HELIOS Autosampler set-up extremely prone to clogging

[yunhualee]

Hi everyone! My lab just upgraded our CyTOF 2 to Helios with their new PSI (sample loader) and we have been testing it our with PMBC samples. However, we encountered clogging in either the sample line, sample capillary or the nebulizer on multiple occasions. Although these clogging occured time to time with CyTOF 2, it was never this bad.

While we are able to unclog the sample lines/capillaries, we were unable to do so with the nebulizer despite following the recommended troubleshooting by Fluidigm. We resorted to using the spare nebulizers for CyTOF 2 we had on hand for now. Even so, the nebulizer still got clogged after running a 2ml PMBC sample (diluted to 0.5mil cells/ml).

Is there anyone out there with the same problem and/or knows why this is happening so frequently? Also, is there a better way to unclog the nebulizer other than soaking it in hot water or detergent as recommended in the user guide?

Any advices will be much appreciated. Thanks in advance!

[mleipold]

Hi yunhualee,

The Helios PSI unit clogs. There's no other way to describe it. It clogs noticeably more than a v1 or v2 instrument ever did with the sample loop, and more than any of the three models do with the SuperSampler.

We have had 2 Helios now for about a year and a half: both of them do it, and have from the time we first got them. At CYTO 2016 in Seattle, every lab that I talked with that had a Helios complained about it.


Fluidigm is aware of the problem, though I have no idea where it is in their list of priorities. In the short term, you only have a couple options:
1. Purchase a SuperSampler ( http://www.victorianairships.com/ ) and use that as your sample introduction system, rather than the PSI unit. You will have to use the PSI for tuning, though.

2. Manage the clogging as best you can. For PBMCs, we dilute to no more than 0.8M/mL (800K/mL), and if the samples are stickier than usual, we dilute further. For whole-blood, we dilute to no more than 0.6M/mL (600K/mL). We make sure to run MilliQ water between each sample, and to run Wash solution periodically: at least once every 1-1.5hr. We also keep the Status Panel open, toggled so we can see the Pressure reading (see below for pressure values we've determined)

In terms of where the clogs originate: we have had clogs at every single junction or in every single piece of tubing in the unit, from the inlet of the sip through the nebulizer itself.

In our experience:
1. Most of the clogs happen at the silica line side of the grounding nut (where the blue fitting screws into the grounding nut). Unscrew the fitting, let the liquid drip for a bit to clear it, then screw back in.
2. Most of the rest of the clogs wind up happening in the nebulizer capillary. We have a bunch of them pre-assembled to swap them out on the fly. We buy the capillary kits from Fluidigm, but usually buy the thin white tubing in a 100ft spool from Cole-Parmer (maybe supplied through Fisher?):

Cole-Parmer
Ultramicrobore PTFE Tubing, 0.004"ID 0.016"OD, 100 ft/pack
Item# EW-06417-72
http://www.coleparmer.com/Product/Ultra ... m=06417-72

From time to time the yellow ferrules wear out and need to be replaced:
Ferrules (10 pack): Fisher cat# 05700386
Manufacturer: Upchurch Scientific™ P259X

3. Some clogs happen in the grounding nut itself. You can use a syringe on a screw fitting (FPLC or HPLC fittings) to backflush the grounding nut.

4. Next, clogs can happen at the inlet of the sip. You can use the syringe/FPLC fittings to backflush the PSI from the blue fitting all the way back into an empty tube in the PSI. I recommend doing this with the PSI unit open, to allow you to watch the flow rate: a clear line should give you at least very rapid drops, if not an actual stream.

