FACS - Fluorescence Activated Cell Sorting • iBiology

Top

00:00:13.09 Welcome to iBiology.
00:00:14.21 My name is Steffen Schmitt.
00:00:17.06 And today I want to introduce the technology of cell sorting to you.
00:00:21.20 So, when I'm referring to cell sorting,
00:00:25.14 I mean FACS in its original sense.
00:00:28.15 FACS is the abbreviation for fluorescence activated cell sorting,
00:00:31.20 and many people say FACS
00:00:34.21 when they just mean flow cytometry,
00:00:37.20 as as shown here in the cartoon on the right side.
00:00:41.23 So, here we have a single cell suspension
00:00:45.16 labeled with fluorescent dyes, which we bring, here,
00:00:49.08 in front of a focused laser beam and the optics,
00:00:52.24 where we then are able to detect the fluorescent signals
00:00:55.20 and convert cellular structures into light signals.
00:01:01.02 But so far, we are only analyzing those cells,
00:01:04.07 and not doing cell sorting.
00:01:06.06 So, what are the reasons that we need a cell sorter at the end?
00:01:11.20 If we would apply a laser on our tube of cells,
00:01:15.16 we would not get really important and valuable information
00:01:19.08 out of that.
00:01:21.22 Instead, we want to do many fancy things with our cells,
00:01:25.02 like cultivation,
00:01:29.06 transplantation to look for stem-ness of our cells,
00:01:31.24 and do all these omics that are available today,
00:01:35.08 like proteomics and genomics.
00:01:38.18 So, for that, we need to do cell sorting,
00:01:42.21 and this was developed in an immunological surrounding,
00:01:48.05 but in the meantime, cell sorting
00:01:51.10 was established as a very important technology
00:01:54.07 in the biomedical research, ranging...
00:01:56.23 with a lot of different applications,
00:02:00.10 starting with GTL... so-called...
00:02:03.14 where we start separating GFP-transduced cells...
00:02:08.01 so, green to the left sorting...
00:02:11.20 because we maybe also want to establish cell clones with them.
00:02:14.09 We want to synchronize cells for cell cycle,
00:02:18.07 transplant them...
00:02:21.02 But in the past, a very high demand
00:02:24.07 was on single cell analysis
00:02:26.08 and, there, single cell transcriptome analysis
00:02:29.22 is getting more and more important.
00:02:34.03 Today, I'm focused on the droplet electrostatic
00:02:37.18 cell sorting technology
00:02:40.01 and all these other possible mechanisms,
00:02:43.09 like the magnetic activated cell sorting.
00:02:45.24 Chip-based and microfluidic devices will not be discussed in detail today.
00:02:52.13 So now, let's have a look at the machine itself.
00:02:56.05 And what we can see is that many features
00:02:59.05 are shared with a common flow analyzer.
00:03:01.18 We have a sheath stream bringing our cells
00:03:05.00 in front of a laser beam.
00:03:07.03 This is done inside a cuvette.
00:03:09.18 But, for sure, we also need
00:03:12.15 additional topics and features like a transducer,
00:03:15.24 a charging device,
00:03:17.18 a nozzle,
00:03:19.10 and deflection plates,
00:03:21.06 enabling us, later on, to separate the drops from each other.
00:03:25.04 So, cell sorting can be thought of
00:03:28.01 as a two-step process, in principle.
00:03:30.05 In the first step,
00:03:32.13 we are doing our conventional flow cytometric analysis
00:03:35.00 as is discussed in detail
00:03:37.20 in different sessions.
00:03:40.03 But then, we will start injecting our stream into the air,
00:03:46.11 break the stream into small droplets,
00:03:49.12 and then, on the way down,
00:03:53.16 we're able to separate these droplets into different positions.
00:03:57.14 So, basically, we are doing droplet sorting
00:04:01.20 and not cell sorting,
00:04:03.06 and we are only calculating
00:04:05.20 when a cell of interest is reaching this particular drop.
00:04:09.21 That means that everything...
00:04:12.20 all the particulates which are included in the drop
00:04:14.18 will be sorted out, finally.
