Perfusion

Cells are cultured on circular glass coverslips on a matrigel substrate. A coverslip is stuck to the bottom of the chamber with vacuum grease and the chamber goes onto the stage of the inverted microscope which forms part of the microfluorimetry system.

The reason that the 1o perfusion looks like a 1ml pipette tip is because it is one. 8 tubes stuffed inside the tip connect to open 20 ml syringes each with its own tap. The tip is positioned using a coarse micromanipulator in line with and 3-4mm away from the centre of the field of view. So long as the solution in any one line is flowing, the cells "see" whatever is in that stream NOT what is in the chamber as a whole. Changing solutions is therefore practically instantaneous. The dead space in the pipette tip adds a 3-4 second lag between opening/closing taps and the new solution reaching the cells.

This system has been developed, tested and refined over many (many) years. It has no moving or mechanical parts (other than the tap). It is extremely reliable and glitches (air bubbles, poor positioning) are easy to spot and to rectify.

perfusion system

The 2o perfusion system is much more tempramental. We can't use the 1o perfusion system for IgG because it runs at 500 μl/min and we would very rapidly run out of IgG. The 2o perfusion is on a much smaller scale and requires about 5 μl/min. In principle, because the 2o perfusion pipette is much closer to the cells they will "see" only the contents of the 2o line even if the 1o perfusion is still running (which it is). In practice, when it works properly, this is what happens. Positioning of the pipette is absolutely critical and depends on two precision micromanipulators (one hydraulic).

perfusion system

Most of the 2o system contains mineral oil except for the glass pipette at the end of the system which is back filled with approximately 25 μl of test solution. Having exactly the same pipette tip every time is crucial for reproducibility. Fortunately, our patch-clamp pipette puller is good on reproducibility even on something as big as this.

Flow through the 2o system is driven by a syringe pump. The slightly higher pressure needed to drive flow out of the pipette tip is also sufficient to create occasional oil leaks. The syring pump and the plastic tubing connecting it to the glass pipette also introduce a significant hysteresis into the system, somewhere between 10-30 seconds.

Overall, the 2o perfusion system is based on a series of compromises. I would like it to be smaller and still closer to the cells. The manipulators we're using were bought for patch-clamp, so I could place the pipette on a cell (and keep it there). Unfortunately, smaller pipettes clog very easily and are practically impossible to back fill.

The 2o perfusion system has only a single line, which makes it hard to change solution. With practice (and luck) it is possible to replace the solution in 6-8 minutes.

So, assuming that the 2o pipette is the correct size and in exactly the right position and it isn't clogged and it hasn't been nudged onto the coverslip and broken and there are no leaks or air bubbles then we can test IgG.

The system is capable of generating false negatives (e.g. IgG appears to have no effect). These are understandable. The Ca2+ signals result from a positive feedback mechanism triggered by receptor activation. Sometimes it is possible to get a maximal response from a small stimulus. Under such circumstances the relatively small reduction in stimulus caused by adding IgG has no detectable effect on the Ca2+ signal. This is (fairly) easy to control by being sure to use an agonist concentration in the dose-dependent range. Unfortunately, the dose-dependent range is narrow and subject to change without warning.

For reasons that I really (really) don't understand, the 2o perfusion system is slightly prone to give false positives. If you put the same solution in both the 1o and 2o perfusion systems then you should see little or no effect on switching between systems. Some days, everything has an effect.

My best guess is that the cells are to blame. Some cells are "hair trigger" and respond to any stimulus, for example the change in flow rate/direction that is unavoidable when switching between perfusion systems. What I am doing is to intersperse control and test runs on a continuous basis and reject everything if it turns out to be a "sensitive" day. This works (I think) but slows things down.

Control data

data

cellsA "perfect" control. Carbachol (CCH) 10 μM from the 1o perfusion (upper treatment bar) stimulated an increase in [Ca2+]i. Switching on the 2o perfusion (lower treatment bar) which also contained CCH 10 μM had no effect. Increasing CCH concentration to 50 μM caused an additional increase in [Ca2+]i. Switching on the 2o perfusion, which still contains CCH at 10 μM, now caused the Ca2+ signal to drop. This experiment is proof of concept that a) activating the 2o perfusion system need not by itself have any impact on the Ca2+ signal and b) when it is activated, the cells respond to the 2o perfusion and not the 1o perfusion.

Each of three traces is the response of an individual cell. If you look closely at the image you can see appropriately coloured circular regions. The line graph represents the change in [Ca2+]i within this region.

Go To Top