Testing for antibacterial properties of less polar compounds

[BabcockHall]

I am collaborating with a microbiologist to test some compounds to see whether or not they possess antibacterial properties.  These compounds are either neutral or bear a single negative charge near pH 7, and they have aromatic rings.  The compounds we have tested so far are soluble in DMSO, but modestly soluble in ethanol (less than 10 mg/mL), and probably not soluble in unbuffered water.  The assay begins with adding the compound to a disk that is placed in contact with bacteria.

 

My question concerns solvents that would be appropriate to use to dissolve the compounds.  I assume that one control is to try the solvent without the compound of interest.  Are there any solvents that are known to be so toxic in this sort of assay that they are best avoided altogether?  If two solvents had about the same polarity, is it better to choice the more volatile or less volatile solvent?  Is there any other property that we should consider?  So far I looked in a microbiology laboratory manual, but I did not find anything.  Perhaps there is a good reference that someone could suggest.  Thank you.

[CharonY]

AFAIK there are no established methods to help you there. For the most part folks stick to solvents that are only mildly toxic for the organism (such as ethanol) and hope for the best. Controls are, as you mentioned, the solvent without the compound of interest. Perhaps trivially, another rule of thumb is getting the stock as highly concentrated as possible in order to minimize the volume of the solvent to be added to the assay. One issue of course is that once in media, it is often unclear how the compound behaves, things like micelle formation or aggregation can become issues, which may need to be analyzed in a cell free system.

Specifically DMSO can be problematic as it enhances membrane permeability quite a bit and if you do molecular assays (such as proteome analyses) you will often see quite some drastic changes in the bacterium in response to it alone. However, while it may be dubious for real-life use, in most cases folks are content (at least for publication) if one shows significant effect above solvent alone.

[hypervalent_iodine]

If they’re soluble in buffer, is it possible to instead measure MIC with microbroth dilution? I’ve never been a great fan of disk diffusion personally. There are just too many other things that can influence zone of inhibition. 

[BabcockHall]

Inasmuch as we are working in collaboration, we are probably limited to disk diffusion for now.  Because the compounds are soluble in DMSO, I was going to try acetone next for one of them.  That way the acetone will simply evaporate, leaving the compound behind.  The other compound came into a solution of K₂HPO₄.  If we did our calculations correctly, we created a phosphate buffer.

[hypervalent_iodine]

I believe any volatile compound it dissolves in is fine then, just so long as it’s fully evaporated before you place it on the agar.

[CharonY]

I somehow missed that it has to be a disc-diffusion, so much of what I ranted about earlier does not really apply (no idea how, it is prominently in OP). DMSO often takes quite a while to evaporate fully.

[PhilGeis]

The MIC/ZOI and other protocols mentioned above can generate data that may show "potential" efficacy but will be largely of questionable relevance.

Many solvents in this regard have some antimicrobial effect -such protocols suffer from  the inability to address additive, synergistic ot even absolute nature of combination.  That is breaking out solvent + test compound  vs test compound.  A solvent-only control does not answer that question.  DMSO is prob less of an issue in this than ethanol.

More importantly, using a solvent  to shoehorn an insoluble compound into a classic MIC/ZOI test is common but offers a result of questionable relevance.  Dispersing the agent in agar is also problematic as it effectively establishes an emulsion in the water colloid.  The target bug may not even "see" the compound.  But you can still get numbers and report potential - if that's all you need at this point.

Do you have application in mind - or is this an academic effort?

[BabcockHall]

Thank you for some food for thought.  It seems to me that a volatile solvent (assuming it is essentially gone before the disk is put into place) solves many of the problems that you mentioned, possibly all of them.  Or am I mistaken?  We are in the process of making additional compounds and testing the ones that we already made.

[CharonY]

14 minutes ago, BabcockHall said:

Thank you for some food for thought.  It seems to me that a volatile solvent (assuming it is essentially gone before the disk is put into place) solves many of the problems that you mentioned, possibly all of them.  Or am I mistaken?  We are in the process of making additional compounds and testing the ones that we already made.

Yes and no. Getting rid of solvents completely just by drying works reasonably well for ethanol, but is difficult with DMSO. PhilGeis has touched on other points. 

[BabcockHall]

We are fairly early along in the overall project, and my thinking on this kind of testing is not overly sophisticated yet.  However, it could be that qualitative results are acceptable at this point.