Some sort of prep-scale chromatography would be typical. You could do normal-phase silica column chromatography, or reverse-phase C18 column chromatography, for example. In each case, you would use a linear solvent gradient, collect fractions, dry down the fractions, and test each one for potency in your MIC assay.

Adams answer is indeed a good one but you will need the proper setup (i.e. preparative HPLC or similar). If you do not have it you can start by dissolving a little extract in methanol, add 19 volumes of water and do consequence liquid-liquid extraction against Hexane, Chloroform, Ethylacetate and Butanol. Dry each fraction and see we're did the activity go. We extract in small volume of the above solvents against the water phase 3 times with each solvent.

Yoram

Out of curiosity were was the MIC values?

Thanks Adam and Yoram.

The MIC was 0.078 mg/mL

Yoram, How much methanol (mg/mL?) do you suggest I add to dissolve my extract. Do I then add 19 X  water to this volume to get a 1:20 dilution?

I will need a separating funnel? - are those solvents immiscible? What ratio/volume do you recommend for each solvent to extract 3 X with water? 

Also, besides methanol MICs - I have also done hexane (0.625 mg/mL), acetone (0.156 mg/mL)  and acetonitrile (0.156 mg/mL)  - should I take any of this into account for my methanol enrichment process?

The fact that aceton and acetonitrile gave better MICs suggest a polar compound. Never the less methanol is usually a better extractant, here apperently much better... You can also try MTBE.

I would suggest starting with ~200mg dry methanolic extract, redisolve it in 1 ml methanol and add 19 ml of water. Then add to the water 10ml of hexane, mix and allow to separate. Repeated twice and unite the hexane fraction. Dry in chemical hood, if possible with N2 flow. Then extract the water 3 times with 10 ml of chloroform, 3X with 10 ml of ethyl acetate, and 3X with 10 ml of butanol. Unite all fraction of each extractant and dry. Weigh the residue, devolve in methanol and do disk diffusion assay or MICs.

This will not give you pure compounds but it's a good start...

No the methanol exctract gave an MIC of 0.078 - better than 0.156 of acetonitrile and acetone :)

Thanks a lot for your help.

As I said, methanol is a very good extractant but not very specific see