Certainly is a stumper.

If the contaminant was in the running buffer or from the gel box, then you would expect a diffuse background, not a sharp band.

To me it looks like the contaminant has to be in direct contact with the sample prior to loading. It is not the sample buffer, so the remaining culprits are the sample tubes and loading tips.

In my post-doc we had a problem with background on Ag stained gels and it was traced back to in house autoclaving of tubes and tips. Switching to commercial pre-sterilized tubes (non-colored!), tips and strict single use of both solved the problem. 

Keratin contamination (from skin) is a common problem in SDS-PAGE. I don't know if that is what you are seeing, but if it is, there are numerous suggestions on-line about how to avoid it.

Since your ladder did not have this contaminating protein at the same intensity and you don't see this protein between the lanes, you can exclude your SDS-Running Buffer (Tris-Glycine) as a source of contamination. Since you did not have this contaminating protein when you run just empty Laemmli Buffer you can also exclude this buffer as a source of contamination. Since you use commercial Bio-Rad precast gels these gels can also be excluded. I would suggest to check all the buffers and reagents that many people share in your lab for cells washing, homogenization, lysis, and other general procedures which people are doing as a routine every day. Some of these buffer(s) or reagent(s) might be contaminated with this protein, and therefore it shows up on all of your gels. I would recommend to change all the buffers and reagents used for the preparation of the lysates for SDS-PAGE especially the ones with the long storage time. 

Hi Adam, it could be a spillover from adjacent lanes if you are loading too much volume.

Thank you all for your considered responses. In response, it appears keratin runs at

I would suggest to run some highly purified albumin (not cell lysates). If your assumption is correct you will see that artifact again (which unlikely). You also might want to check water (usually used by different labs from the same source), common cell incubator(s) for possible contamination (like mold) by using some other protein mixture instead of cell lysates, and self-prepared gels to make sure that the gels are not involved.

Certainly is a stumper.

If the contaminant was in the running buffer or from the gel box, then you would expect a diffuse background, not a sharp band.

To me it looks like the contaminant has to be in direct contact with the sample prior to loading. It is not the sample buffer, so the remaining culprits are the sample tubes and loading tips.

In my post-doc we had a problem with background on Ag stained gels and it was traced back to in house autoclaving of tubes and tips. Switching to commercial pre-sterilized tubes (non-colored!), tips and strict single use of both solved the problem. 

I will try some different (non-colored) tubes for prepping the standards. I tend to be very clean with tips and pipetting and am using my own private setup so trying different tips or vials is something I've neglected trying as I was assuming they were clean. I have remade running buffer stocks numerous times and have tried three different batches of precast gels (all Bio-Rad but different density and lot number), so I'm fairly convinced it's not coming from the gels. I used to heat my samples for 4 minutes at 95 but found it was detrimental to the transmembrane protein subunit I'm investigating right now, but I see that the artifact is in there regardless. I will try some non-colored vials and a different brand of tips. Running some purified BSA (66-69 kDa) would be easy enough to include on my next gel as well. That would test my lysis buffer and laemmli's for contamination vs. the brain homogenates I've been loading. Thanks gentlemen for your ideas. I'll report the results.

Reran the rat brain homogenate adjacent to Thermo-Pierce purified BSA standard that I get for their BCA assay kit. The protein standand ladder is Thermo's NIR PageRuler which contains as expected glycerol, SDS, and DTT along with typical purified proteins conjugated to a 700nm flurophore. I stained the gel in colloidal coomassie so protein would fluoresce at 700nm as well and scanned on a LICOR imager as before. In this case I mixed protein samples in new uncolored DNA/RNAase free vials. 50ug of homogenate was run in lane 2 and 2ug of BSA run in the same laemmli in lane 4. Every lane except the BSA shows a clear sharp banding below the 130kDa marker. The 90kDA marker does not appear. The BSA ran just below a 70kDA marker at the bottom of this image. All maker lanes, even distal from lane 2 homogenate, shows the band, suggesting its not from spillover. (20uL total volume was cleanly loaded into 30uL wells here). However, the purified single protein BSA (in laemmli) ran clear without artifact. It can't be my laemmli loading buffer. The markers come in their own buffer and I've added nothing to that. Incidentally, I used the same box of tips in making up both samples and a second box for loading both samples. Ladders were loaded with smaller tips from a third box. The smaller tips were used for measuring out the BSA and homogenate both. Yet the BSA lane is clear. This suggests its not from the tips, and new natural color vials made no difference. What is similar between the standards and homogenate but uncommon with the purified BSA? There is a trace of sodium azide in the BSA suspended in saline. I'll run something more complicated as another control. I have some normal goat serum I use as a blocking agent.

Wow. my head is getting sore for all the scratching... ;)

Hi Adam, this is sounding like "stump the chumps" segments on "Car Talk" (RIP Tom). I have one more straw to grab.

Is this band a protein? Does it also show up as a sharp blue band in visible light?

I agree with Steingrimur's opinion that the contaminant should be in direct contact with the samples prior to loading. Assuming that you loaded all the samples at equal volumes in your last picture, the much more intense band of the contaminant in the rat brain homogenate (relative to MW markers band of the contaminant) might indicate that the contaminant presence is protein dose-dependent. That points out to some common procedure that have been used for the samples before the electrophoresis. Is it possible that this band might be somehow associated with conjugation of the proteins with 700nm flurophore ? It does not look like a protein (relative to the appearance of MW standard proteins). As Steingrimur suggested, do you see it in visible light ?

Hi Adam, have you verified that this sharp band is a protein?

