Help Migrating bands

[camherst]

I have a problem with migrating protein bands. My main band of interest is 25kDa with the higher bands being protein aggregates in slide 1. While in the second picture (western 2) the bands are slightly to the left of the more intense bands however sometimes these bands can appear slightly below the intense main bands. Looking though it might be a doublet.
In slide 1, these bands are migrating to the left below the main band of interest. This blot was made using the iBlot by Invitrogen at 20v for 7 min (slide 1) the gel is 8-16% tris glycine gel. I also have the same issue when using the Xcell II blot module ( a semi-wet transfer unit) by Invitrogen transferring at 25V for ~2 hours in 20% methanol, 10% 10X tris glycine SDS buffer (western 2). However, it is important to note that this problem does not appear in every blot. The frequency of this problem is about 1 in 5 blots.
In both gels the high intensity band is loaded at 500ng while the control is 10 ng, I used Nitrocellulose, and using BCIP/NBT for colormetric detection. Any ideas should I decrease the concentration of protein??

Thanks for your help

[postdock]

Hiya, cool question and welcome to the forums.
Are you loading the same amount of protein in gels 1 and 2?
It seems that even though you may be loading the same amount of protein, the band intensity and the presence of extra bands changes from western 1 to western 2 (less bands).

You may be loosing some when lysing or loading, or the antibody concentrations (primary or secondary) may be different this time.
In general though, it is hard to say. I thought that the prepared gels from companies were pretty consistent however this may not always be the case.
Something to consider also is salt concentration. You should try keeping the salt concentration the same, and also make sure you load sds loading buffer into all the lanes even the empty ones and throw some into the marker lane just in case after loading the ladder.
Sometimes, the bands run in a smilie or weird curved lines because of missing sds loading buffer (or empty lanes).

Anybody else have a different opinion?