00:04:18.05 A real picture of an instrument,
00:04:22.00 where you can also see that all the components
00:04:24.16 are in a fixed and a precise position.
00:04:29.00 There is one exception.
00:04:30.23 This you can call the heart of a flow cytometer
00:04:34.17 and a cell sorter --
00:04:37.06 this is this nozzle part, which is sitting here.
00:04:39.07 This is flexible.
00:04:41.02 It can be taken out,
00:04:43.02 and therefore this allows us to adjust the instrument settings
00:04:46.16 to our experimental needs.
00:04:50.05 This is how modern nozzles look like.
00:04:53.04 Here, we have this orifice coming
00:04:57.20 in different diameters,
00:04:59.12 ranging from 50-130 micrometers.
00:05:01.16 And for sure, the size of this diameter
00:05:05.09 is later on, then, influencing very much
00:05:08.13 our stream characteristics.
00:05:11.14 So, let's have a closer look
00:05:14.09 at how the droplets are really formed in the instrument.
00:05:17.11 Like in a loudspeaker,
00:05:20.11 we have a vibrating membrane,
00:05:23.23 which is later on transferring a stable wave
00:05:27.10 onto our stream.
00:05:29.14 So, below the nozzle we have a camera
00:05:32.23 so we can observe this droplet formation.
00:05:36.21 And as indicated here on the left,
00:05:39.04 we have here different nozzle sizes
00:05:44.05 and running with different frequencies.
00:05:47.03 That means for each nozzle,
00:05:48.19 you're generating a specific amount of drops per second,
00:05:53.18 which are then finally...
00:05:56.07 allow us to adjust the speed of our sorting process.
00:06:01.18 There are different possibilities
00:06:05.00 for how to control our droplet breakoff.
00:06:08.23 One of those is the amplitude...
00:06:12.03 and again, when you think about a loudspeaker,
00:06:14.17 the audibility of the tone would be comparable, here,
00:06:20.09 to the amplitude, to the force we're applying to the stream.
00:06:23.08 And with that, we can adjust, here,
00:06:25.24 the drop 1 formation in a very rough manner.
00:06:31.08 The second thing we can control with the amplitude
00:06:33.24 is the so-called gap,
00:06:35.22 which is the distance between the first free drop
00:06:38.23 and the last attached drop on the stream.
00:06:41.06 And the gap is, later on,
00:06:44.01 important to get a stable side stream,
00:06:46.09 because the gap is serving as an insulator
00:06:49.04 of the different charges we apply,
00:06:51.11 and only with an optimal gap
00:06:55.06 will we have nicely separated side streams later on,
00:06:57.22 without spraying and potential contamination later on.
00:07:03.16 So, drop 1 itself... we will hear in a few slides...
00:07:07.18 it's very important that we keep this position quite stable
00:07:11.18 and that we can precisely calculate
00:07:14.14 where this drop 1 is located.
00:07:18.09 So, once we've generated our drops,
00:07:21.09 we can charge them,
00:07:25.00 and later on we can deviate them in an electrical field.
00:07:29.00 So, we are using physiological salt solutions
00:07:32.15 as a carrier stream, as a sheath.
00:07:34.13 So, this sheath stream, the sample stream,
00:07:37.07 can be charged,
00:07:39.24 and this charge is then traveling down the stream
00:07:43.00 to the droplet breakoff point,
00:07:45.14 and in the moment when the drop is leaving the stream,
00:07:49.12 it also carries the same charge
00:07:52.06 as the stream before.
00:07:53.22 And by going down, it will then move, here,
00:07:57.02 to an electrical field,
00:07:59.12 and it's then deflected in the right manner
00:08:02.19 and can then be collected with the right collection device.
00:08:09.08 So, now we have to bring all the things together.
00:08:12.00 So, we have the point where we're deciding
00:08:15.24 if we want to sort the cell or not.
00:08:19.08 This is the laser interrogation point.
00:08:21.23 We have a droplet break off point.
00:08:24.13 And now we have a charging of this drop.
00:08:29.13 And now all these three parts have to be synchronized,
00:08:31.18 and this can be done with a so-called drop delay,
00:08:35.00 which is basically the timespan or the distance
00:08:38.17 a cell needs to move
00:08:41.22 from the interrogation point down to the break off point.