Hi Steingrimur, 

In regards to whether this band is protein, I can say it stains with coomassie in gel, and Ponceau S (on the blot). Also it does not generally appear in lanes with no protein loaded--except that it did not appear in the test lane with purified BSA. This anomalous band also labels well with a monoclonal anti-sodium pump antibody (in the lanes containing whole homogenate), suggesting there is some protein in there. However, in that immunoblot, the ladder lanes do not get labeled with anti-NKA but the band fluoresces albeit less brightly. I have not yet sent it off for mass spec analysis. I suppose that is next.

I suspect something like stacking gel conditions are concentrating protein together -- perhaps by precipitation. It is oddly reliable in it's MW location regardless of how long I run a gel. Always near 100kDa. Always very sharp tight band traveling precisely equal to adjacent lanes even when this band "front" has an overall curve across the width of the gel.

Hi Adam. If it shows up with Ponceau S along with the MW markers at the same sharpness and intensity as in IR, then it is a major protein in the samples you are loading.

You are not making this easy. Another straw. Blood RBCs have a lot of Na pumps and pipetters can become contaminated with blood. Can you try barrier pipette tips to eliminate that possibility?

Hi Steingrimur, my pipette set has never seen blood. I've only worked with tissue or cell culture lysates since the pipettors were new. I can certainly try a different set of pipettors. I just added some new ones to the lab that have not been used at all. I am also trying to get access to an ultracentrifuge to isolate the membrane fraction to see if this helps clean things up. I will also run some non-biotinylated samples again to see if there's a difference there. It's the most obvious thing I can imagine at this point, though avidin is 66kDa and "should" be dissociated--but might not be as I don't heat my samples. I'm wondering how semi-native avidin would run. Probably would look heavier than 66kDa. I still suspect something is odd in this band being so intense and tightly formed. It does not look real. Even the purified ladder proteins spread quite a bit in comparison. [sound of head scratching]

OK. from your previous gel

"Every lane except the BSA shows a clear sharp banding below the 130kDa marker."

Can you mix together that BSA with your sample to see if the band disappears?

I guess, you need to focus on what exactly was different in sample preparations (step by step) between purified BSA (which did not have that contaminant) and all other samples (which did). You are right, the biotinylation step might me the problem. I would also suggest to run some other protein MW markers (from other company) without any protein modification with them. I'm sure they will not show that contaminant.

Steingrimur, I will try that. Good idea. I can run it again itself next to a lane with it plus homogenate.

Viktor, I have two different ladders I have run and it occurs across both. Odd that the BSA standard did not get "contaminated" whereas the ladder's do. I think it's time to rerun some non-biotinylated lysates again. It's been a while. I should double check that. Since the purified BSA is purchased as is I have no idea their process. This is simply the Pierce standards used for BCA assays. I ran 2ug before diluted in my lysis buffer and the same purchased 2x laemmli I'm using for the other lanes. Obviously the ladder lanes don't get any laemmli or lysis dilution.

Thanks gentlemen for your suggestions.

[SOLVED] possibly...

For what it's worth to anyone, I have found a possible cause of this artifact banding (see attached images). I ran various lysates with and without biotinylation, with and without purified BSA added, empty lanes, lanes with only lysis buffer + laemmli, and ladders near and far from lanes loaded with homogenates. In the end, none of this matters. :)

This gel I decided to stop eariler than my usual, not running much low MW protein off the bottom. The artifact band(s) were there, BUT, now up at 130kDa rather than further down around 100kDa. So run duration is a modifier. This supports my suspicions that this effect is due to some change in pH or ion concentration in the upper or lower chambers during an extended gel run. Perhaps the buffer is being depleted somehow. It's as if the stacking gel effect is coming down into the separating gel.

The first image is the artifact hitting a 130kDa ladder protein (which I could see occurring during the run which motivated me to quit the gel early). The second image is the same but strongly brightened (imageJ) and some arrows added to show a very faint transition line going across the whole gel, even in between lanes. This is what I think might be the front of depleted ions coming through the separation gel and reconcentrating large MW proteins that were held back there.

Comments appreciated.

I am running into the exact same issue, but I am seeing the band at 200 kDa. I have also been using BioRad TGX pre-cast gels. Unfortunately, the band I have been probing for is supposed to run around 200 kDa so I can not differentiate a real signal from this artifact. Did you ever find a way to actually get rid of the band? Have you tried different gel percentages?

Hi Brittany,

I was using TGX gels as well, and no, I never did solve this completely. Perhaps to try to run your gel longer than you need to, letting small proteins run off the bottom. Perhaps you can at least get the artificial "band" to end up below your area of interest.

Perhaps it's something to do with the TGX gels?

I already run my gels out pretty far (so that the lowest molecular weight left is about 70-90 kDa), but I suppose I could try running it out further. I'm considering trying other gels because this is driving me nuts. I have attached an example of one of my images.

That looks exactly like what I was wrestling with Brittany.

So try the opposite--run some samples just to the bottom of the gel. I know this won't separate your bands how you want but you can atleast confirm that it the artefact band lands in a different area (or not). I haven't been at the bench recently so haven't gotten further with this issue. Maybe go back to Tris-Hcl gels or use a store-bought running buffer solution. ?? Sorry I don't have a fix for you.

Hi Brittany, are any of these lanes run with no sample, just sample buffer?

After walking through memory lane of the answers there was one thing that I did not consider. Maybe the SDS sample buffer has been contaminated.

I remember now from my post-doc days that ghost bands were appearing on everybody's gels because one grad student was using the same tip to put SDS sample buffer into samples and then dipping the same tip into the sample buffer solution to repeat the process.

He was a weird specimen. He did not want to waste tips.

That's when I decided to make my own sample buffer and let nobody else use it.