A Gene Frame replacement and an unexpected lesson in physics

by Christian Rudolph

Some months ago I wrote a post asking a simple question: why had Fisher Scientific discontinued the Gene Frame, a product that did its job perfectly well at a relatively reasonable price? It was, I noted at the time, a corporate mystery for the ages.

The mystery remains. But after quite a search, I have at least found a replacement, and it arrived in a shiny white envelope.

The replacement

The frames are made by Grace Bio-Labs and sold as SecureSeal. The headline specifications look familiar – 15 × 16 mm internal, 25 × 25 mm external, 0.25 mm thick, which is what Thermo quoted for the Gene Frame.

That is where the resemblance ends. These are not the same product wearing a different label. They are white rather than blue, the internal cut-out is smaller, the plastic covers are quite different, and consequently they handle differently and hold a different volume.

Two of those differences matter. One is an improvement, and one is a nuisance.

The improvement nobody will write a press release about

To understand this, a word on how we use these frames.

They are adhesive on both faces and come with plastic covers on each side. One cover is solid, closing the whole aperture; the other is cut out, covering only the sticky frame and leaving the window open. You peel off the solid cover, press that face onto a slide, pipette in molten agarose, and immediately cap it with a second slide.

The remaining cover is what stops that second slide bonding to the frame – it is acting as a release layer. Once the agarose has set, you slide the top slide off, peel away the open plastic cover, dry the pad, add 1.5 µl of sample, and apply a coverslip, which sticks to the freshly exposed adhesive and seals the chamber.

Simple enough. Except that on the Gene Frames, the plastic covers were thick, and the solid one was thicker than the open one. Reaching for the solid cover, one would as often as not peel off the open one instead, ruining the frame. Getting it half off, having it re-adhere, and ending up with a rippled frame that spoiled the slide was also not a rare occurrence. There was a technique to it – a particular use of the fingernail – and it worked well enough once learned, but it had to be taught to every new student.

A consumable that requires training to open is a badly designed consumable. But at ~60 pence per frame the occasional waste of a frame is not the end of the world.

Because the Grace frames have much thinner covers they simply peel. I had no trouble with a single one. This is not the sort of thing that appears in a product comparison, and it is worth considerably more than it sounds.

The nuisance

The internal cut-out on the Grace frames is 14 × 14 mm rather than 15 × 15 mm, and the covers, as noted, are thinner. The frame depth is identical. But the volume the well accepts is substantially less, and both differences contribute.

We had established over the years that a Gene Frame took 96 µl of molten agarose – enough for a solid pad, with a little air left over, which matters for time-lapse work, because it generates an oxygen reservoir. Here is the interesting part: 15 × 15 mm at 0.25 mm depth is a volume of about 56 µl. We were routinely pipetting in nearly twice that. The extra was sitting in headroom created by the thick plastic cover, which stood proud of the frame and gave the liquid somewhere to go. We were never filling the frame. We were filling the frame plus the plastic.

Which means the 65 µl volume printed on the packet was never the number anyone actually needed. It describes a well that only exists if the frame is flush and empty, which is not how we are using it. It simply happens to land much closer to reality for the Grace frames, because there is less plastic sitting on top.

Sixty-five microlitres, while definitely not enough for the original Gene Frame, was not bad at all. My suspicion is that it is slightly on the low side, and that 70 µl will prove to be the sweet spot, but we will test volumes such as 70 and 75 µl properly.

And this is where the improvement and the nuisance turn out to be the same thing. The thick plastic was the worst feature of the Gene Frame and simultaneously the reason 96 µl was the magic number. Remove one and you lose the other.

Everything we had established empirically is now wrong.

The volume is not the only casualty. We had also established, empirically and precisely, that Gene Frame pads needed drying for 14 minutes at room temperature – or five minutes at 42 °C, and not a second longer.

Those numbers are now void. A smaller volume of agarose in a shallower well will dry faster, and drying time is not a detail one can hand-wave. Under-dried pads shrink during imaging, which pulls cells out of focus, which is somewhat of a nuisance both for snapshots and time-lapse experiments.

My guess is 10 to 12 minutes at room temperature might work best. But it is a guess, and re-establishing it properly needs more testing than I had time for.

