I tried handling the beads more gently: no vortexing and slower pipetting. Surprisingly, this reduced the clumping.
The slowing it down here doesn’t surprise me with beads, keep us posted!
Updates?
1 Like
Sorry to have gone silent. We were busy going into the holidays and then I was away. I haven’t had as much time as I’d like to test but I think moving away from using the on-deck shaker to resuspend the beads has been helpful.
I’ve done a round of testing with Dyanbead MyOne C1 streptavidin (our standard bead for this method, hydophilic) vs T1 beads (hydrophobic). Using pipetting alone the C1 beads have much less clumping than I observed when using the shaker and the T1 beads are nearly monodisperse. This is promising enough, that I’m going to put the new pipetting steps into the full day of protocol and try building some real libraries.
Separately, I tested whether the wash volumes used in the Bravo protocol are problematic by themselves. Nope, beads handle just fine in reduced wash volumes. No clumping.
Tentatively, it does appear the shaker may, per se, be causing trouble. Does the magnetic drive affect the beads? Static from shaking?
thanks,
At length, I tested my latest version of the protocol. Avoiding shaking and more extensive pipetting appear to have resolved the clumping issue. Oddly, my yield is still only ~25% of manual yield but that could be for molecular biology reasons not related to the Bravo.
In parallel, I’ve been testing a scaled down manual version of the protocol in PCR strips with conditions very similar to the Bravo protocol. Yields are a little lower than manual prep in 1.5mL tubes (80-90%) but much better than on deck.
I plan to test by starting a batch of samples on the Bravo and switching one sample after each step to the manual protocol. Any thoughts on factors most likely to affect yields when moving to automation?
I’m working with Brooks FrameStar low binding plates (4TI-LB0960/RIG) because Eppendorf twin.tec plastics became scarce a few years ago but I don’t know how low binding they really are. Does anyone have experience using them for low input nucleic acid handling?
Good to hear from you again! Regarding your low yield… Non-specific binding in the tips would be my first suspect. Again, I never use the pipette tips to pipette mix. It adds a lot of plastic surface area.
For the shaker… What kind of shaker do you have on your deck? The amplitude, especially with your plate and low volumes, may not be sufficient. It’s an interesting hypothesis but I have not yet heard of interference from the shaker motor causing issues with the beads.
Thanks, binding to the tip material is a good thought. I’m so accustomed to using low binding resins for my manual tips I hadn’t given much consideration to the Bravo tips. In my case, the possibility of loss to the tips may be lower because my RNA is biotinylated and bound tightly to the beads. Insofar as the beads are all removed from the tips by the magnet–I have a pipet mix on the magnet to recover stragglers–the RNA should stay in the wells though whether free or stuck to plastic isn’t clear. I do tend to see a light coating of beads along the well walls by the end of the protocol.
I have an Inheco Teleshake and was vortexing 30uL of buffer as fast as possible without over-topping the wells (1200-2000 rpm depending on the buffer). This reduced clumping relative to previous pipetting-only approaches but didn’t eliminate it.
I’m not sure whether hypothetical magentic interference from the shaker itself could be an issue or not. At the start of my protocol, I bind my RNA to the beads for 20min with constant shaking at 900rpm without a problem. Maybe it’s only at the highest speeds or maybe it’s a red herring and the clumping is driven by something else. Static? Sheering forces?
Just some food for thought along the shaker optimization - possible red herring/rabbit hole.
I always refer to Qinstruments knowledge base when evaluating shake speeds - they’ve done some great work evaluating force required for proper mixing (Link, see the “Knowledge” tab). Keep in mind that orbital radius influences shake speeds tremendously between shakers.
Further, they have some papers on optimizing mixing such as this one (Link) and of particular interest might be the profiles of shaking liquid in Case A, B, & C. If optimal shaking speed can not be obtained due to volume or any other factor, a good trick is to add required volume in portions (e.g. add 500µL, mix fast, add 500µL more, mix slower) or to add volume while shaker is already in motion. Both of these solutions alleviate the problem of providing lift in Z direction when the shaker itself only provides X/Y movement. Using this scheme has saved me many headaches with magnetic bead based chemistry, but your miles may vary. On a similar note, any “wobble” of the shaker nest in Z greatly undermines the mixing efficiency of the shake. I don’t have any papers for that observations, just data from my own tests using worn out shakers.
Hopefully this helps!
3 Likes
Really handy resource. thanks!
Any sense of how the recommended speeds should be adjusted when using 3 mm orbit? All of the values on the Qinstruments knowledge tab are for 2 or 1.2 mm orbits?
You can use an angular velocity formula to solve for a larger radii.
I made an Excel worksheet a while ago that I might still have… I’ll try to find it tomorrow and upload it here. It wasn’t super fancy but it automated the math well enough. Worst case scenario, you can email QInstruments and they have a few internal white papers that they’re usually happy to share which helps solve for 3mm shake speeds. I think Rolf was the guy I talked to way back when - really awesome guy.
Late edit - the excel sheet is in an external HD that I can no longer find after moving. It might have been lost to the void. If I ever find it again I’ll be sure to upload & share.
1 Like
I’m having a similar issue but with DNA and low yields, I plan to use an alternate magnet (currently using Alpaqua ring) as suspect overdrying could be the problem and our pellets are small.
You mention it has 6-8 hours of handling, could all the steps where its not covered with liquid be contributing to this clumpiness?
For improving the yield, we found that leaving a small bit of the previous wash supernatant behind (~20ul), doing a quick shake and then adding the next wash onto this (multi-dispense then shake) has helped somewhat. Of course, this depends on your application, wash volume, # of washes etc. as to whether this would help you.
We have the Alpaca 96S magnet in our lab as well, but too be honest I am not to fond if this ‘ring’ magnet. On our Tecan I have decided to swap it out for the DynaMag 96 Side Skirted plate. This one pulls the beads to one side and is therefore easier to loosen with the elution buffer.
As for the drying. We try to remove all the EtOH with a 200uL pipette tip and then when all wells are done with the FCA we try once again to remove any remaining liquid with a 50uL tip. This works quite well.
Thanks @ClaireB, after a bunch of trial and error I ended up with a similar solution. I leave 30uL on the beads at all times which is a volume I’ve found for my solutions/plates that keeps the bead ring submerged.
I also took an additional step to make absolutely sure the beads don’t dry but I don’t really like it on principle. To minimize the time the beads spend dry while removing supernatents, changing tips and picking up the next wash, I’m handling the last bit of wash supernatent and the next wash buffer in the same tip. I aspirate the next wash buffer, a big air gap, and then the residual 30uL of the previous wash from the beads. It is much faster but I’m sure it reduces my washing power and it makes me a bit nervous the air gap will collapse.
Ultimately, keeping the beads from drying did help me determine that the clumping I was getting wasn’t due to drying. It appears, somehow, to relate to either shaking or the shaker. Vortexing at up to 2000rpm was unable to prevent clumping so I reverted to an all pipet mix which preserved the bead quality far better than any of my vortex-based attempts. No idea what’s happening. G-forces? Static charge? Subtle magnetic field from the shaker? Vortexing on an analog lab vortexer has no effect on the beads.