I agree with Hussein and Philipp above. When you do not see a transfer it is due to 2 things: the proteins are still in the gel (you can confirm by staining the gel, post transfer) or most likely, they have transferred through the membrane (use a second membrane and stain it for protein). I have routinely used PVDF "Problott" which was sold by Applied Systems, to capture specifically, small proteins, for protein sequencing.
you can't detect any protein smaller then 50 kDa (with a dye) or a specific protein smaller then 50 kDa (with a antibody)?
Small proteins run faster, so reducing the transfer time ist a good idea. Test different times. Additionally you could place a second membrane behind the first. You also could try a PVDF membrane.
I agree with Hussein and Philipp above. When you do not see a transfer it is due to 2 things: the proteins are still in the gel (you can confirm by staining the gel, post transfer) or most likely, they have transferred through the membrane (use a second membrane and stain it for protein). I have routinely used PVDF "Problott" which was sold by Applied Systems, to capture specifically, small proteins, for protein sequencing.
One more thing u need to consider is pore size of membrane (used for transfer). for small molecular proteins it should be 0.2μm (http://www.bio-rad.com/en-us/product/nitrocellulose-membrane-0-2-um) so that protein will not come out of membrane. U need to keep less time for transfer.
I am generally using 0.45 µm pore size Nitrocellulose membranes to transfer proteins smaller than 50 kDa, in fact up to 14-16 kDa - and there is no problem. My technicians are going for tank blotting (standard 50 min, 100 V), whereas I am going for semi dry blotting - 45 min, 15 V using 10% gels (I am interested in proteins coming in between 30-150 kDa under these conditions). Both is working fine. When using 15% gels to get only low molecular weight bands like 14-16 kDa up to 70 kDa I reduce the blotting time to 35 min.
We are routinely doing Ponceau staining of the membrane after blotting to ensure that proteins are really on the same.
Unless getting the information how exactly you work it is tough to analyze your problem.
So first question - did you do Ponceau staining of your membrane that you can ensure that the protein is on the membrane? Or do you only try to detect some specific protein using some maybe not working antibody?
Second question - what are your exact conditions? What did you try?
And I would really suggest to go for Ponceau staining of your membrane, you can of course do complimentary Comassie staining of your gel, but when going for membrane staining you can immediately see air bubbles and if you have protein on your membrane or not.
Get rid of SDS from your transfer buffer if you have any and keep the methanol concentration at 20%. Small proteins are more likely hindered from binding to membranes in presence of SDS.
What kind of transfer method do you use? Semi-dry or tank blot? Semi-dry method seems more suitable for small MW proteins, and requires less (transfer)time.
I use liquid transfer. The worrying thing is that everything was working fine until suddenly only my ladder was transferring below 50 kDa - this was detected by Ponceau Red.
My transfer buffer is just glycine and Tris with 20% methanol. The transfer times I tried are: O/N for 16 h 70 mA, 3 hours 300 mA, 1.5 hours 300 mA.
If your ladder is transferrred this indicates that it shouldn't be a problem of the transfer itself as in principle all protein should be transferred. Do you have some protein leftover that you can load onto a gel, just to run it and check the gel itself with Coomassie staining if there is at least in the gel protein below 50 kDa visible?
It should not be any problem to transfer small proteins with wet-blotting or semi-dry methods. Probably, the condition of WB transferring is very hard, due to high current over the gel-filter or prolonged transfer time. You can do two things in order to overcome this problem.
(1) Double check that the condition of transfer is according ot literature. Check your buffers, just ot make sure that you do have to much ion.
(2) Used PVDF membrane instead of nitrocellulose form protein. Nitrocellulose membrane for DNA does not work for proteins.
I agree with Daniela, if you can see the ladder transferring then it is the case with your gel electrophoresis. Best would be to check if you have any proteins on your gel after the transfer using standard gel staining techniques. This would tell you if they stay in your gel and they are not transferring or they are just not there in the first place. Do you run a sample gel of your samples before you do Western?
as suggested by many people above, you can use 0.2 microne nitrocellulose (or PVDF)
Keep 20% methanol in your transfer buffer and if you are using semi try transfer 30 -45 mins is sufficient. But even if you transfer for 1hr 45 minutes you would still get enough proteins which are in between 25 to 50 kDa.
I transfer for 1hr 45 minutes at constant current ( equal mili amperes as to the area of gel in cm square)