5. Otherwise, the clogs can happen in unit components themselves. Before you try this, have your Fluidigm FAS teach you how to unscrew, flush, and properly reassemble the silica lines from the underside of the PSI and the sip line from the top of the flow meter (boxy part on the right side of the PSI, usually covered by a metal shield. It connects to both the sip line from above and the long silica line from below).
a) long silica line with blue fitting. Disconnect from bottom of the PSI and backflush with the syringe. You should get a strong steady stream coming out.
b) while the long silica line is disconnected, place a tube of MilliQ or Wash solution in the PSI chamber and start running sample. This will allow liquid to drip out the bottom of the flow meter to clear any clogs there.
c) reconnect the long silica line, and very carefully unscrew the sip line from the top of the flow meter. The chamber will have to be at least partially open, to allow the sip line to straighten up out of your way. Backflush from the blue fitting of the long silica line up through the flow meter to clear any clogs that might have been too big to get through the flow meter. You won't see a stream or anything, just a bit of water bubbling out the top of the flow meter.
d) while the sip line is disconnected from the flow meter, very carefully close the chamber (avoid hitting or crimping the sip line with the handle). Turn sample introduction on to forward-flush the sip line. The software will only run the sample for a bit: since the sip is disconnected from the flow meter, it can't register the default 30uL/min flow rate and will probably shut down (might give you the "DRY" reading in the Status panel).
e) carefully reconnect the sip line to the top of the flow meter.

- as your FAS should teach you, it's important with both silica lines (sip and long line) to have the tubing protrude at least a bit beyond the tan screw fitting before screwing things back in. If you don't, you can crimp the tan fitting over the end of the silica line, *creating* a choke-point which would lead to more clogs and also probably causing some leakage because you don't have a tight seal.


Some pressure recommendations we've learned over the years (all values are at 30uL/min flow rate default):
1. If everything is connected (PSI through nebulizer), your Pressure should be 12-17psi, definitely below 20psi. As you start forming clogs, the pressure will start creeping up. Also, your event rate will start to fluctuate.
2. If everything but the nebulizer is connected (PSI through nebulizer capillary), your Pressure should also be 12-17psi, definitely below 20psi.
3. If PSI through long silica line are connected to the grounding nut (but nebulizer and nebulizer capillary are *not* connected), your pressure should be below 4psi.
4. If PSI through long silica line are connected but long silica line is *not* connected to the grounding nut, your pressure should also be below 4psi.
5. If PSI is connected but long silica line is *disconnected* from the bottom of the PSI, your pressure should be 1.5-2.5psi (definitely below 3psi).

- if your Pressure values differ from the above, you probably have a clog, and can use the above values to help isolate where the clog is.
- For example, your pressure with all lines but no nebulizer (#2) is 25psi, but your pressure without the nebulizer capillary (#3) is 3psi, the clog is probably in the nebulizer capillary and you should replace it.


Sorry for the long post, but hopefully you (and others) will find it useful.


Mike

[AdeebR]

Hi Mike,

I was told recently that the flow rate meter on Helios PSI will stop acquisition if it detects a clog.

When running tissue derived samples, we've found that the CyTOF regularly suffers from "microclogs" that are apparent from transient fluctuations in event rate, but these seem to resolve as the the sample continues to run and don't necessarily require interventional de-clogging.

I'm wondering if these same fluctuations in flow rate may trigger the clog detection on the PSI and causes the acquisition to stop, causing a minor clog to be perceived/treated like a major clog, whereas if liquid was continued to be forced through the system it would force these push these minor clogs out and allow acquisition to proceed.

Do you think thinks could be a contributing factor or are the clogs really just clogs?

[FlowjoVA]

Thanks Mike for the thorough explanations. I totally concur with your assessments. We too have all the same issues you describe. One of the biggest issues we have found is the is the shaving of the plastic female connector on the neb gas line to the metal male connector. These plastic shavings end up in your nebulizer and are almost impossible to get out. We have found soaking the nebulizer for several days in 4% nitric oxide eventually dissolves it out. I have attached pictures to show you what I am referring to. If you look closely at the picture of the nebulizer you ca see the small thread of plastic peaking out of the end of the nebulizer. and you can also see the threads on the plastic connector are shaved. Once this happens it just gets worse and you will notice that you have more difficulty with screwing the connectors together.
https://www.dropbox.com/s/6gux4mbp4dpd4 ... .jpeg?dl=0
https://www.dropbox.com/s/ajjp88ppgtxc6 ... .jpeg?dl=0
https://www.dropbox.com/s/cm3igttj4p5mx ... .jpeg?dl=0
https://www.dropbox.com/s/ppzc56v0mmaqu ... .jpeg?dl=0
The community needs to put pressure on Fluidigm to get this resolved and I don't mean as something else we need to buy. This is not a feature, this is a malfunction and it needs to be addressed and addressed quickly. We are looking at the Super Sampler now, which will correct some of the clogging issues, but not the one I described here.
Best,
Joanne

[mleipold]

Hi Adeeb,

There are microfluctuations in event rate, definitely: maybe fluctuating between 200-250 cells/sec.

