We’ve been struggling for quite awhile to get automated transient transfections of various mammalian cell lines to meet the same transfection efficiencies when done manually.
As of now we have a pretty good protocol that produces equivalent results to manual transfections but we can only do one plate at a time. If the MessengerMax is left for any amount of time on the automation the subsequent plates have lower transfection efficiencies. We aren’t sure if this is due to the MMax “settling out,” degrading over time, or any number of other factors.
Ideally we’re looking to automate transfections of up to ~20 96-well plates, which would take a few hours on our automation. So the MMax needs to be on the system for that amount of time and ideally behave the same for the first plate as it does for the last plate.
How do you treat the MessengerMax when manually preparing the cells?
I used to work with Viafect in smaller batch sizes (it was 14 plates I think) which then got automated. I remember the efficiency being lower, but still high enough for our measurements, and that was with treating the plates basically the same as when processing them manually (eg preparing plates with Viafect, medium and plasmids in each well so there was a 1:1 well ratio for cells vs transfection mix).
How do you treat the MessengerMax when manually preparing the cells?
That’s the million dollar question. If you ask 10 different scientists you’ll get 15 different answers. However, what has worked for both automation and manual is to combine the MMax and OptiMem (cell media) in a 1.5mL Eppendorf tube, vortex, incubate at room temp for 5 mins, aliquot into plates and then immediately or within ~5mins combine that with mRNA. The problem for the automation is that vortex step. We haven’t found a way to mix MMax and OptiMem in a manner that is equivalent to vortexing offline. We have attempted a vigorous pipette mixing, but it doesn’t perform nearly as well.
For testing we’ve mixed up enough MMax/OptiMem for a total of 3 plates. The first plate gets run immediately (T0), next plate 60 mins later, and then a third plate 120 mins after the first. Those second two time points have horrible transfection efficiency. The MMax/OptiMem really needs to be used within ~10 mins of a proper mix. Our next experiment will be to combine/mix MMax/OptiMem at T0, but also mix offline on a vortexer prior to the T60 and T120 time points. This is not amenable to our current automation, but we’re hoping to determine what sort of instrument we could add to the workcell to help us with this problem if it works.
We made master mix, aliquoted manually to plates and then placed the plates on the deck. This would not be a great strategy for a large number of plates of course.
We pre-made and froze our DNA plates for use later. When needed we just thawed them, made the master mix, aliquoted it and placed the plates on a Bravo to stamp the DNA into the MM for complexing. We’d reconfigure the deck with a trough of cells and destination plates for the final move of the mix into the cells.
I would have done this different with a bigger deck or if I only had an 8 channel.
We ran 3 plates at a time.
We have a workcell with a Precise Arm, refrigerated Liconic, 37C cell incubator Liconic, HRB AmbiStore (ambient plate/tips storage), two Bravos, centrifuge, sealer, desealer, some ODTCs, and a Formulatrix Mantis. Everything running on Cellario. We can configure one Bravo with a 96LT head and the other with a 96ST head if needed, and do occasionally for protocols that require high and low volume pipetting.
The fastest the system could work through transfecting a single 96-well plate from start to finish is about 10 mins. There can be a little bit of overlap with the second plate, so perhaps two plates in ~18mins. The goal is to have a walkaway protocol that could chug through 20 plates with decent repeatability from first to last.
I’ll mention that we have successfully developed a protocol for DNA transfection using JetOptimus, but our current science doesn’t need DNA transfection at the moment. We started that protocol development with Lipofectamine 3000 which is a two reagent system and found it not amenable at all for automation.
We do the same type of process with mRNA.
It sounds like you have more trouble with the scheduling of all the components. That they can’t move the plates around fast enough.
Do you use the Mantis to deliver your MM? That could be placed on something like a Spin Vessel if you think that mixning is the issue and you do not want to pipette to mix.
So far we haven’t found the MMax/OptiMem mix to be good for anymore than 10mins or so. When run manually most scientists mix enough of those for one plate at a time and then another aliquot gets mixed for their second plate, etc.
We have tried the Mantis for delivery of the MM, but never got good results, I believe this is because we can’t achieve a proper mix of the Opti and MM. Putting a pre-mixed MM/Opti mix on the Mantis would get us back to where we are now, it’s OK for ~10mins or so and then efficiency drops off. We do have a SpinVessel, it is next on the list for testing to see if we can get a protocol working. We’ve so far stayed away from it because the combo of it and the Mantis is a really unreliable system. The SV will easily shake the tubing lose from the Mantis chip.
Are you doing a forward transfection (adding transfection mix to plated cells)?
Might try a reverse transfection (adding cells to transfection mix). Sometimes there are things that make this unfeasible such as having to have your cells in a certain phase or confluency.
I’ve had better luck transfecting via reverse transfection as opposed to forward for messenger max.
Plate RNAs, add MMax/OptiMem to the plated RNAs, mix, let sit to allow the complexes to form, then add cells via a combi or something.
Worth a comparison if it’s possible to do it this way, even if it means you run a 30 minute process to plate your complexes then take it offline to the hood to add cells with a combi or something. The less handling you do with the complexes the better in my experience.