00:08:46.17 We can accurately calculate this time span
00:08:50.05 because we are running the system with constant velocity.
00:08:54.08 And what we also mentioned already
00:08:57.12 is that the break off point is very constant.
00:09:00.17 So, also the time is constant
00:09:02.13 and can be precisely calculated.
00:09:07.05 The number we are getting out
00:09:09.04 is not the time in seconds.
00:09:11.00 It's basically really the distance
00:09:13.04 which is fitting inside,
00:09:15.16 so the drop delay is given as numbers
00:09:18.13 which are fitting in the distance
00:09:20.15 between the interrogation point and the droplet break off point.
00:09:25.08 And in this example, it would be 6.5 drops.
00:09:30.21 To define the drop delay,
00:09:33.06 there are different systems available.
00:09:36.03 The most prominent one is the so-called Accu Drop system.
00:09:39.16 This consists of a separate laser
00:09:43.24 which is equally illuminating
00:09:46.15 the deflected drops and the waste stream.
00:09:50.01 So, when we're now using fluorescent beads,
00:09:53.02 we can simulate the sorting process.
00:09:56.07 And when we calculated the right drop for sorting,
00:10:01.08 all the beads should be sorted out,
00:10:05.12 and no beads should be left in the waste stream.
00:10:10.03 So, this is how it can look in an instrument,
00:10:12.15 in a software-aided system.
00:10:15.08 So, we have, now, two areas:
00:10:17.12 one for the waste and one for the deflected stream.
00:10:21.12 And you see, with a correctly calculated drop delay,
00:10:25.06 all the fluorescently labeled beads are sorted out,
00:10:28.14 and none are left here in the waste stream.
00:10:33.21 So, we are looking for a minimum here
00:10:37.03 in the waste stream,
00:10:40.19 and the calculation between sorted and unsorted events
00:10:43.17 should sum up at the end to 100%.
00:10:49.18 There is a method published by a colleague of mine,
00:10:54.17 Andy Riddell, and Rui Gardner
00:10:57.18 which is called the Rmax method,
00:10:59.22 which is a quick and independent way
00:11:01.24 to prove your instrument settings
00:11:04.11 besides the Accu Drop system.
00:11:09.13 What you also can see is that the actual drop delay values
00:11:15.09 can vary from nozzle size to nozzle size
00:11:17.23 and are actually quite bigger than the one
00:11:20.16 we had in the example.
00:11:23.23 So, we now heard about the theory.
00:11:28.15 We now know how to justify
00:11:31.20 if you want to sort or not.
00:11:34.01 Now let's assume a more realistic daily situation.
00:11:38.18 You have an experiment where you need a very pure sample
00:11:43.12 with a maximum of your cells.
00:11:45.22 But as always, time is money,
00:11:48.05 and you need it very fast.
00:11:50.23 Of course, all these three things
00:11:53.24 -- purity, yield, and time --
00:11:56.04 cannot be maximized at the same time.
00:11:59.03 So therefore, such a request
00:12:02.16 would just not be possible.
00:12:04.16 So, to understand this,
00:12:07.02 we first will have a look at what are realistic sort speeds
00:12:10.00 and how fast we can go.
00:12:13.20 We already saw that with a smaller nozzle,
00:12:16.13 like a 70 micron nozzle,
00:12:18.17 we can generate more drops.
00:12:20.23 So, with 90,000 drops,
00:12:23.10 we can run about 30,000 cells/second,
00:12:26.09 so this is giving us 100 million cells/hour.
00:12:31.16 This sounds quite much at the beginning,
00:12:34.04 but if you now start thinking
00:12:37.00 about real cell populations,
00:12:39.03 you will end up after a full day of sorting
00:12:42.07 with not more than 2 million cells.
00:12:45.02 If you now have very complex experimental setups,
00:12:47.24 this might already be a limiting cell number to go on.
00:12:57.10 So, why do we have to limit to 30,000 cells/second
00:13:00.17 if I just told you that with a 70 micron nozzle
00:13:03.11 we have 90,000 drops available.
