In order to investigate why the SmartFlares seem to be fluorescent in puncta within the cells, we designed an experiment to try to inhibit endosomal maturation, thus (hopefully) preventing SmartFlare processing and release of fluorescence.
Endosomal maturation occurs through the sequential fusion (and fission) of compartments, a process which is regulated by many things including the presence of Rab GTPases, the phospholipid composition of the vesicle and the pH within the endosome. The pH is decreased partly through the action of the Vacuolar ATPase (VATPase) that drives the pH down to create the highly lytic lysosome.
One hypothesis for why both the VEGF and Scrambled SmartFlares are fluorescent in punctate structures is the idea that nucleases resident in the endocytic pathway cleave the oligonucleotides on the SmartFlares, releasing (and thus de-quenching) the fluorophores.
A (seemingly simple) way to test this hypothesis is to inhibit the endocytic maturation of the SmartFlare-containing vesicles and to see if signal is still evident. Unfortunately it’s very difficult to stall the endocytic pathway without having the knock-on effects either on endosome stability (IE disrupting the endosomal membrane), endocytic uptake (preventing uptake all together) or flat out killing the cells.
Briefly, the experimental protocol was to:
- Seed HeLa cells at ~50,000 cells /mL and allow to adhere for 4-6h
- Treat the cells with SmartFlares, 10kDa Dextran (to label the entire endosomal pathway) and with and without 100μM Chloroquine and 500nM DMOG
- Incubate for 18h
- Wash and image the cells live
As always, the results of the experiment can be found on our Open Data Repository date stamped [2015-05-22] and [2015-06-04]. These experiments only use the VEGF SmarFlares and don’t include the Scrambled or Uptake Controls.
Below is the VEGF with no drugs (which we’ve seen plenty of times before):
Then the same conditions, but with the addition of 100uM Chloroquine added at the same time as the SmartFlares:
Most of the chloroquine treated cells show a considerable degree of vacuolation:
This has been documented in the literature for over 40 years. The vesicles are thought to be lysosomal in nature and where the chloroquine accumulates in the cell (ibid).
Another interesting thing is that the Fluorescence Intensity of both of the SmartFlare and dextran channels is decreased by about 65%
This was very roughly quantified by taking the background subtracted integrated density for the whole image and dividing it by an estimate (whence comes the roughly) of the number of cells per field (for the 8 fields per condition). This resulted in a mean (± SD) of 3548514 ± 775404 for the Control and 1205146 ± 563586 for the chloroquine treated sample. Even with this very rough estimate, the p-value for a non-paired, two-tailed, equal variance T-Test is 7.36 x 10-6.
When normalised to the control, the data look like this:
In order to better image the localisation of the signal (and to see if there was any signal in the cytosol), the latest chloroquine experiment was imaged using two different acquisition settings. This dataset is all imaged using the same settings, while this dataset (just for the chloroquine conditions) was imaged at a higher contrast setting (by increasing the laser power and PMT gain). The stage positions are the same so you can directly compare similarly labelled images.
Looking at the high contrast imaging, you can see that some of the SmartFlare fluorescence is present in the dextran-positive compartments (as we’ve seen before, but perhaps more? Analysis to follow), and there really isn’t any detectable diffuse signal.
The next post will deal with possible interpretations of the data.