Are you using a rotating shaker plate to oscillate your blot, or a linear rocker (see-saw action)?  The "bald patch" in the middle of your blot sounds distinctly like the effect you can get if using a rotary shaker where a shallow spot can develop in the middle of the dish.  This in contrast doesn't happen so much if you incubate your membrane curled up inside 50ml Falcon tubes revolving on rollers, or in square flat dishes shaking back and forth linearly rather than in a circular motion.  Alternatively -  if you use a linear rocker, do you always put the blot on the rocker in the same orientation?  If you tried rotating 90 degrees, does the bald batch rotate 90 degrees as well?

More pertinently - looking at your sample blot - and don't take this next bit personally : as a PhD student or postdoc, my PI would have absolutely crucified anyone in the lab who produced blots where even the properly-transferred and exposed lanes looked like the darker lanes on your blot in Figure 2.  

Firstly - the bands themselves are not crisp.  Is your loading buffer heavy enough (high enough glycerol or similar []) to get the samples to sit down in the wells?  Secondly - do you wash out the wells after removing the comb to get rid of the casting contaminants that get left behind, as often these are denser than running buffer so can cause the bands to be wibbly if loaded direct into this.  Finally, fuzzy or "out of focus" bands can occur from poor transfer stack construction or not great transfer buffer composition.  What is the Methanol concentration (final) in your transfer buffer?  This can be critical.  I routinely found that 20% final was necessary to get crisp transfers.

Next up : taking the shapes of the bands away and the uneven staining, looking at the darkest bands at the edges, there are uneven Rorschach ink blot outlines going on which suggests that there is glowing liquid lying on top of the membrane during exposure.  In my experience, this can happen if the blot is in Saran wrap for the exposure and has not been adequately washed of PBS containing secondary antibody - the result is that you get pools of ECL in the Saran wrinkles that glow gently and cause these uneven shapes.  Before putting the blot into the wrap after secondary washes, dunk it 2-3 times in changes of clean PBS, then drain by wicking the excess liquid away into tissue paper, before putting onto stretched Saran, adding the ECL, ensure the ECL fully covers the membrane uniformly and the excess is then removed by using a roller / swab before folding the Saran cover over and trim with a scalpel to remove excess Saran to avoid wrinkles.

Above all : don't let it get to you.  Relax, try a few different things, and don't get stressed : worse things can happen in your career than a blot misbehaving!

best of luck!

Is your antibody aliquoted and are you using a fresh aliquot each time? If not, maybe something happened to it in the meantime (e.g. heat exposure).

Are you using a rotating shaker plate to oscillate your blot, or a linear rocker (see-saw action)?  The "bald patch" in the middle of your blot sounds distinctly like the effect you can get if using a rotary shaker where a shallow spot can develop in the middle of the dish.  This in contrast doesn't happen so much if you incubate your membrane curled up inside 50ml Falcon tubes revolving on rollers, or in square flat dishes shaking back and forth linearly rather than in a circular motion.  Alternatively -  if you use a linear rocker, do you always put the blot on the rocker in the same orientation?  If you tried rotating 90 degrees, does the bald batch rotate 90 degrees as well?

More pertinently - looking at your sample blot - and don't take this next bit personally : as a PhD student or postdoc, my PI would have absolutely crucified anyone in the lab who produced blots where even the properly-transferred and exposed lanes looked like the darker lanes on your blot in Figure 2.  

Firstly - the bands themselves are not crisp.  Is your loading buffer heavy enough (high enough glycerol or similar []) to get the samples to sit down in the wells?  Secondly - do you wash out the wells after removing the comb to get rid of the casting contaminants that get left behind, as often these are denser than running buffer so can cause the bands to be wibbly if loaded direct into this.  Finally, fuzzy or "out of focus" bands can occur from poor transfer stack construction or not great transfer buffer composition.  What is the Methanol concentration (final) in your transfer buffer?  This can be critical.  I routinely found that 20% final was necessary to get crisp transfers.

Next up : taking the shapes of the bands away and the uneven staining, looking at the darkest bands at the edges, there are uneven Rorschach ink blot outlines going on which suggests that there is glowing liquid lying on top of the membrane during exposure.  In my experience, this can happen if the blot is in Saran wrap for the exposure and has not been adequately washed of PBS containing secondary antibody - the result is that you get pools of ECL in the Saran wrinkles that glow gently and cause these uneven shapes.  Before putting the blot into the wrap after secondary washes, dunk it 2-3 times in changes of clean PBS, then drain by wicking the excess liquid away into tissue paper, before putting onto stretched Saran, adding the ECL, ensure the ECL fully covers the membrane uniformly and the excess is then removed by using a roller / swab before folding the Saran cover over and trim with a scalpel to remove excess Saran to avoid wrinkles.

Above all : don't let it get to you.  Relax, try a few different things, and don't get stressed : worse things can happen in your career than a blot misbehaving!

best of luck!