00:13:07.22 And this is because we don't get an equal distribution
00:13:10.23 of our cells inside the stream,
00:13:13.00 like in this ideal situation.
00:13:15.18 Under this situation,
00:13:18.05 we would have no problems to discriminate rare cells
00:13:20.13 and to sort them out very efficiently.
00:13:24.01 Unfortunately, the reality looks a little bit different,
00:13:28.01 so we have a more or less random distribution
00:13:31.03 of the cells inside the stream,
00:13:33.10 and therefore, also, a quite high possibility
00:13:36.07 that two cells will be matching up in one drop,
00:13:40.01 and therefore we have some difficulties
00:13:43.11 that rare cells are completely free of
00:13:46.02 unwanted, contaminating neighbors.
00:13:49.20 So, now let's have a look, again,
00:13:53.15 at our numbers and what is given by the physics.
00:13:57.19 We already learned that with...
00:14:01.24 going down with the nozzle size,
00:14:04.13 we can increase our sort speed,
00:14:07.06 but we have to pay a price for that.
00:14:10.21 To get a stable droplet break off with a small nozzle,
00:14:15.01 we have to do a high pressure...
00:14:17.08 apply a very high pressure on the stream.
00:14:20.01 This induces shear stress and forces to the cell,
00:14:24.14 and therefore the cell viability
00:14:27.22 will be higher than when we are choosing a larger nozzle size.
00:14:32.17 To summarize some of these features,
00:14:36.10 we can come up with some rules of thumb.
00:14:39.09 For example, cell size is mattering.
00:14:42.05 It should not exceed 1/3 of the nozzle diameter;
00:14:45.23 otherwise, again, the shear stress
00:14:49.03 will become more important
00:14:51.20 and the cells start dying sooner after sorting.
00:14:56.01 And also, the side stream
00:14:59.03 will not be as stable as possible.
00:15:03.04 Another thing is the optimal event rate
00:15:05.24 you can run on those machines,
00:15:08.16 and this is roughly 1/3-1/4 of the drop frequency.
00:15:13.19 Therefore, also, we are coming down to the 30,000 drops/second
00:15:18.13 we can run at maximum.
00:15:21.14 So, you see,
00:15:23.22 all these settings together are influencing our decisions
00:15:28.12 and also are dependent on our downstream assay.
00:15:31.24 So, for sure, if you want to have viable cells
00:15:35.09 for transplantation or for cell culturing,
00:15:37.21 then you would choose a more gentle treatment of your cells
00:15:42.07 at the beginning.
00:15:44.14 We heard about these three terms:
00:15:47.08 yield, purity, time.
00:15:49.18 And we can put that here in a kind of triangle
00:15:52.23 because all three terms are influencing each other,
00:15:57.24 and one of the reasons is that you cannot really maximize
00:16:01.20 all of them at the same time --
00:16:04.21 you always have to find a compromise to that.
00:16:08.03 If you have a focus on your yield,
00:16:11.13 then you have to sacrifice your purity
00:16:13.24 or you have to increase the time.
00:16:17.20 In the same way, if purity is an issue,
00:16:20.20 then either your yield is going down
00:16:23.11 or, again, time does matter.
00:16:27.09 And like in our example, if you have low... less time,
00:16:33.15 then you have to live with a lower output
00:16:36.00 and also with less purity.
00:16:40.21 Sometimes there is a mixing or misunderstanding in terms.
00:16:44.17 That's why I'll explain the meaning of yield to you.
00:16:49.07 So, let's assume you have 10% positive cells in your sample,
00:16:53.17 and we are analyzing 100,000 cells.
00:16:57.00 That means, at the end,
00:17:00.01 we had 10,000 cells which could have been sorted out.
00:17:03.20 Now, after the sorting process,
00:17:06.11 the machine tells you that you only have 9,000 cells sorted,
00:17:11.00 so the yield is then 90% --
00:17:14.15 90% of that which you potentially, or optimally,
00:17:17.10 could have sorted out.
00:17:19.10 But why are we losing, now, 1,000 events?
00:17:24.17 This is due to the fact, as I said,
00:17:27.05 we have a random distribution,
00:17:29.03 and therefore cells are matching up in one drop,
00:17:32.01 or coming too close,
00:17:34.09 that the purity would be impaired.