However, what I'm referring to are cases where you're running a sample at, say, 250 cells/sec, and all of a sudden your event rate drops to ~120 cells/sec, usually coupled with an increase in pressure of at least 2-3psi (often 5+ psi). In many cases, a couple minutes later you'll see a *bunch* of cells come out over 15-30 seconds and the event rate will jump to something like 400 cells/sec and the pressure will drop back to the original value as your microclog burst through. Those are irritating, and also a sign you should probably clean with Wash or 3% nitric immediately after that sample.


The PSI unit will stop itself if the pressure goes to 45-50psi and it still can't achieve the 30uL/min flow rate. If that happens while you're running a sample, you'll get a pop-up prompting you to load more sample and press "OK" button, or hit "Cancel" button to stop acquisition. The Status panel will also show "DRY" rather than "RUNNING" in the PSI line. Basically, the PSI thinks that the reason it can't achieve 30uL/min is because your tube ran dry and it's pumping air. That's also the reason why literally the *only* way to clear the "DRY" setting is to open and close the PSI chamber: there's no software button to do that. That resets "DRY" to "IDLE" and then you can hit "OK" to continue.

If you get a "DRY" signal and you still have sample left, then the reason for your 50psi pressure is because of a solid clog. In that case, you *still* must open and close the chamber to clear the "DRY" setting back to idle; if you do not do that and still hit "OK" (maybe after changing the nebulizer capillary), it will act like it's starting sample acquisition, but after the 60 second delay before acquisition, it will throw the same error message.


In more recent versions of the firmware for the PSI unit: if you are starting the PSI from "IDLE" and it starts to ramp up pressure without getting any flow rate, it will stop trying at maybe 15psi rather than 50psi with 0uL/min flow rate. Similarly, if you are troubleshooting the various lines and connections, this version of the firmware does not like larger pressure jumps.

For example, if you have the blue fitting of the long silica line disconnected from the grounding nut and your pressure is 3psi, simply connecting the fitting back to the grounding nut and nebulizer capillary will cause your pressure to jump to 12-16psi, even without a clog. This is enough of a pressure jump in a very short time period to make the PSI stop. From what I understand, this was added to the firmware because of concerns about high pressure causing a large-droplet spray from the nebulizer which would overcool the plasma ("blow it out") like would happen with v1 and v2 instruments when a strong nebulizer clog would burst through. However, in two machine-years of experience with our instruments, we had that happen *no more than* one time. This is irritating, as it gets in my way of troubleshooting clogs, but is manageable.

There's also a separate issue with that firmware version of a 15-second delay in pressurizing and starting sample introduction, every single time that you turn the PSI on. Again, irritating, but manageable....though those 15 second increments add up quickly over the course of a full run-day.


Mike

[FlowjoVA]

We have found a different source of clogging that is not related to the sample itself, but rather to the clogging of the neg gas line resulting from the shaving of the plastic female fitting from the metal male fitting (bad combination). These slivers of plastic end up in the nebulizer and can only be unclogged after soaking in 4% Nitric acid for several day. We have spoken to Fluidigm about it and they are aware of the problem. Here are some pictures to show you what I am talking about.
https://www.dropbox.com/s/6gux4mbp4dpd4 ... .jpeg?dl=0
https://www.dropbox.com/s/ajjp88ppgtxc6 ... .jpeg?dl=0
https://www.dropbox.com/s/cm3igttj4p5mx ... .jpeg?dl=0
https://www.dropbox.com/s/ppzc56v0mmaqu ... .jpeg?dl=0
Best,
Joanne

[vtosevski]

Hi Yunhua,

As you also asked about cleaning the nebulizer, let me add that we clean ours by pushing or sucking water/cleaning agent through the capillary with a syringe. For super-heavy clogs we also use fused silica wires that nebulizer manufacturer sells. There were clogs that I wasn't able to resolve in any other way.

