We’ve so far, not seen any response of the VEGF SmartFlares to treatment of our cells with DMOG. Combined with the punctate distribution of the signal, we have no evidence that the SmartFlares make it to the cytosol in 18 hours. As we’re not sure if the problem lies in the SmartFlares or the delivery, I microinjected the SmartFlares directly into the cells thus bypassing the need for endosomal uptake.
On the volumetric variability of microinjection…
Microinjection is a versatile and downright cool technique for loading things into a cell. One of the problems with microinjection however, is that there is always variability in the injection volume, due to many different factors including the solution viscosity, needle angle and even the membrane composition and degree of convolution of the cell.
This is normally not a huge problem, except that with SmartFlare injection, we’re expecting a VEGF-mediated change in signal intensity. If the cells have more SmartFlare to begin with, how do we know that the signal is truly VEGF-mediated instead of having more or less SmartFlare.
… and how to deal with it
Thankfully we already have everything we need to correct for this. By injecting a mixture of SmartFlares with fluorescent 10kDa dextran, we can take into account variability in injection volume (as well as changes in cell shape and volume – although this is less of a problem when imaging on a widefield microscope). This way, changes in fluorescence are attributable to just the SmartFlare response.
As with previous experiments, HeLa cells were seeded in Ibidi glass-bottomed dishes and left to adhere. On the widefield microscope at the CCI, we have a Narishige micromanipulator with an Eppendorf Femtojet pressure system. For injection, 2µL of 1:1 SmartFlare:dextran were loaded into an Eppendorf Femtotip II. After confirming a good seal, the back pressure was set to 75hPa. Injection was carried out passively by introducing the needle into the cell and relying on the back pressure to expel the solution.
The data are date stamped [2015-06-26].
Technically, the microinjection was a success. The dextran is detectable as a diffuse signal in the cells:
After co-injection with a fluorescent dextran neither the Uptake Control (Cy5 labelled) nor the VEGF SmartFlares (Cy3 labelled) were detectable. In some cases a tiny amount of very feint signal can be seen in the corresponding channel, however this is likely bleed-through from the green channel.
So what’s going on?
Even as a stock solution (although remember we used it at 50% to incorporate the fluorescent dextran), the SmartFlares are probably not at a high enough concentration to be visualised after injection.
The product is not provided with a concentration, however, this can be determined from the extinction spectrum using the appropriate extinction coefficient (see Haiss et al for details). We’ve not done this yet, but do plan to.
For now, we can use a rough estimate of 5 nM based on visual observation (and lots of experience working with these particles). For a typical injection volume of 50 fL (an estimate based on rough cell-volume calculations and the degree of swelling post-injection), this would result in about 150 particles being injected. If there are 40 fluorophores per particle, this would be ~ 6000 fluorophores per cell. It might be that the sensitivity is too low to detect this number of molecules dispersed throughout the cell especially since this is a very rough estimate.
Authors: Dave Mason & Raphaël Lévy