I think it would be interesting to see the protein content in these vesicles. If it is a limited subset of proteins that appear in these vesicles, then these are 'true' vesicles and not a reconstitution of the phospholipids from the cell debris.

You can't really be sure. You need to do electron microscopy and look for vesicles. (usually ~50nm spheres, and you can see the membrane) Then you just use the same purification protocol for future preps and hope they are the same.

I don't think you can be sure at all. However, "normal" MV come from the outer membrane, whereas those formed from lysed bacteria include inner membrane vesicles as well, so you can assay for IM contamination using some assay for respiratory chain enzymes or other integral IM protein.

Also, I have a vague recollection that the OM and the IM exhibit differential solubilities in some common detergent (Tween, Triton or something else found in every lab), so you could try that too (solubilize, recentrifuge, then see what remains).

Thank you for your replies,

 Andrew M Frey, I  do observe spheres by TEM, but again, I don't know whether these are vesicles the bacteria released, or re-formed cell debris.

Alejandro Martin,  I actually work with Gram-positive bacteria, so all vesicles I observe are IM vesicles.  I don't think any markers have been found for these to check for the enrichment in "normal" MV.

You should ilude control of dead cell, which can not activly produce vesicles

There is a good understanding of cell death and lysis of bacteria in log, post-log, and stationary phases. During log phase there shouldn't be a significant contribution of lysed cells to add to your membrane vesicle population, although some might be found, somehow. Lysing cells will grow in the population at other phases. Regardless, could they be an equally-important part of the population? Consider the existence of MV from lysed cells as more than accidental.

Additionally, you may be able to circumvent total cell lysis by working with a mutant strain. Dr. Kenneth Bayles at the University of Nebraska in Omaha, NE, USA likely has S. aureus lysis-deficient isogenic mutants that you could use for experiments or making similar mutants in your species/strain of interest.

Elhanan, this could be a good control, but not knowing how many cells actually die and contribute to my MV population, hard to kill exact number cells for proper comparison.

Ethan, I used to work with overnight culture, so concerns were quite high. Now I'm witching to mid-log and for now will assume that contribution of lysed cells is insignificant. However, I do know that my strain released phage so there is an inevitable cell death even during the early stages of growth. I'll have a look at the S. aureus paper, thank you for the reference. but I think the whole point is actually study vesiculation rates from established pathogenic strain, rather then lysis-deficient mutant

PMID:   24826903

Look at the control in this article

Extracellular vesicles produced by the Gram-positive bacterium Bacillus subtilisare disrupted by the lipopeptide surfactin.

Brown L, Kessler A, Cabezas-Sanchez P, Luque-Garcia JL, Casadevall A.

Mol Microbiol. 2014 Jul;93(1):183-98. doi: 10.1111/mmi.12650. Epub 2014 Jun 4.

If you take dead cells, and take them through the same condition as your experimant, and dont get vesicles, it apparent that the vesicles in the real experiment are not form passively by mambrane enclosures or something like this.

It is hard to distinguish the MV ( or Plebs) from MV made from lysed cells. However, the best approach is to determine the percentage of the cells that lysed by assaying the supernatant for the presence of a constitutive intracellular enzyme - like glucose-6-Phosphate dehydrogenase (G6PD). This way you can determine the percentage of lysed cells and their potential contribution to your total isolated MVs. Of course you need to determine the total activity of G6PD in a given volume of the culture (i.e., Cells + supernatant). Therefore, % lyses = (Total G6PD activity per ml of culture / G6PD activity per ml of supernatant) 100. Good Luck!

From my experience it is good to verify in which fraction after gradient you have the highest amount of OMV. I did it using DLS assuming that typically OMV are 50-150 nm. Lysis products are usually much bigger. I also did mass-spec analysis of OMV and bacterial cells. There is significant reduction of house keeping proteins and central metabolism so I assume that I am really working mostly with OMVs. But I also like Hosni M Hassan idea. 

Cheers!