As for the tendency to clog - I can also +1 the notion that the system seems to be clogging more often after we have upgraded to CyTOF 2.1. The reason for that is unknown to me.

@Mike - how does your sample transmission efficiency on beads compare between super sampler and PSI? I am having a feeling that SS underperforms there but haven't thoroughly tested that.

Best,
Vinko

[mleipold]

Hi Vinko,

We haven't done a ton of head-to-head Supersample vs PSI comparisons. In the few that we did, we didn't notice a significant difference between the two while running cells.


Regarding cleaning the nebulizers: we keep our nebulizers stored filled with 5% citranox, submerged in 5% citranox. At the beginning of the day, we backflush through both the sample port and the gas port, then forward-flush through the gas port to examine how steady and coaxial (both X and Y) the stream is.

If there are multiple streams, or anything else funny, we backflush both ports with water and then carefully backflush both ports with concentrated nitric acid (be careful!). The 2mL microcentrifuge tubes that come with spin filters are the perfect length to use: the gas port sidearm keeps the nebulizer tip suspended 1-2mm above the bottom of the tube, so you only need 500-1000uL of conc nitric in the tube. Leaving it overnight at room temp like that is usually sufficient to clear anything.

Alternatively, Meinhard sells the Pearl Nebulizer cleaning apparatus. It's worked really well for us; in that case, the cleaning agent is 80:20 5%citranox:Methanol. 1-2 flushes usually takes care of it.


At the end of the day of sample running, we backflush with 5% citranox twice through each port, and then leave filled with citranox, submerged in citranox, until the next use.


Mike

Mike

[desireeBCRT]

Hi,

we also have a lot of clogging issues after we upgraded our CyTOF2 to Helios. There are two things that seemed to help in our hands: 1) improving sample quality and 2) changing from silica lines (sample and probe line) to flexible PEEK-sil lines.

1) We are a core unit and get the samples fully stained to us. To make sure we get samples of good quality we go through *all* protocols with the users - from isolating single cells to when the user brings the samples to our lab.

- we specifically check that cells are not being centrifuged too fast as long as they are alive (max. 310g, 10min, 4°C). This improved sample quality a lot and reduced the risk of clogging.
- for cells isolated from tissue (e.g. mouse spleen cells, human gut cells, etc.) an extra 5min benzonase step (37°C in cell culture medium) has been introduced.
- we filter the cells through a 20µm filter directly before the run (making sure to verify before that the user is not losing cells of interest of course)

2) The other thing that helped was a change of sample and probe line from silica to flexible PEEK-sil material. It is a spare part from a BD FACSVerse that I got by chance from our service engineer. I connected it with the connections that came with the Fluidigm lines. It goes without saying that this not a practice that has been recommeded by Fluidigm but as it really helped a lot in our hands I thought it is fair to mention it here.

With these lines actual clogs are rare and if a clog builds up it can be rather easily removed by backflushing the sample/probe line from the grounding nut end. We did not have any clogs in the sample capillary or nebulizer if we made sure to stop the sample when a clog was building up (higher pressure on PSI, events/sec going down, etc.) and backflushing sample/probe line.

While this is working well for us since some weeks now I have to say that I did not extensively test going back and forth with silica lines and PEEK-sil lines and I am not going to try again with the silica lines as long as the system is running …


Desiree

[ChristophS]

Hi Desiree,

we also experience quite some problems with the PSI lines but unfortunately do not have FACSverse around, thus no spare parts to swap in.
Would you happen to know the length and inner diameters of the ones you are using?
Sigma has them in 10cm, 20cm and 50cm length (you cannot cut them yourself) with IDs of 25um to 150um.
I would think 20cm for the probe line and 50cm for the sample line would be good length but for diameter I am not sure whether smaller is better (clogs likely to block them early, rather than the nebulizer?), what do you think?

Best regards

Christoph