I should be equally plain about something else: I have not yet imaged cells on these pads. The pads set correctly, they behave as they should, and once the coverslip goes on, the frame seals very well. Everything I have seen suggests they will work perfectly well, certainly for snapshot microscopy. But suggesting is not demonstrating. That, along with the drying optimisation, is work still to be done, and it will get a post of its own in due course.

The price

Now the part that changes the calculus entirely.

The list price for the Grace frames is £155 for 25. For the same money, we used to buy something in the order of 100 Gene Frames.

There is one small consolation. The Gene Frames came bundled with flexible plastic coverslips, which Grace sell separately. For microscopy these coverslips are completely and utterly useless, so not receiving them is, if anything, a further improvement. But it does not begin to close a fourfold gap in price.

At sixty-odd pence a frame, one does not think very hard about whether a frame is necessary. At four times that, one does.

What we used to do instead

Which sent me back, rather usefully, to a technique we used routinely when I was a postdoc, and which requires no frames at all.

Take a plain glass slide, and pipette 800 µl of molten agarose solution onto it, quickly. The liquid forms a puddle that runs to the top and bottom edges of the slide but not the sides, producing a heap. This sounds crude, and it is, but the important feature is that the centre of the heap is smooth and very nearly flat. That flat centre is the entire trick.

Level matters here. I set the slide on a perspex plate and use a spirit level, because the shape of the pad depends on the bench being genuinely horizontal, which our lab benches are definitely not.

Leave it to dry for about 18 minutes at room temperature. The pad will be larger than the aperture in the microscope's slide holder, but once set it can be trimmed to shape with a scalpel blade. Add the usual 1.5 µl of sample, rotate the slide gently until the liquid has spread and evaporated, and then simply balance a coverslip on top. No adhesive, no seal, nothing else at all. The capillary forces between the pad and the coverslip are strong enough to hold cells firmly in place, even through longer exposures.

The cost of this method is a slide, some agarose, and a coverslip.

Three caveats, one of which only occurred to me while writing this

The first follows from the shape of the pad. A heap has curvature. Image at the centre, where the gradient is shallow, and all is well. Move towards the edges and the gradient steepens, to the point where half a frame can be in focus while the other half is not. Levelling the surface reduces the problem but cannot eliminate it, because one is contending with the shape of a drop, not merely the tilt of a bench.

The second is that this is definitely not a method for time-lapse microscopy. Time-lapse requires that the pad does not dry, and a coverslip resting loosely on a dome is not a seal. For that, there is no substitute for a frame.

The third occurred to me only as I was writing this post, which is a useful reminder of how easily one carries an assumption for twenty years without inspecting it. When we used this method, we were working on an upright microscope. Gravity was quietly holding the coverslip against the pad. These days we use an inverted system, as most people do – and on an inverted microscope, gravity is pulling the coverslip away from the pad, with nothing but capillary force resisting it.

I have checked, and it works in principle: the coverslip does not simply fall off. But that is precisely all my test establishes – not less, and certainly not more. Whether it holds steady through the longer exposure times that fluorescence imaging demands is a different question, and one I need to answer properly before recommending the method to anyone using an inverted scope.

So: for an upright microscope I can say with complete confidence that this works, because we did it for years and published the images. For an inverted one, watch this space.

What were we actually buying?

Which leads to a thought I would not have had if the price had stayed where it was.

The frame was never doing anything the physics could not do unaided. Agarose sets flat if you let it. Capillary action immobilises cells whether or not there is adhesive around the edge. What the frame provided was two things: a defined, reproducible geometry, and a seal.

The seal is genuinely necessary, and for time-lapse there is no way around it. There is also no doubt that it helps with snap-shot microscopy by making it easier. But easier is not necessary. The geometry is a convenience – a real one, worth paying for at sixty pence, but worth thinking rather harder about at four times the price.

A discontinued consumable is an irritation. It is also, occasionally, an invitation to ask what the thing was actually for.

Similar and related blog posts:

• Thermo Fisher discontinuing the Gene Frame
• An environmental chamber for time-lapse microscopy
• An excellent LED illumination system for microscopy