00:17:36.18 And this is then called
00:17:40.03 a coincidence or an electronic abort,
00:17:42.08 if you run them too fast.
00:17:44.09 So, all these effects have the consequence
00:17:47.19 that your yield is normally always lower
00:17:50.10 than what you originally calculated.
00:17:55.15 But even with a yield of 90%,
00:17:58.01 you cannot be sure that you will really re-find
00:18:00.20 these 9,000 cells in your collection device.
00:18:04.21 What you really find is the recovery of your cells,
00:18:08.07 which is then normally between 70 and 80%
00:18:12.11 of your yield.
00:18:14.01 Why is that?
00:18:16.24 As I said, cell sorting
00:18:19.10 is very stressful to your cells,
00:18:21.15 so they start dying, getting fragmented, or...
00:18:26.11 cells are living entities.
00:18:29.11 We are using mostly plastic collection devices,
00:18:33.13 so the cells can stick to your plastic tubes and the walls,
00:18:37.22 and therefore can get lost.
00:18:41.06 Also, the sample preparation and processing after the sorting c
00:18:45.02 ould impair the recovery.
00:18:47.02 Think about centrifugation --
00:18:49.16 also there, you can lose some of your cells.
00:18:56.14 Now it would be very stressful and demanding
00:18:59.24 to find, always, a good compromise between
00:19:04.13 this yield, time, and purity.
00:19:07.01 And therefore, the machines are giving us
00:19:09.23 already predefined options to choose from,
00:19:12.20 which are the so-called sort precision modes,
00:19:16.23 and they are coming, here, in the three different flavors.
00:19:19.03 Of course, there are more, but I'm just introducing these three to you.
00:19:23.08 And already, what the names are suggesting...
00:19:26.17 one or the other term is then prioritized in that.
00:19:32.09 And these precision modes are always a mixture and combination
00:19:36.10 of the so-called sort masks.
00:19:39.14 So, now I'm already introducing the next term,
00:19:43.09 so now, let's have a closer look at what the sort mask is about.
00:19:48.14 Most cell sorters are using three different sort masks:
00:19:51.17 the yield mask,
00:19:53.19 the purity mask,
00:19:56.19 and the phase mask.
00:19:57.20 So, each mask is dividing the drop, virtually,
00:20:01.22 into 32 segments,
00:20:05.13 and then a cell can be touching or located
00:20:08.21 in one of more of the segments,
00:20:10.22 and this will then influence the outcome
00:20:13.08 of my sort decision.
00:20:15.24 For example, the yield and the phase mask
00:20:19.09 are looking at the drop we just calculated
00:20:21.24 in the drop delay,
00:20:23.21 whereas the purity mask is looking
00:20:26.13 at the trailing and the leading drop
00:20:29.04 to look if there are contaminating neighboring cells
00:20:33.04 next to our cell or interest.
00:20:37.15 So, all these three masks are now combined
00:20:40.19 in these sort precision modes,
00:20:43.10 which I'll summarize now in the next tab.
00:20:47.10 So, from the sort software,
00:20:52.04 you can now choose, here, yield, purity, and single cell.
00:20:54.19 And if you have chosen the yield mask,
00:20:58.17 you see the yield is set here to 32.
00:21:03.12 So, that means a cell will always be touching a segment,
00:21:06.18 and therefore always two drops
00:21:10.08 are sorted to make sure that
00:21:13.14 the cell of interest really makes it into our collection device.
00:21:17.12 So, the aim is to get all of our positives,
00:21:21.00 and therefore we are getting a very high recovery.
00:21:24.09 But because we are sorting two drops, always,
00:21:27.22 we are also getting a higher volume.
00:21:32.12 So, in the yield setup,
00:21:34.15 you see we don't care about the purity.
00:21:37.17 This is coming in the next combination where we...
00:21:42.13 in addition to the yield mask,
00:21:44.22 we are now adding the purity to maximum
00:21:47.17 so that we only get the positive ones
00:21:52.03 with a very high purity.
00:21:54.06 But that means if we have a contaminating neighbor
00:21:56.09 somewhere in the ejecting drops
00:22:01.07 that we are also losing cells of interest.