We are facing the probelem recently, we got good results with the same samples previously. But we try to reproduce for publication with the same antibody and almost had same condition but results are weak. Even we have also tried three different hands, different appartus, but invain. As Nora said we also speculated the antibody freeze thawing condition. So order fresh antibody and waiting for the antibody. It will be so nice if you resolve or found the reason Chi and do let me know

Hello, do you work with colored stand marker? Did you noticed if it changed the color during the final minutes of the running stage?

Maybe you have a problem at your electrophorese apparatus and the electric current is not working homogenous along the apparatus.

Is the voltage fixed?

Here we had pH changes during the last minutes of the eletrophoreses due to the voltage and our gel "melts" and we had images like yours.

Did you keep the blocking stage at constant moving?

Isn't that band fail due to bubbles during the transfer stage?

Did you check the gel with Pounceau after the transfer?

I think these are some point to check. If you made everything right, than I don't know what is happens. Sorry and hope to helped somehow.

Good luck

Firstly, thank you for all the comments and suggestions. Let me answer the points raised:

• I keep my antibody in blocking solution at 4oC and reuses it several times. My thinking is that if the antibody failed then all the band would not show up/show up equally badly
• I use a linear rocker for antibody staining and I can see the antibody solution rolling across the entire blot. I also flip the blot around so both sides get antibody.
• My transfer buffer is from Bio-Rad and I put in 20% methanol. I also tried 30% methanol but it did not solve the problem.
• I agree that the blot show was not the highest quality and I appreciate very much the tips for making sharper bands. I have other blots with much sharper bands at the edges and no band in the middle
• During washing I would dunk the blot into fresh TBS-Tween 2 times then perform longer washes of 20 min for 3 times
• Fig 1 is a gel with color protein size markers in every well. Based on that blot I concluded that my transfer is not where the problem is
• I tried changing the transfer module and it did not work. I am using constant voltage for running the gel and transfer
• Blocking was done on the linear rocker and I checked that there were no air bubbles in the sandwich
• I have not stained the blot with Ponceau (ordered) and I will stain the gel with Coomassie after transfer to confirm nothing is left on the gel
• My current thinking is that I have put too much protein/antibody into the blot and somehow the ECL was used up in the middle within 10 sec of me putting it on the blot and capturing the image on the Chemidoc. I tried using 20X less protein and had the same result. My PI suggest trying 200X less protein and less antibody.

I'll update when I finish my latest attempt. Once again thanks for the support.

You wrote "same tube".

Are you repeatedly freezing and thawing same sample/tube?

If so, there's no mystery.  Your proteins will be michanically disruted during repeated freezing and thawing.

To avoid the issue, you have to aliquot your protein samples immediately after preparation, and keep it frozen. 

By same tube I mean every time I would thaw one sample and add the same volume of that sample to each well of the gel. So in theory every band should look the same at the end.

I see.  Sorry for my misunderstanding.

So you suspect that the antigen protein could be too much and the substrate could be consumed at the centre of the blot.

To confirm the possibility, it might be worth trying colorimetric substrates, like NBT/BCIP.  Unlike chemiluminescence, these colorimetric substates give cumulating signals.

Thanks for the answers above.  If you are getting "burn-out" of the ECL in the centre, it would be visible in the dark room immediately as you add the ECL providing you are in the dark and your eyes well adapted.  You could slow the reaction down slightly by using chilled ECL reaction mixture and cooling down your blot.

Regarding the suggestion above to use a silver stain to check if your proteins are present - if you don't have access to the silver stain reagents, the easy alternative to this is to use an imidazole-zinc reverse stain which is more sensitive (usually) than Coomassie and uses much more available chemicals than the silver stain.  After the run, rinse the gel with mQ water, incubate the gel in 200mM Imidazole for 5 minutes, rinse with mQ, then incubate the gel in 100-200mM ZnSO4.  Very quickly you should see the gel turn a milk-white colour, with the protein bands remaining colourless, visualise against a dark background.

This can then be destained by incubation with 2% citric acid in mQ for several minutes, and can then be processed for immunoblotting.

Finally - I note from re-reading the various posts above that you are using a 1-step Anti-HA-HRP conjugate antibody diluted 1:5000 (suggested range 1:1000-1:10000).  What is your blocking solution %?  We typically used non-fat milk at 10% or 5%, or 5% BSA in the buffering agent.  Do you have the option to try a traditional primary : secondary combination against HA (for instance 12CA5 against HA followed by mouse secondary)?  We routinely used 1:1000 for 12CA5, then 1:5000 goat anti mouse-HRP.  

best of luck.

Hi all just a brief update, I tried to use even less protein and antibodies. It turns out too little antibodies = no bands (Duh!) and 50X less protein then usual also did not improve the situation. I also tried to run the gel at 100V in case 200V is causing problems in the gel. Again did not work. The Bio-Rad technical support suggested more cooling during transfer by inserting a cooling block in the tank and pre-chill the buffer. So I'll try that next. I'll update with the result.

Thanks for all the comments and fingers cross!