This is...entirely to be expected. Larger proteins move more slowly, because they are larger.

When you run a gel from top to bottom, the small proteins run right to the bottom in the time taken for the large proteins to barely move at all. And....this is the point of running a gel, after all.

The exact same rationale applies to blotting from that gel to a membrane: small proteins will move onto the membrane very quickly, larger proteins will take a lot longer.

This is most obvious with your prestained markers, but it's applying across the gel as a whole.

And this is normal behaviour: it's nothing you're doing wrong. You can change the behaviour by using lower acrylamide% gels, by increasing transfer time or current (or all of the above), but maybe this isn't necessary: the key question is "what is the size of your target protein?"

For example, I work with dystrophin, which is annoyingly huge (427kDa). To even get this protein to move significantly into the gel, we use hand-made 4% gels and run for two hours, sufficient time for anything bigger than 30-40kDa to have already run all the way through and fallen out the bottom: we accept this and just don't try to probe for dystrophin and anything small at the same time (we use vinculin as a loading control, as it's ~150kDa).

For transfer, similarly we use 250mA per gel, and run the transfer for 4 hours in an ice bucket. We also omit methanol from the transfer buffer (it tightens the acrylamide matrix and makes protein transfer more difficult), and include more SDS so the dystrophin doesn't precipitate.

On these gels, prestained high molecular weight markers transfer just fine, because we're optimising the whole system to ensure exactly that.

If we were interested in, say...GAPDH, none of this would be necessary, and as long as anything below ~150-200kDa transferred ok (as shown by the prestained markers), we'd be fine.

You're interested in something that 17kDa and a loading control that's 32kDa, so it does not matter in the slightest that your high mol wt proteins are not transferring completely: all the proteins you are interested in will transfer to completion.

Hello

First and foremost, you need to find the concentration of total protein in the sample using Bradford assay or BCA assay. Then you need to load the samples on the gel in terms of ug and not in terms of ul. The amount of protein to be loaded on the gel is around 20ug to 30ug per well. Sometimes when the target protein is low in abundance it may be necessary to load higher amount of protein on the gel.

PVDF membrane is a better choice for low molecular weight proteins as it has a greater capacity to bind proteins than Nitrocellulose membrane.

Use of semi-dry transfer is better than wet transfer as semi-dry is less vulnerable to over transfer.

In semi-dry transfer be sure that the membrane and filter paper sheets are cut to the gel size without overhangs, and the gel and filter paper are thoroughly equilibrated in transfer buffer. The use of extra-thick filter paper in order to hold more transfer buffer for transfer would be appropriate.

A short rinse (15-30 secs) of PVDF membrane in 100%methanol before transfer is essential as it will hydrate the membrane and allow improved transfer and protein binding.

I hope this helps.

Best.

Hi Trung, Are you re-using a transfer buffer? Try soaking the PVDF membrane, papers/sponges and gels at least for 10 mins in the transfer buffer before transferring to the machine. The screws of the semi-dry transfer machine should be tight enough. Also re-check the methanol amount in the buffer. You also need to semi dry the papers since it absorbs the buffer before running the machine. Try using a low voltage (15V) and run for 45 -50 min (for low molecular weight protein).

Its possible you are not loading sufficient protein as the cell number seems very low to me. As stated above quantify your protein from the lysate. If this is an endogenous protein then levels can be very low and you need to load more. Suggest completing an initial gel with different loadings going up to >10+ microg

Your transfer look inefficient to me: I would not expect HMW standards to be as prominent as they are on your gel after 1 hour transfer? Stain the PVDF with ponceau S or amido black to confirm transfer of LMW protein and to confirm you are getting uniform transfer. These are reversible stains and do not effect WB. Before processing your gel check to see if standards have transferred, if not then continue with transfer. As stated above make sure your buffers are correct.

Dear John,

Thanks for your comment. I check the gel after transferring and the whole lane still remained on the gel, not only the high molecule weight. So, I can not understand what's happening.

In both your pictures above, the transfer looks fine, so perhaps including some pictures of what you're actually seeing would be helpful.

More to the point, is your actual western working? Because it seems like you're worrying too much about things that might not even matter.

Dear Malcolm Nobre ,

Thanks for your infomation.

Dear Kashi Raj Bhattarai

I used the fresh transfer buffer. Your suggested current looks like it is used for wet method, right? I'm thinking about trying this way.

Dear Ian R Wilkinson

Now I don't have any idea about total protein amount that I applied to a well. That's a problem for me now. Thanks for your sharing.

Hi John Hildyard

I attached photos, please take a look at them.

My purpose is compare the different expression level of casp3 between cells w/ and w/o herb extract. Because it is an relative quantitation method, I think I should transfer all the target protein to the membrane (casp3 and GAPDH for ref). So, I'm stuck with this.

Honestly, that looks fine to me.

Comparing the background in the gel before and after, it's pretty clear a lot of protein has transferred, and the ponceau staining of the membrane reflects this. Coomassie is pretty sensitive, and it looks like you're loading a LOT of protein, so some residual is to be expected.

As long as your samples are both run on the same gel, you should assume they're comparable, because any differences in transfer are going to be a function of size rather than a function of specific lane.