Some extracts may dissolve better in water, or water buffered at a certain pH (acidic, neutral, or basic) than in DMSO if the extract contains a lot of charged substances.

Some people use Dimethyl Formamide (DMF) as a solvent and also Methanol as it is. But you will have to add a control to check the antibacterial action of such solvents.

DMSO is the best solvent for the purpose, but DMSO should also be used as control

Yes osborn loop Obey

Of course some of your extracts will not be fully soluble in DMSO, as the point of doing extractions wit different solvents is exactly to get different compound sets with different solubility characteristics. Trying then to dissolve all dried extracts in only one solvent will partly lose that diversity again.  The problem with dissolving in aqueous buffer, potentially, is to keep it sterile. Most organic solvents including DMSO, ethanol, methanol, acetone, chloroform, would effectively sterilse your dried pellet. Adding a sterile aqueous buffer might work if you can be certain the pellet, tube and pipette tip are all sterile - if not you risk all kinds of contaminations. One way around it might be to redissolve in a small volume of the same (or similar) organic solvent you did the extract in, ensuring that you lose nothing, and then diluting that down with a sterile buffer like PBS before applying it to the plate. 

An extract dissolved in water can be sterilized by passing it through a sterile membrane filter.

@Adam,

That is possible of course, but something I try to avoid doing. Ultrafultration of small volumes has a number of serious drawbacks. In order of importance: (1) Anything not perfectly dissolved at the moment of filtration is lost. If there is no activity you will ever know whether you have not filtered out the active component; (2) Even when dissolved a compound may bind to your filter and be (mostly) lost from the solution. I had this issue with some neuropeptides, rather unluckily; (3) No matter what filter you use, you are going to lose a certain volume of extract  with this method. Depending on how much you have and how precious/replaceable the substance is, that is a disadvantage. (4) The cost 0.22 microm filters may be significant depending on your budget and how many samples you want to test.

If you are using disc-diffusion assay, you should use a volatile solvent to dissolve your test extract (ideally the one into which extraction was performed), so that it will leave minimum traces on the disc after drying. DMSO is advisable, when you are doing broth-dilution assay.

As was said there is no true Universal Solvent, but I would argue, assuming you work with small molecules (and not peptides), to go with methanol (but read further). 

Re-suspend your extract in methanol, and absorb the liquid, absorb the liquid - step wise (10ul at a time) - on Whatman 3 paper disk. To facilitate removal of the methanol you can place your disks inside an open glass petri dish on top of a hot plate (but keep temperature low ~50C, to avoid thermodegradation). The idea behind the step-wise addition of liquid is to avoid saturation of the paper disk with liquid and spill out.

Alternatively, use which ever solvent your initial extraction was in for the resuspension. Just make sure to dry it out before you start. From that perspective methanol is better then DMSO or DMF due to much lower boiling point (67C, 189C and 153C accordingly).

Obviously, do the controls by absorbing the same total volume of clean solvent to a disk.

The attached papers might be useful.

Yoram

Thank you.... all

Can I keep the disc to evaporate the solvent (like methanol) in which extract is dissolved? 

Sure. Keep in mind some solvent (DMF, DMSO, n-Benzen) take a very long time to evaporate.