00:22:04.23 The last option is the single cell option,
00:22:09.12 and this is chosen... very special
00:22:12.09 when we, for example, think about cloning our cells,
00:22:16.10 where we really want to have single cells
00:22:18.24 in our 96-well plate, for example.
00:22:21.14 There, we set the yield mask to 0
00:22:24.17 because we want to sort only one drop.
00:22:28.15 We want to sort with a very high purity.
00:22:31.20 And to make sure that the drop is not empty,
00:22:37.01 we're also setting a phase mask to 16,
00:22:39.20 so only a drop is sorted
00:22:42.13 when the cell is located in the middle of a drop
00:22:45.14 and there's no chance that this cell
00:22:48.09 is moving to the next or to the drop before.
00:22:53.12 So, we're getting our cells in very low volume,
00:22:56.23 but for sure,
00:22:59.08 we're losing most of our cells of interest
00:23:01.20 because it has to come together
00:23:04.23 that they're located in the middle of the drop
00:23:08.13 without any contamination surrounding.
00:23:12.09 Please be aware if you set the phase mask to 32
00:23:16.17 then no drop will be sorted,
00:23:19.19 because the cell is then always in the phase mask,
00:23:22.17 and this will lead to 100% abortion of your cells.
00:23:27.18 At the end of the talk,
00:23:29.20 I want to discuss some strategies
00:23:32.16 for how you can improve your sort quality and the outcome of your sorting experiment.
00:23:37.20 As you're used to in conventional flow cytometry,
00:23:42.09 you should exclude doublets and dead cells
00:23:47.20 from analysis. [00:23:49.04You should think about your target, collection medium...
00:23:53.07 because this can have an influence on the viability of the sorted cells.
00:23:57.13 If time is an issue,
00:24:00.07 for sure you should think about if you can
00:24:03.10 pre-enrich your sample beforehand.
00:24:07.10 And at the end, I will show you some examples
00:24:11.19 of how the setting of threshold and gating strategy
00:24:14.22 might influence the outcome of your sort.
00:24:21.16 Now, we do recommend that you also
00:24:26.05 need to do a proper compensation.
00:24:28.13 Even though in conventional flow cytometry,
00:24:31.08 you can change your compensation settings afterward,
00:24:35.22 keep in mind that then, at this time,
00:24:39.16 you already have sorted your cells,
00:24:42.04 and then you cannot change your sort gates anymore.
00:24:48.09 Let's come to the threshold.
00:24:50.08 So, here are two examples
00:24:53.01 for differently set thresholds.
00:24:55.08 In one situation,
00:24:57.09 we have a very low threshold,
00:24:59.13 where all the particulates can be seen,
00:25:02.02 whereas in the second situation
00:25:04.04 we have a higher threshold,
00:25:06.07 which is also leading to a higher percentage
00:25:08.21 of the target populations.
00:25:13.18 So, the advantages might be obvious,
00:25:17.06 namely that you can analyze your cells faster,
00:25:22.03 because rare populations get visualized better in your plot,
00:25:27.10 so you have an overall higher throughput of sorting,
00:25:31.16 and maybe you're also, then,
00:25:34.03 therefore getting slightly higher yield of your cells of interest.
00:25:37.18 But please always keep in mind that particulates
00:25:42.03 which are below the threshold
00:25:44.13 cannot be seen by the machine and therefore are not analyzed.
00:25:48.19 So, your purity will be impaired very significantly
00:25:53.10 because you cannot exclude unwanted events
00:25:56.10 from the sort decision process.
00:26:02.15 To come to an end,
00:26:05.06 we will also have a look at the gating strategy.
00:26:07.22 Here, I chose an example for regulatory T cell staining
00:26:11.06 with CD4 and CD25.
00:26:16.13 So, you can see two populations:
00:26:20.21 one CD4 single positives
00:26:23.17 and one CD4/CD25 double positives.
00:26:27.15 And for both populations, we set, now,
00:26:30.11 two sorting gates for a high and a low expression
00:26:34.11 of these markers.
00:26:36.10 And if we now have a look
00:26:39.01 at the reanalysis after sorting,
00:26:41.10 we can clearly see...
