) If  your droplets are heterogeneous in size with some going going down to the sub-micron field, as we are used with cell-derived microparticles, then you may enter into more troubles: precisely defining a standardized threshold for scatter-based analysis becomes tricky and needs appropriate tools (e.g. Megamix-Plus from BioCytex, Marseille, F). I would highly recommend using the LOW speed of your instrument and check that you have a linear relationship between your counts and your sample dilution factor in dilution experiments. If you do not have linear correlation, you may be facing some coincidence artifacts: too many particles at the same time in the laser beam and the smallest, normally below the threshold and thus not detected, can generate enough signal to be counted as one event, meaning that you now count droplets that were too small to be normally detected at low speed and/or higher dilution. At the minimum (LOW) speed, the core-stream (sample fluid) is already about 10 µm in diameter . At higher speeds, its diameter increases, the fluids are less laminar (more turbulences between sample fluid/core-stream and sheath fluid) and the small paricles go everywhere.

In conclusion, in order to have absolute counts of your droplets in FCM:

1) standardardize/master your scatter threshold

2) use counting beads of a size and concentration comparable (if not similar) to that of your droplets

3) prefer the LOW speed if that is enough to get significant numbers of events in a few minutes run

4) apply dilution experiments on your samples to check linearity of counts with dilution and thus avoid coincidences and so-called "swarming" effects.

It sounds like you have gotten all the pieces of advice I would give, so I will summarize.

Flow rate increases the diameter of the core stream which yields;

- more fluid per unit time

-more cells per unit time (but not in direct correlation with more fluid), in a wider stream cells don't line up and are more likely to go through as doublets

- you must include counting beads with your sample to get accurate counts.  Alternatively, you count your cells separately.  Put a known number of cells/tube.  Then use the qualitative percentages from your flow and apply them to the number you placed in the tube (XX% per million cells)

Additional note: if doing sorting, faster flow rate puts more pressure on the computer to do calculations quickly.  The computers response can be to provide false negatives or increase the stringency of the selection and thus you would get a lower volume... something to keep in mind.

Finally summary: Purchase counting beads to do accurate cell counts with flow cytometry.

I agree with you all that if you want to do accurate counts, you would have to purchase a quantification bead and you add that bead to your sample at a precise concentration and count the required number of beads and then you can determine an accurate count. I will repeat the answer of  Irina Fernandez ·“• A higher flow rate is generally used for qualitative measurements such as immunophenotyping. The data are less resolved, since the cells are less in line in the wider core stream, but are acquired more quickly.”

“• A lower flow rate decreases the width of the sample core and restricts the position of the cells to a smaller area. The majority of cells passes through the center of the laser beam; thus the light illuminating the cells and emitted from the cells is more uniform. A lower rate is generally used in applications where greater resolution is critical, such as DNA analysis.”

So DNA analysis is very sensitive to the flow rate than immunophenotyping.