Besides possible overloading indicated above, it's possible that the later fractions might have much higher salt concentration and that affect the mobility of the proteins in those fractions.

Hi Meropi,

At first glance it looks like the later lanes are massively overloaded - you should maybe try loading around 1/50th of that amount of protein for those lanes.

Best wishes

Andy

Besides possible overloading indicated above, it's possible that the later fractions might have much higher salt concentration and that affect the mobility of the proteins in those fractions.

Hi Meropi,

I agree that the lanes are overloaded. You are right, loading the same volume is important, but you can optimize so that you are loading the same quantity of protein in the volume of your choice by adjusting the amount of water in your samples.

Also, this result can be caused by air bubbles between the gel and the membrane during transfer. Or, your gel may have been unevenly poured such that polymerization was not consistent. Something may be wrong with your buffers. You may want to check the pH of your buffers.

Good luck!

Noelle

Does it happen is multiple runs? Its possible that gel was not evenly formed and later wells tend to run like this. 

I agree with Noelle that it could be a transfer problem. Since you use a pre-cast gel, it is very unlikely to be anything wrong with that. When setting up transfer, make sure the gel is contacting the membrane tightly and flat with no air bubbles. Use a roller or pencil to squeeze all the air bubbles out, and repeat this after placing the filter paper sandwich on top.

Still another possibility is that you have those samples, where the migration is weird, contaminated with DNA that interferes with the running. Good luck!