00:26:45.00 here are the CD4/CD25 double positive cells.
00:26:48.07 We are getting, here, a good signal;
00:26:50.22 therefore we are able to separate
00:26:54.06 P2 and P3 very nicely.
00:26:57.15 Whereas in the CD25 negative population,
00:27:02.21 P5 and P4 cannot really be distinguished from each other,
00:27:08.01 because here we are in the background,
00:27:11.06 and if you reanalyze a background
00:27:13.24 it always gives you the background back.
00:27:16.14 This is just due to the data spread
00:27:19.03 inherent to the measurement,
00:27:21.18 and it has nothing to do with being more positive or more negative.
00:27:29.18 When you are performing cell sorting,
00:27:32.05 always ask some questions beforehand
00:27:36.00 so that you can then,
00:27:39.02 from those things,
00:27:41.10 conclude what kind of instrument setups,
00:27:44.00 what kind of nozzles and sort precision modes
00:27:46.11 you are choosing.
00:27:48.21 You know, what is your downstream assay?
00:27:51.20 Do I need cell viability?
00:27:54.09 What is the desired number of cells
00:27:56.24 I am having to do so?
00:27:59.13 What is inside my cell?
00:28:01.22 Can I reach these desired cell numbers?
00:28:04.09 What is the percentage of my target cell?
00:28:06.20 Is there maybe a high chance for coincidence to do so?
00:28:10.14 Therefore, which parameter do I need to maximize?
00:28:13.17 Is it purity, yield, or is it time,
00:28:16.13 because my instrument is heavily booked
00:28:19.08 and it's not accessible all the time?
00:28:21.19 Can I do a pre-enrichment?
00:28:24.19 Or what is my target collection device?
00:28:28.03 So, many things can be clarified beforehand,
00:28:31.10 before you even go on the machine
00:28:33.20 and start your process.
00:28:37.20 If you're now getting interested in this technology,
00:28:40.07 I can recommend some books and publications to you
00:28:46.00 for further information.
00:28:48.09 So, there are two books.
00:28:50.05 One is the Howard Shapiro... the Practical Flow Cytometry.
00:28:54.19 Howard is a famous teacher, at the moment,
00:28:58.11 here in flow cytometry,
00:29:00.13 and his book is dealing a lot with flow cytometry overall.
00:29:05.20 Whereas the book of Andreas Radbruch
00:29:08.01 is also focusing a little bit more on cell sorting.
00:29:12.01 And quite recently,
00:29:14.19 there was a publication in the European Journal of Immunology,
00:29:17.03 where some colleagues and I
00:29:19.20 put some stuff together
00:29:22.14 to give you a good starting point
00:29:25.08 for flow cytometry and cell sorting overall.
00:29:29.21 Well, at the end,
00:29:32.02 please keep in mind we are always dealing with living cells.
00:29:35.16 We're doing very harsh things to them
00:29:38.17 and trying to treat them as gently as possible.
00:29:42.02 Thank you for your interest.

This Talk

Speaker: Steffen Schmitt

Audience:

Educators of H. School / Intro Undergrad
Student
Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: May 2018

Talk Overview

FACS (fluorescence activated cell sorting) differs from conventional flow cytometry in that it allows for the physical separation, and subsequent collection, of single cells or cell populations. FACS is useful for applications such as establishing cell lines carrying a transgene, enriching for cells in a specific cell cycle phase, or studying the transcriptome, or genome, or proteome, of a whole population on a single cell level. Dr. Steffen Schmitt explains the components and basic function of droplet-based cell sorters. He also provides strategies to optimize the key values in cell sorting (e.g. yield, purity, recovery time, and viability) depending on the downstream assay to be performed on the isolated cells.

Speaker Bio

Steffen Schmitt

Dr. Steffen Schmitt studied biology at the Ruprecht-Karls-University of Heidelberg and completed his PhD in the Department of Cellular Immunology at the German Cancer Research Center (DKFZ). After a short post-doc, he established and led a flow cytometry core lab at the Natural Science and Medical Research Center (NMFZ) at Johannes Gutenberg University of Mainz…. Continue Reading